Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Investigating the linkage between business objectives and information technology objectives : a multiple… Reich, Blaize H. 1992

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1993_spring_phd_reich_blaize.pdf [ 16.81MB ]
JSON: 831-1.0086389.json
JSON-LD: 831-1.0086389-ld.json
RDF/XML (Pretty): 831-1.0086389-rdf.xml
RDF/JSON: 831-1.0086389-rdf.json
Turtle: 831-1.0086389-turtle.txt
N-Triples: 831-1.0086389-rdf-ntriples.txt
Original Record: 831-1.0086389-source.json
Full Text

Full Text

INVESTIGATING THE LINKAGE BETWEENBUSINESS OBJECTIVES AND INFORMATION TECHNOLOGY OBJECTIVES:A Multiple Case Study in the Insurance IndustrybyBlaize Homer ReichB.A., University of British Columbia (1968)M.Sc.(Bus. Admin.), University of British Columbia (1988)A THESIS SUBMITTED IN PARTIAL FULFILMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYin THE FACULTY OF GRADUATE STUDIESFaculty of Commerce and Business AdministrationWe accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIADecember 1992© Blaize Homer Reich, 1992In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature) Department of Certftedte if kentre7) 6117/10009The University of British ColumbiaVancouver, CanadaDate Deg /0, /99 ,a, .6xle,i2-..DE-6 (2/88)llABSTRACTThis research study used a multiple case methodology to investigate two questions:1) how can the degree of linkage between business and information technology (IT)objectives be measured, and 2) what factors are associated with the attainment of linkage.The linkage within multi-divisional life insurance companies was studied at both thecorporate (3 sites) and business unit (10 sites) levels.Linkage was defined as a state in which there is a high level of "mutualunderstanding between IS and business executives about each others' mission, objectivesand plans." It was measured by assessing 1) cross-references between IT and businessplans, 2) mutual understanding of current objectives by information systems (IS) andbusiness executives, 3) congruence in long term vision for IT among executives, and 4)subjective assessments of linkage by interviewees.The research model included four factors which would potentially influencelinkage: 1) shared knowledge between IS and business executives, 2) previous ITimplementation history, 3) frequency of communication between IS and businessexecutives, and 4) connections between IT and business planning processes.The data collection process was designed to enable new factors of interest toemerge from the sites. A total of 57 long, semi-structured interviews were held with 45informants. Written business and IT strategic plans, minutes from IT steering committeemeetings, and other strategy documents were collected from each of the three corporateunits and the ten business units.Data analysis indicated that both short term and long term linkage existed withinthe units. Of the four linkage measures, mutual understanding of objectives (by IS andbusiness executives) and congruence in IT vision were found to be the best measures ofshort and long term linkage, respectively. Data on these measures showed littlecongruence, indicating that short and long term linkage are two separate dimensions ofilllinkage.In the sample, Corporate IS departments provided services for two major internalclients: the company as a whole (represented by the Chief Executive Officer, and theSenior Vice Presidents of Finance, Corporate Development, and Administration), and thevarious business units. Corporate-level linkage proved difficult to isolate until linkagebetween corporate IT and corporate business objectives was separated from linkagebetween corporate IT and business unit objectives. Two of the three companies weresuccessful in achieving the first type of linkage, none was successful in achieving thelatter. No long term vision for the future of IT within the company was present in anyof the three corporate sites.Linkage between corporate IT and corporate business objectives was associatedwith: 1) the presence of written corporate objectives, 2) shared beliefs about the value ofIT, 3) communication between IS and business executives, and 4) CEO involvement inIT management.Linkage between corporate IT objectives and business unit objectives wasuniformly low in the sample companies and was associated with: 1) very low levels ofcommunication between corporate IS and business unit executives, 2) a disagreementabout the nature of linkage and 3) a "double-bind" for corporate IS departments in whichthe technology leadership requested by business unit executives was resisted by themwhen it was provided.It appears likely that the tension between corporate IS and business units couldprovide positive benefits if managed appropriately. However, it may be unrealistic toexpect corporate IT objectives to exhibit high levels of linkage simultaneously with bothcorporate objectives and BU objectives since actions taken to improve one aspect oflinkage seemed to reduce the other.Within the ten business units, there was a wide variance in achieved levels ofivlinkage, both long and short term. The factors which seemed to influence the level oflinkage attained were: 1) shared knowledge between IS and business executives, 2) asuccessful IT implementation history, 3) shared beliefs about the value of IT, and 4)communication between IS and business executives.Findings from this study must be interpreted cautiously, not only because of thesmall sample size, but also because of the factors present in the insurance industry duringthe time of data collection. This industry, for the first time in several decades, was understress. Margins were being eroded, profits were decreasing, and many of the traditionalbusiness practices were under review. In a more stable industry, the relationships betweenthe factors and linkage might have been different.This study separated factors into antecedents (shared knowledge, IT implementationhistory) and current practices (communication and connections in planning). An importantoutcome was the emergence of several other influential antecedents (e.g. shared beliefs,double binds) and the finding that antecedents seemed to strongly influencecommunication which, in turn, influenced both measures of linkage: mutual understandingand vision for IT. These findings suggest that both practitioners and researchers shouldreposition their work away from the technical aspects of IT planning and organizationalinterventions (e.g. IT steering committees) and towards the longer term issue of creatingshared knowledge and shared beliefs between business and IS executives in organizations.VTABLE OF CONTENTSAbstract ^ iiList of Tables viiiList of Figures ^ xAcknowledgements xiiCHAPTER I: INTRODUCTIONA. Why is the Study of Linkage Important? ^  1B. Research Objectives ^ 3CHAPTER II: THEORETICAL FRAMEWORKA. The Linkage Construct - Defmition and Dimensions ^  5B. A Model of the Factors Influencing Linkage  14C. An Integrated Model to Guide Research into Linkage  28CHAPTER III: RESEARCH METHODOLOGYA. The Research Model ^ 36B. The Units of Analysis  38C. Case Selection  40D. Data Collected  44E. Operationalizing the Research Model ^  52F. Data Gathering and Analysis  61G. Issues in Reliability and Validity  64H. Potential Contribution of the Research Methodology ^  67CHAPTER IV: FINDINGS CONCERNING LINKAGE and the MEASUREMENT ofLINKAGE at the CORPORATE LEVELA. Company A ^ 71B. Company B 77C. Company C 83D. Across-Site Findings  87viCHAPTER V: FINDINGS CONCERNING FACTORS WHICH POTENTIALLYINFLUENCE LINKAGE at the CORPORATE LEVELA. Factors - Company A ^ 98B. Summary and Analysis - Company A ^  109C. Factors - Company B  122D. Summary and Analysis - Company B  128E. Factors - Company C  ^ 137F. Summary and Analysis - Company C ^  147G. Across Site Findings  156CHAPTER VI: FINDINGS CONCERNING LINKAGE and the MEASUREMENT ofLINKAGE at the BUSINESS UNIT LEVELA. Business Unit 1 ^  169B. Business Unit 2  173C. Business Unit 3  177D. Business Unit 4  182E. Business Unit 5 ^  184F. Business Unit 6  189G. Business Unit 7  193H. Business Unit 8  198I. Business Unit 9 ^  202J. Business Unit 10  207K. Across-Site Findings  210L. Overall Linkage Ratings for the Business Units   224CHAPTER VII: FINDINGS CONCERNING FACTORS WHICH POTENTIALLYINFLUENCE LINKAGE at the BUSINESS UNIT LEVELA. Factors - Business Unit 1 ^  231B. Summary and Analysis - Business Unit 1 ^  238C. Factors - Business Unit 2  245D. Summary and Analysis - Business Unit 2  251E. Factors - Business Unit 3 ^  257F. Summary and Analysis - Business Unit 3 ^  262G. Factors - Business Unit 4  267H. Summary and Analysis - Business Unit 4  271I. Factors - Business Unit 5 ^  277J. Summary and Analysis - Business Unit 5 ^  283K. Factors - Business Unit 6  289L. Summary and Analysis - Business Unit 6  295viiM. Factors - Business Unit 7 ^  301N. Summary and Analysis - Business Unit 7 ^  3070. Factors - Business Unit 8  313P. Summary and Analysis - Business Unit 8  318Q. Factors - Business Unit 9 ^  324R. Summary and Analysis - Business Unit 9 ^  331S. Factors - Business Unit 10  336T. Summary and Analysis - Business Unit 10  341U. Across Site Findings ^  347V. Other Findings  360W. A Revised Model of Business Unit Linkage ^  367CHAPTER VIII: CONCLUSIONSA. Linkage at the Corporate Level ^  369B. Linkage within the Business Unit  379C. Implications for Future Research  386D. Implications for Practise  389E. Contribution of the Study ^  391REFERENCES ^  394APPENDICESA. Sample Interview Guides ^  400B. A Brief Overview of the Companies and Business Units ^ 412C. Measures of the Linkage Construct ^  429D. Scales used to Measure the Factors  435E. Reasons given by Respondents for their Subjective Ratings of Linkage ^ 438viiiLIST OF TABLESTable 11.1 Processes and Stages in the Formulation of Business and IT Strategy . . . . 10Table 11.2 Items in the Social Process Dimension of Linkage   13Table 11.3 Results of Empirical Studies on Factors Associated with Linkage and RelatedConstructs ^  19Table 11.4 Factors which may Influence Linkage ^  26Table 111.1 Interviewees within each of the Corporate Units ^ 46Table 1112 Interviewees within each of the Business Units  47Table 111.3 Archival Data Collected within each Unit of Analysis ^ 49Table 111.4 Galbraith's Lateral Relations Typology with Examples of Communicationbetween IS and Senior Executives ^  58Table III.5 A Comparison of IT Planning Typologies  60Table IV.1 Summary of Linkage ratings Between Corporate IT and Corporate BusinessObjectives ^  88Table IV.2 Summary of Linkage ratings Between Corporate IT and Business UnitObjectives  91Table V.1 A Summary of the Factors and Linkage Ratings for Company A ^ 110Table V.2 A Summary of the Factors and Linkage Ratings for Company B  129Table V.3 A Summary of the Shared Business and IT Experience in Company C ^ 140Table V.4 A Summary of the Factors and Linkage Ratings for Company C ^ 148Table VI.1 A Summary of the Linkage Findings in the Business Units ^ 211Table VI.2 Linkage Ratings for the Business Units ^  227Table VII.1 A Summary of the Factors and Linkage Ratings for Business Unit 1 • 239Table VII.2 A Summary of the Factors and Linkage Ratings for Business Unit 2 • 252Table VII.3 A Summary of the Factors and Linkage Ratings for Business Unit 3 • 263Table VII.4 A Summary of the Factors and Linkage Ratings for Business Unit 4 • 272Table VII.5 A Summary of the Factors and Linkage Ratings for Business Unit 5 . 284Table VII.6 A Summary of the Factors and Linkage Ratings for Business Unit 6 • 296Table VII.7 A Summary of the Factors and Linkage Ratings for Business Unit 7 • 309Table VII.8 A Summary of the Factors and Linkage Ratings for Business Unit 8 . 319Table VII.9 A Summary of the Factors and Linkage Ratings for Business Unit 9 . 332Table VII.10 A Summary of the Factors and Linkage Ratings for Business Unit 10 342Table VII.11 The Effects of the Shared Knowledge Factor   348ixTable VII.12 The Effects of the IT Implementation Success Factor ^ 350Table VII.13 The Effects of the Communication Factor ^  353Table VII.14 The Effects of the Connections in Planning Factor  355Table VII.15 The Effects of Emergent Factors ^  357Table VII.16 Demographic and Functional Factors in the Business Units ^ 366Table VII.17 The Relationship between the Age of the IS Department and OverallLinkage ^  367LIST OF FIGURESxA Proposed Model of Factors Influencing Linkage ^ 29A Two-Stage Model of Linkage and the Factors Influencing Linkage ^ 30The Research Model for this Study ^  38Corporate-Level Linkage ^  39^Linkage at the Business Unit Level   40Some Aspects of "Ideal" Corporate Linkage ^  70Corporate Linkage Findings in Company A  77Corporate Linkage Findings in Company B  83Corporate Linkage Findings in Company C^  87Expected Relationships Between the Factors and Linkage ^ 97Corporate IT to Corporate Business Linkage: Causal Relationships in CompanyA ^  116Corporate IT to Business Unit Linkage: Causal Relationships in Company A 120Corporate IT to Corporate Business Linkage: Causal Relationships in CompanyB ^  133Corporate IT to Business Unit Linkage: Causal Relationships in Company B 135Corporate IT to Corporate Business Linkage: Causal Relationships in Company   152Corporate IT to Business Unit Linkage: Causal Relationships in Company C 155Corporate IT to Corporate Business Linkage: Across-Site CausalRelationships   161Corporate IT to Business Unit Linkage: Across-Site Causal Relationships^  165"Ideal" Linkage at the Business Unit Level ^  167Relationship Between Two Linkage Measures: Mutual Understanding ofObjectives and Cross-References in One Year Plans   214Relationship Between Two Linkage Measures: Shared Vision for IT and Cross-References in 5 Year Plans   216Relationship Between Two Linkage Measures: Mutual Understanding ofObjectives and Subjective Assessments   218Relationship Between Two Linkage Measures: Shared Vision for IT andSubjective Assessments   219Relationship Between Two Linkage Measures: Subjective Assessments andInvolvement in New Product Development   221Figure II.1Figure II.2Figure III.1Figure I11.2Figure III.3Figure N.Figure W.2Figure IV.3Figure W.4Figure V.1Figure V.2Figure V.3Figure V.4Figure V.5Figure V.6Figure V.7Figure V.8Figure V.9Figure VI.1Figure VI.2Figure VI.3Figure VI.4Figure VI.5Figure 'VI.6xiFigure VI.7^The Relationship Between Measures of Short and Long Term Linkage:Mutual Understanding of Objectives and Shared Vision for IT .. 226Figure VII.1^Expected Relationships Between the Factors and Linkage ^ 230Figure VII.2^Causal Relationships in BU 1 ^  244Figure V11.3^Causal Relationships in BU 2  256Figure VII.4^Causal Relationships in BU 3  266Figure VII.5^Causal Relationships in BU 4 ^  276Figure VII.6^Causal Relationships in BU 5  288Figure V11.7^Causal Relationships in BU 6  300Figure VII.8^Causal Relationships in BU 7 ^  312Figure VII.9^Causal Relationships in BU 8  323Figure VII.10^Causal Relationships in BU 9  335Figure VII.11^Causal Relationships in BU 10 ^  346Figure VII.12^Business Unit Linkage: Across-Site Causal Relationships ^ 368Figure VIII.1^Corporate IT to Corporate Business Linkage ^ 370Figure VIII.2^A Revised Model of Corporate Level Linkage: Corporate IT to CorporateBusiness Objectives ^  373Figure VIII.3^Corporate IT to Business Unit Linkage ^  374Figure VIII.4^A Revised Model of Corporate Level Linkage: Corporate IT to BusinessUnit Objectives ^  378Figure VIEL5^Linkage within the Business Unit ^  380Figure VIII.6^A Revised Model of Business Unit Linkage ^ 385xiiACKNOWLEDGEMENTSThis research project was a labour of love for me, primarily because the subject oflinkage had been an important concern during my consulting career. However, even alabour of love needs to be strongly supported when it stretches over several years andconsumes many resources. Fortunately for me and for this project, that support wasprovided willingly from several sources.My advisor, Izak Benbasat, was a constant source of encouragement and reason as weentered into uncharted waters. Without his critical eye, this project would have lacked therigour that it needed. Without his support, I may have lacked the stamina to complete it.I am very much in his debt.The committee members, Al Dexter and Larry Moore, were very conscientiousreaders and supportive of my wordy attempts to "show evidence" . They both have a keeneye for the nuances of organizational life and I benefitted from their perspectives.The three organizations who participated in this study were most generous with theirtime and their confidence. I am grateful to them all, and most specifically, to thecorporate IS executives who facilitated the data gathering process and spent many hoursto help me understand their history.My husband, Allan, and children, Stephen and Geoffrey, bore the major brunt of thisproject as I regularly failed to turn up for dinner and spent many weekend hours in frontof the computer. Although this effort went well beyond their understanding of"homework", they provided understanding and love at the critical times. They, however,will be most glad that it is over. Not only will I be more available, but my sense ofhumour may return, too.I thank you all.1INVESTIGATING THE LINKAGE BETWEEN BUSINESS OBJECTIVESAND INFORMATION TECHNOLOGY OBJECTIVES:I. INTRODUCTIONA. Why is the Study of Linkage Important?In recent surveys of information systems managers (Niederman et al., 1991; IndexGroup, 1988; Brancheau and Wetherbe, 1987; Dickson et aL , 1984), Information Systems(IS) Planning has consistently been rated as one of the most important issues. Since ISplanning is a complex activity which includes forecasting, evaluating, prioritizing, andallocation of resources, it is important that an investigation into IS planning determinefirst what aspects of planning are problematic. A review of the empirical literature revealsthat one issue, the linkage' of IS plans with organizational objectives, is always amongthe top few problems reported by IS managers and planners (Galliers, 1987A; Ledererand Mendelow, 1986).There is both theoretical and empirical support for the concerns of IS managersregarding their ability to link IS and organizational plans. Emery (1969), in his generaltheory of organizations as systems governed by a hierarchy of plans, remarks that "Thebehaviour of an organization ultimately depends, of course, on the composite activity atthe lowest level. However, if the organization as a whole is to achieve purposefulbehaviour... lower-level activity must be guided by a hierarchy of higher-level planningconstraints. Otherwise, lower-level success tends to be local rather than global" (p. 124).Davis and Olson (1985) support this objective in an IS context, stating "A very important1 Several terms other than "linkage" are used in the literature - alignment (Galliers, 1987; Henderson andVenkatraman, 1989), fit (Venkatraman, 1989; Das, Zahra and Warkentin, 1991), and coordination(Lederer and Mendelow, 1989). Because the underlying concepts are very similar, we will not distinguishbetween them and will use "linkage" exclusively.2fundamental concept of information system planning is that the organizational strategicplan should be the basis for the MIS' strategic plan. Alignment of MIS planning is oneof the central problems of MIS planning" (p. 446).The importance of linkage was corroborated empirically by Cresap, McCormickand Paget (1983), who found that linkage of IT objectives with business objectives wasan important predictor of success in IS planning.Evidence from surveys suggests that most organizations have not been verysuccessful in linking their IS and business plans. Galliers' study (1987A) of IS plannersin the UK inquired about the extent of the link between IS planning and corporateplanning. Of the 129 respondents, 8% reported that the plans were "totally isolated",34% were "tenuously linked", 48% were "somewhat linked" and 10% were "inextricablytied". These results are very similar to another UK study in which Earl (1987) reportedonly 41 % of the 42 respondent organizations as linking their IT plan with their businesslong-range plan. Cresap, McCormick and Paget's (1983) findings in their U.S. studyvaried widely by industry. However, an average of 29% of respondents replied "appliesprecisely" when asked if their "IS Plan refers to Business Plans". Of the remainder, 55%replied "applies somewhat", and 16% replied "does not apply". A similar percentage wasfound in a large Canadian study (Conrath et al., 1992/93), in which 31% of respondentsclaimed to base their IS plans on corporate plans.Although the need for linkage has been established and companies report lowsuccess rates in attaining it, there are few studies of how companies perceive the linkageissue or how they actually organize and act to achieve linkage. The authors suggest that2 In this document the acronyms IT (for Information Technology), MIS (Management InformationSystems) and IS (Information Systems) are used in two ways. When referring to other research, theacronyms used in those articles are used unchanged. We use "IT" holistically to mean the technology,systems, and people in the industry or in a particular organization and "IS" to refer to the functionaldepartment in an organization which oversees the IT environment.3a more comprehensive model of the linkage construct might help to explain currentfindings and guide the design of new studies to investigate it.B. Research ObjectivesThe specific research objectives were to:1) Develop a theoretical framework of the factors influencing linkage,2) Define and develop ways to measure the linkage construct,3)^Describe how several organizations have achieved or failed to achievelinkage as influenced by the factors in the framework.It was felt that achievement of these objectives would increase our knowledgeabout linkage and the factors which influence it. The theoretical framework wasdeveloped to guide future investigation and the development of a measurement theoryabout the linkage construct was designed to lay the foundation for the creation of validand reliable instruments.Underlying this research is the assumption that achievement of high levels oflinkage between organizational and IT objectives is beneficial to a significant number offormal organizations'. This study makes no attempt to identify or quantify these benefits.Organization of this DocumentIn Chapter II, a definition of linkage and a survey of the literature which pertainsto it are presented. A research model is developed after dimensions of the linkage3 McFarlan et al. (1983) suggest that linkage is less important in organizations for which IT, either nowor in the future, is not a strategic resource. This view is considered to be valid by many researchers andis included in this research approach through the selection of sites to be studied. This is discussed in the"selecting the cases to be studied" section of Chapter III.4construct have been explored. In Chapter III, the research approach is described and thedetailed data collection protocol is presented.Chapter IV and V present findings on linkage at the corporate level, chapter VIand VII present findings on linkage at the business unit level. These four findingschapters are long because of the decision made to present evidence in detail, includingquotes from interviewees and quotations from written documents. Readers wishing to"fast track" through the findings should skip to the Across Site sections at the end ofchapters IV through VII. Chapter VIII summarizes the findings and presents implicationsfor research and practice.5II. THEORETICAL FRAMEWORKThis chapter will first discuss the linkage construct and propose a definition. Thenthe factors which are hypothesized to influenced linkage will be reviewed. From thisreview, a theoretical model of linkage is developed.A. The Linkage Construct - Definition and DimensionsIn order to build a multi-dimensional view of linkage, we will 1) use concepts fromthe strategic management literature to define a broad range of connections betweenbusiness and IT, 2) differentiate between the "cause" and "outcome" views of linkage,and 3) distinguish between the "intellectual" and "social" processes in the formulationand assessment of linkage.1. Developing a Strategic Management View of LinkageIn the IT literature, the concept of "linkage" emanates from the IT planningperspective and authors suggest that IT "plans" should the linked to other artifacts inbusiness - business plans (Lederer and Mendelow, 1986; Conrath et. al, 1992/93),business strategies (Pyburn, 1983) or business objectives (Galliers, 1987A; Zviran, 1990).While these terms (i.e. plans, strategies and objectives), are considered interchangeableby some IS researchers, they are carefully differentiated by strategic managementresearchers.Although there are many variations in the strategic management literature, we willbegin by using the following terms (definitions adapted from Thompson and Strickland,1990):Missions are long term visions of what an organization seeks to do and what kindof an organization it intends to become.Example: Become the predominant U.S. manufacturer of knitwear.6Goals are specific measurable performance targets (i.e. the "ends") an organizationseeks to produce through its activities and the competitive position the enterprise wishesto occupy in the market.Example: Increase market share to 25% by 1995.Strategies are the are the approaches (i.e. the "means") which will be used topursue the goals.Example:IT Goal = Reduce the application backlog to 20 person-years by 1995.IT Strategy = Have the users develop systems using a 4GL.Plans are the detailed roadmaps of the direction and course that the organizationpresently intends to follow in conducting its activities.Example:IT Strategy = Have the users develop systems using a 4GL.IT Plan Item 1 = Install the Software AG 4GL product.IT Plan Item 2 = Write the investment system in the 4GL as a pilot.There is usually not a single goal/strategy pair created before plans are made. Ahigh level goal (i.e. increase ROI profitability by 10%) can be attained in several ways(i.e. reduce cost, increase revenue, reduce investments). Once it is chosen, the strategyof "increase revenue" becomes a goal, when quantified. This goal too can be achievedin many ways. And so an "ends/means" hierarchy can exist above any item in the plan 4 .An item in a plan can potentially support more than one goal and, therefore, theconnections between plan items and goals and/or strategies are not tree-like, but mayform a network.Because of this complex ends/means hierarchy, we introduce another definition tosimplify the discussion and subsequent research:Objectives are the aggregate of the goals and strategies of an organizational unit.Objectives denote the intentions of the organizational unit, representing both the4 We are not suggesting that these ends/means hierarchies would be explicitly developed in allorganizations but that, upon reflection, a senior manager could reconstruct ends/means chains to explainthe rationale behind any significant item in the current organizational plans.7future performance targets and the means selected to achieve them.Because the IT literature is not consistent in describing what the IS plans shouldbe linked to, we will take our direction from the strategic management literature, andlook at the broadest possible set of linkages between the IS function and the business.'Our preliminary definition of linkage will be "the degree to which the IT mission,objectives, and plans reflect the business mission, objectives, and plans."' This departurefrom the more simple notion of linking IT and business plans may increase thecomplexity of measurement of the linkage construct but will allow the maximum possiblerange of investigation.A more precise definition of the meanings of the word "reflect" used in the linkagedefinition are developed in the next two sections. We argue for the definition of linkageas a "state" rather than as a "process" and we identify two dimensions of it - an"intellectual" dimension (i.e. "linkage as a high level of fit between IT and businessstrategies") and an "social" dimension (i.e. "linkage as a high level of mutualunderstanding' between IS and business executives).2. Distinguishing Causes from Effects in LinkageIn the IT Literature, there has been little discussion of whether linkage is an5 There is some support for a broad conceptualization of the linkage construct. Lederer and Mendelow(1989), extending King (1978), suggest that "coordination is achieved when the IS strategy set (systemobjectives, constraints, design strategies) are derived from the organization's strategy set (mission,objectives, strategy)."6 We make the distinction between mission, objectives and plans here since they become different sub-dimensions of the linkage construct later in this paper.The idea of "mutual understanding" was introduced by Churchman and Shainblatt (1964) in the contextof the interaction between management scientists and managers. They suggested that the relationshipbetween the scientist and manager was central to any theory of implementation of scientific managementtechniques and warned that "we believe intensive research is required to discover its (i. e. mutualunderstanding) real implications".8organizational process or is an outcome resulting from these processes. Webster's ThirdNew International Dictionary describes linkage alternately as a process ("the manner orstyle of being fitted together or linked") and as an outcome ("the quality or state of beinglinked"). Some empirical studies have measured it as a process (e.g. Conrath et al.(1992/93) asked "Is your EDP/MIS long range plan modeled after the corporate plan?")and others have measured it as an outcome (e.g. Galliers, 1987A asked "How close is thelink between the IS plan and the corporate plan?".)In our view, certain organizational processes lead to the outcome of "beinglinked" . We therefore consider the organizational processes as potential "causes" oflinkage and the outcomes as the "effect" (i.e. linkage itself). Since our objective in thisresearch is to create a model of the factors which influence linkage, it is necessary toseparate out cause from effect and to define linkage as an outcome of the variousorganizational processes which may influence it. This approach leads us to deconstructsome of the previous frameworks developed to investigate linkage.The first attempt to dimensionalize linkage is found in the accounting literature(Shank et al., 1973). They suggested that business plans and budgets could be tightly orloosely linked, depending on three characteristics: 1) content linkage (between the plansand budgets), 2) timing linkage (between the planing and budgeting systems) and 3)organizational linkage (between the people doing the planning and budgeting). These threedimensions were adopted by IT researchers (Lederer and Mendelow, 1989), who statedthat "coordination... can be achieved in three dimensions - content, timing andpersonnel" .The only study of IS planning which used multiple items to measure linkage(Cresap, McCormick and Paget, 1983) also reflected this three-part view. They createdthe following five items of which four can be directly related to the Shank et al.'s (1973)terminology:91. The business plan states information system needs. (content linkage)2. The IS plan makes reference to items in the business plans. (contentlinkage)3. IS plans are closely checked against business plans.4. Line and staff managers participate actively in information systemsplanning. (organizational linkage)5.^Business planning calendars and IS planning calendars are carefullysynchronized. (timing linkage)These studies, while explicitly recognizing the multi-dimensional nature of linkage,did not separate the efforts made toward achieving it (organizational and timingdimensions, which we view as causes of linkage) from linkage as an outcome (contentdimension). After separating these two ideas, we have:- the "organizational" and "timing" dimensions as potential influences on linkage.(e.g. who is involved, when is planning done).- the "content" dimension as a way to objectively measure linkage (CresapMcCormick and Paget (1983) suggest looking at written plans).3. Differentiating Intellectual from Social Process DimensionsIn the mainstream strategic management literature, there is a clear distinction madebetween the formulation and the implementation' of strategy. Formulation pertains to thecrafting of strategy and implementation to the realization of strategy. This research isdealing exclusively with the issues in the formulation of IT and business mission,objectives and plans.Another important distinction can be made between the issues relating to contentand those relating to style. This distinction was identified by Horovitz (1984) wholabelled the two dimensions "intellectual process" and social process". The intellectualprocess refers to the particular methodologies, techniques and data used in the formulation8 We are indebted to Professor Raphael H. Amit, University of British Columbia, for bringing thisdistinction to our attention.Table II.1Processes and Stages in the Formulation of Business and IT StrategyProcesses \Stages Influences on Linkage(CAUSE)Linkage(ElIFECT)What methodologies forformulation of IT and businessmission, objectives and plans wereused and what were the steps andtechniques?A high quality set of IT andbusiness mission, objectives andplans that are complementary.Intellectual ProcessChoice of actors, timing, andsupport for IT and businessstrategy formulationHigh levels of organizationalunderstanding of the mission,objectives, and plans.Social Process10of strategy. The social process dimension refers to the choice of actors and levels ofinvolvement and support given to the chosen methodologies. Put simply, the distinctionis between strategy formulation techniques and styles.After combining the idea of cause and effect and the distinction betweenintellectual and social processes, we have a framework within which to build dimensionsfor the study of linkage. The preliminary framework is presented in Table II.1Chapter III of this paper discusses the various factors to be included in the"Influences on Linkage' part of the framework. The next section discusses the intellectualand social dimensions of linkage.4. Dimensions of Linkagea) Linkage as the Result of an Intellectual ProcessThere are very few empirical studies centred around the intellectual process issuesin the formulation of business and IT linkage. Zviran's study (1990) identified separateIT and business strategies. It reported the presence of 16 IT objectives in the 13111companies which responded to the survey/questionnaire. Tavakolian (1989), found astrong relationship between generic business strategies (i.e. defender, reactor, prospector)and the structure of the IS function - the more aggressive the business, the moredecentralized the IS function.Although these early studies are inconclusive, strategy research into otherfunctional areas such as production and marketing is much better developed and is anindication of the importance of this intellectual, content-oriented dimension of linkage.One possible "intellectual" dimension of linkage is the concept of"comprehensiveness" from the strategic management literature (Fredrickson, 1984;Henderson and Venkatraman, 1992). This is the defined as "the extent to which anorganization attempts to be exhaustive or inclusive in making and integrating strategicdecisions" . However, our separation of causes from effects places "comprehensiveness"into the set of factors which might influence linkage. A recent article (Henderson andSifonis, 1988), suggests that strategic plans need internal consistency and externalvalidity. In the linkage context, these notions would result in two aspects of theintellectual dimension of linkage:1) Business and IT planning outputs are internally consistent. (i.e. the IT mission,objectives and plans chosen are consistent with the given business mission andobjectives). 92) Business and IT planning outputs are externally valid (i.e. they are comprehensive andbalanced with respect to external data from the business and IT environments).These two aspects will be used in the preliminary model of linkage.Cresap et al. (1983) suggested a measure for this dimension: "the degree to which the written IS andbusiness objectives and plans reference each other". We also, as explained in the next section, expectarticulated plans and objectives (i.e. those recounted by managers in interviews or questionnaires) to bemutually referencing in situations characterized by high linkage.12b) Linkage as a Result of a Social ProcessThe preoccupation in the IT literature with issues of process is understandable onthree counts - 1) the historical position of the IS function as a second string player(Hossack, 1989), 2) the need for strong organizational processes in order to movesmoothly and effectively into implementation phases of strategic management, and 3) thedifficulty of tracing IT strategy through to organizational outcomes. The focus for ISmanagers, and hence for most researchers, has been on creating a social process to linkthemselves to the formulators of corporate strategy, rather than investigating the businessoutcomes of strategic choices.On the social process dimension, linkage has been identified as a high level ofmutual understanding by organizational members. In the previous section, we noted thepossibility of looking into the wording of the written business and IT plans to identifyunderstanding. This approach, while intuitively pleasing since it is amenable tooperationalization, poses a significant practical problem - the absence of written IT and/orbusiness plans. Several studies (Lederer and Mendelow, 1986; Calhoun and Lederer,1990), have reported that detailed, written business plans are often not produced. Sincethere is no theoretical reason why a lack of written business plans means that linkage isnonexistent, we need to search for other sources and manifestations of plans andstrategies.One other source of business objectives is the "minds" of senior managers. If theIT objectives are highly linked to the objectives and plans of the business, we suggestthat:1) the senior IS executive would understand and be able to articulate the businessobjectives and2) business executives would understand and be able to articulate the current ITobjectives.13It also seems possible that executives may lack understanding of current ITobjectives, while an over-arching mission or vision unites the organization and provideslinkage. If both business executives and IS executives share a common vision of the wayIT will support the business mission', one might predict that decisions made within thisshared knowledge will be congruent and "linked" . Therefore, another proposed dimensionof linkage is the existence of a common vision of the future role of IT within theorganization.These items in the social process dimension can be characterized along two axes -their source and their timeframe. These are depicted in Table 11.2.Table 11.2Items in the Social Process Dimension of LinkageTimeframe\Source ofLinkageIS Function Business UnitCurrentThe 1-3 year plans andobjectives.IS executives understand currentbusiness objectives.Business executives understandcurrent IT objectives.Future(The 3-10 year mission andvision).IS and business executives sharea common vision of the role andcontribution of IT to theorganization's mission.IS and business executives sharea common vision of the roleand contribution of IT to theorganization's mission.5. SummaryOur discussion to this point would result in a definition of linkage as the state inwhich the set of IT and business mission, objectives and plans are:1) of high quality and complementary and2) are well understood by executives in the e.g. "IT will allow our company to place significant decision making responsibility in the hands of ouragents and clients in support of our vision of superior service."14The linkage construct would have the following dimensions:Dimension I (Intellectual Dimension) - The quality of the set of business and IT mission,objectives and plans:1. High internal consistency2. High external validityDimension II (Social Dimension) - Mutual Understanding between IS and businessexecutives1. Mutual understanding of and commitment to current business and IT plans andobjectives.2. Shared vision of future IT role and contribution to the business mission.B. A Model of the Factors Influencing LinkageIn this section, the literature which directly and indirectly relates to linkage as anoutcome will be reviewed and used to formulate a model of the potential factorsinfluencing linkage.1. Factors derived from research into LinkageIn the only study in which linkage was the dependent variable, Lederer andMendelow (1989) interviewed 20 IS managers. They found that a mandate from the CEOwas a significant enabler of linkage. Inhibitors of linkage were 1) lack of a stable, clearbusiness plan, 2) lack of communication between IS and business executives, 3) IS notbeing involved in business planning, and 4) unrealistic expectations of users. Thepresence of an IS steering committee or the reporting relationship of the IS director didnot seem to mitigate these problems.152. Factors derived from empirical research into the dimensions of IS planningSeveral recent studies which have investigated the factors affecting the quality ofIS planning (ISP), included dependent variables which are associated with the linkageconstruct as defined earlier in this paper.Raghunathan (1985) theorized that ISP is done at three levels : strategic ISP,systems planning, and implementation planning. Using these planning dimensions, hedeveloped, through exploratory factor analysis, 11 dimensions of ISP. Two of thestrategic ISP dimensions, 1) the level of strategic content in the ISP process and 2)integration of the IS function into the business, directly relate to our definition of linkage.Later studies (Raghunathan and Raghunathan, 1988; Gupta and Raghunathan, 1989)reported that top management support and the presence of an IS Steering Committee wereassociated with high levels of these two ISP dimensions.In a survey of 192 IS managers, Raghunathan and Raghunathan (1989) reportedthat the rank of the most senior IS manager and increasing use of an IS SteeringCommittee (as opposed to decreasing or no use) was significantly related anorganization's ability to have its IS plan "linked to organizational concerns". Jang (1989)created a uni-dimensional measure of the extent of business strategy in the ISP processby asking respondents whether or not their ISP process included certain activities,including 1) top management involvement in ISP, 2) IS attendance at senior businessmanagement meetings and 3) user managers' involvement in ISP. He found that the extentof business strategy in the ISP process could be modeled as a function of IS access tosenior management and maturity of the IS function". Environmental volatility was alsosignificantly positively related to the extent of ISP.Calhoun and Lederer (1990), in a study of eighteen matched pairs of IS and11 This was defined as a function of sophistication in data processing, telecommunications and officeautomation technologies.16business managers, found no relationship between the quality of the business planningprocess and the extent' of IS planning. They did, however, find a significantcorrelation between communication of the business plan (to IS managers) and the extentof IS planning.3. Factors derived from research into IS planning successSeveral studies investigated factors associated with ISP success. Although there isno detailed examination of the "success" construct in any of them, our belief (supportedby Cresap et al., 1983) that direct measures of ISP success will implicitly includemeasures of linkage suggest that these studies are relevant to the search for factorsinfluencing linkage.Cresap et al. (1983) measured ISP success using two items: "business programsare assured of needed IS support" and "scarce IS resources are allocated wisely".Approximately 50% of the sample of 334 companies reported ISP success based on thesecriteria. They found a strong relationship between certain business planning practices,namely wide distribution of the business plan, commitment of top management toplanning, perceptions that the plan was realistic, monitoring performance based on thebusiness plan, and ISP success. They found no relationship between the choice of ISPmethodology' and ISP success.Pyburn (1983), in an in-depth analysis of seven companies, created a typology ofpotentially successful planning styles, labelling them "verbal/informal" , "written/formal" ,and "written/informal" . He found that the success of any planning style is contingent on12 The "extent" of IS planning was measured by asking about the content of the written IS plan (i.e.whether or not it included objectives and strategies, a portfolio approach to applications, and financialprojections for each time period in the plan).13 e.g. Business Systems Planning (IBM, 1974); Critical Success Factors (Rockart, 1979); Stages ofGrowth (Nolan, 1979).17several moderator variables, such as senior management style, and status, proximity ofthe IS manager to business executives, complexity of the IT environment and volatilityof the business environment.Sullivan (1985), after analyzing data from 35 firms, found that successful planners(as represented by effectiveness ratings) matched the choice of ISP methodology' to thelevel of infusion and diffusion of IT in the company. Infusion was defined as the strategicvalue of IT to the company. Diffusion was a function of the distribution of IT technologyand the responsibility for IT decisions throughout the business processes of theorganization.Vitale, Ives and Beath (1986), found two factors to be highly correlated withreported satisfaction with strategic ISP processes. There was a positive correlation with"the existence of IS-knowledgeable line managers" and a negative correlation with"environmental turbulence". To measure planning effectiveness, Waibel (1987) createda 12 item scale which primarily measured the comprehensiveness of the ISP process, notthe outcome of the ISP effort. He found a significant relationship between the presenceof an IS Steering Committee and planning effectiveness. There was no relationshipbetween planning effectiveness and top management presence on the Steering Committeeor top management support for ISP.Raghunathan and Raghunathan (1990) found that higher levels of top managementsupport and comprehensiveness in IS planning were positively associated with the"fulfilling ISP objectives" dimension of their ISP success measure.Lederer and Sethi (1990) in a survey on implementation of strategic IS plans,found that, after more than half of the planning horizons had been passed: 1) only 24%of planned projects had been initiated, 2) 38% of projects begun since strategic IS14 see previous footnote.18planning were not in the plan, 3) only 50% of recommended changes in the ISorganization were done. The reported level of strategic ISP implementation was the bestdiscriminator between ISP satisfaction and dissatisfaction.Galliers (1987B) surmised that failures of linkage might partially be explained bythe fact that business planners were seldom involved in IS planning and only 6% oforganizations conducted a formal review of the IS plan, although they had ranked thisactivity as being very important.The findings from these empirical studies, grouped by independent variable, aredisplayed in Table 11.3. The dependent variables are not directly comparable and thefindings are not consistent between studies. However, there are indications that highlevels of top management support, increased use of IS steering committees and certainbusiness planning practices (such as distribution of the business plan and commitment toplanning) are associated with higher levels of linkage (and surrogates of linkage).Table 11.3Results of Empirical Studies on Factors Associated with Linkage and Related ConstructsINDEPEND'TVARIABLE(IV)DEPENDENTVARIABLE (DV)CITATION FINDINGS(relationship between IV and DV)EnvironmentalturbulenceISP SuccessSatisfaction with ISPExtent of business strategyin the ISP processPyburn (83)Vitale et al. (86)Jang (89)The combination of volatile environments and informal ISPstyles are associated with ISP success.Turbulence was negatively related to satisfaction.Turbulence was positively related to the dependent variable(DV).IS Maturity Extent of business strategyin the ISP processJang (89) IS maturity was significantly related to the DV.BusinessPlanning (BP)SystemISP SuccessProblems in Implementingthe IS PlanComprehensiveness of theIS PlanLinkageCresap, McCormickand Paget (83)Lederer and Sethi(90)Calhoun and Lederer(90)Lederer andMendelow (89)Certain business planning practices (distribution of plans,realistic plans, commitment to planning, monitoring performanceto plans) are important to ISP success.Tactical business planning results in a higher severity of ISPproblems than strategic business planning.1) No relationship between the quality of business planning andthe DV.2) Significant relationship between the high levels ofcommunication of the business plan and the DV.Lack of a business plan and lack of IS management involvementin business planning inhibits linkage.Table 11.3Results of Empirical Studies on Factors Associated with Linkage and Related ConstructsINDEPEND'TVARIABLEON)DEPENDENTVARIABLE (DV)CITATION FINDINGS(relationship between IV and DV)ISP System(the socialprocesssurrounding ISplanning)ISP SuccessLinkageLinkagePyburn (83)Lederer andMendelow (89)Gainers (87B)ISP styles must match the characteristics of an organization andits environment.Unrealistic expectations of users inhibit linkage.Low levels of reported linkage may result when businessplanners do not get involved in IS planning.ISP Metho-dology (BSP vsCSF vs SOG)Fulfilment of ISPObjectivesISP SuccessISP SuccessRaghunathan andRaghunathan (90)Cresap, McCormickand Paget (83)Sullivan (85)The level of capabilities of the ISP methodology is positivelyrelated to the fulfilment of ISP objectives.No relationship was found between the use of a particularmethodology and ISP success.Organizations which reported success in ISP had methodologieswhich matched the level of diffusion and infusion of IT in theorganization.IS-knowledgeableLine ManagersSatisfaction with ISP Vitale et. al (86) There was a significant relationship between having IS-knowledgeable managers and satisfaction with IS planning.Table 11.3Results of Empirical Studies on Factors Associated with Linkage and Related ConstructsINDEPEND'TVARIABLE(IV)DEPENDENTVARIABLE (DV)CITATION FINDINGS(relationship between IV and DV)TopManagementSupportFulfilment of ISPObjectivesExtent of Strategic ISPLinkageComprehensiveness of theIS planning process.Raghunathan andRaghunathan (90)Raghunathan andRaghunathan (88)Lederer andMendelow (89)Waibel (87)Top management support was significantly related to fulfilmentof ISP objectives.A positive relationship was found between the level of topmanagement support and the emphasis placed on strategic ISP.A mandate to link IS and business plans is important in order toachieve linkage.No relationship between top management support for ISP andcomprehensiveness of the IS planning process.Rank of theCIOLinkageISP is linked toorganizational concernsISP implementationproblemsISP SuccessLederer andMendelow (89)Raghunathan andRaghunathan (89)Lederer and Sethi(88)Pyburn (83)No relationship was found between the rank of the CIO andlinkage.A significant relationship was found between the rank of theCIO and the DV if the CIO is one rank below the CEO.The lower the CIO in rank, the more severe were the ISPimplementation problems.The combination of a high status IS manager and an informalplanning style was associated with ISP success.NTable 11.3Results of Empirical Studies on Factors Associated with Linkage and Related ConstructsINDEPEND'TVARIABLE(IV)DEPENDENTVARIABLE (DV)CITATION FINDINGS(relationship between IV and DV)Presence of anIS SteeringCommitteeComprehensiveness of theIS Planning ProcessLinkageExtent of Strategic ISPISP is linked toOrganizational ConcernsWaibel (87)Lederer andMendelow (89)Gupta and Ra-ghunathan(89)Raghunathan andRaghunathan (89)The presence of an IS Steering committee is related to ISPeffectiveness.No relationship was found between the presence of an ISsteering committee and linkage.A significant relationship was found between the presence of anIS steering committee and the extent of strategic ISP.An IS steering committee is related to ISP which has strongerlinks to organizational concerns.TopManagementPresence on theSteeringCommitteeComprehensiveness of theISP ProcessWaibel (87) No significant relationship between having top management onthe steering committee and comprehensiveness of the IS planningprocess.Access to andCommunicationwith ExecutivesLinkageExtent of business strategyin the ISP processISP SuccessLederer andMendelow (89)Jang (89)Pyburn (83)Lack of communication with top management inhibits linkage.Ease of access by the IS executive to the CEO is related to theDV.Proximity of the IS manager to the CEO is important in thesuccess of informal planning styles.Table 11.3Results of Empirical Studies on Factors Associated with Linkage and Related ConstructsINDEPEND'TVARIABLE(IV)DEPENDENTVARIABLE (DV)CITATION FINDINGS(relationship between IV and DV)Implementationof the ISPSatisfaction with ISPLinkageLederer and Sethi(90)Galliers (87B)The higher the perceived implementation of the IS plan, thehigher the satisfaction with ISP.Low levels of reported linkage may be attributed to lack offormal review of IS plans by business executives.244. Factors derived from the Prescriptive/Theoretical LiteratureThere have been several prescriptive and descriptive articles on ISP in the recentIT literature. Those which add to our understanding of linkage are described below.A key theoretical factor which might influence linkage was proposed by McFarlanet al. (1983). They suggested a contingency framework called the Strategic Grid whichseparated companies into four quadrants based on the strategic value of their current andfuture IS and claimed that different IS strategies were needed for different quadrants. Inparticular, they stated "for companies where IS is and should be in the support role, ....less effort needs to be made to ensure alignment of IS and corporate strategy" . Themanagement techniques prescribed for companies in the "strategic" quadrant include 1)devotion of significant management time to the IT planning process, and 2) closeorganizational relationships between IT and senior management (Cash et al., 1988). Froma descriptive point of view, organizations in different quadrants of the strategic grid mayexhibit different levels of linkage.' From a prescriptive point of view, achieving linkageis more important for organizations in the strategic quadrant.A recent article by Das, Zahra, and Warkentin (1991) proposed a model whichlinked strategic IT planning and business strategy and related it to competitive advantage.It distinguished between the "content" and process" of strategic IT planning, anddiscussed the importance of integrating these dimensions in order to increase the "fit"between strategic MIS planning and business strategy. The model presented in our paperextends the content/process discussion into the definition of linkage: "intellectual process"and "social process" become the two dimensions of linkage.King (1978) developed the earliest planning model which explicitly linked strategicIS planning to organizational strategy. This model is composed of 1) the organizational'Recent research into this hypothesis by Takuna and Weber (1991) has reported "moderate support forthe usefulness of the strategic-grid model in accounting for some types of IS planning phenomena" .25strategy set (i.e. a mission, objectives, strategy for each claimant group), 2) the MISstrategic planning process and 3) the MIS Strategy Set (i.e. system objectives, constraints,and design strategies). Henderson and Sifonis (1988) suggest that IS planning is anattempt to create an internally consistent and externally valid set of means-ends chainswhich gradually become concrete. Their three stage planning model includes 1) strategicbusiness planning, 2) strategic IS planning and 3) IS investment planning. In their model,the outputs of the business planning process (i.e. vision, critical success factors, andgoals) are fed into the strategic IS planning process, which identifies the set of criticaldecisions, assumptions and business processes. These are used to prioritize IS projectsand create an IS plan.For both the King and the Henderson and Sifonis models, the existence of astrategic business plan and planning process is a prerequisite to the creation of linkage.5. A SummaryUsing both the empirical and prescriptive literature, we have placed the factorswhich have the potential to influence linkage into five groups as shown in Table 11.4. Ingrouping these factors, there were several choices which might be considered debateableand are somewhat arbitrary. These are discussed below.The External Influences group contains factors which exist at the level of theenvironment or industry and have been shown to influence the success of IT planning andthe extent of business strategy in the IT planning process. One major factor is thevolatility and turbulence in the environment of an organization. Another factor of interestis the current use of IT in the industry (i.e. the "strategic grid" position of the industry).We have chosen to place the strategic grid factor at the organizational level (it is includedin the IT Characteristics group) since significant variations in the "perceived" strategicgrid position may exist within industries.26Table II.4Factors Which May Influence LinkageFactor Group Independent VariableExternal Influences -environmental turbulence-complexity of the IT environmentCharacteristics of IT inthe Organization- maturity of the IT function- perceived strategic grid position- infusion and diffusion of IT in the organization- IS-knowledgeable line managersConnections between theIT and BusinessPlanning Systems- availability of the business plan to IT planners- timing between IT and business planning- particular methodologies or techniques which connect theISP and the business planning processes- involvement of senior management and business planners inISP- involvement of the senior IS manager in business planning- review and approval of the IT plan- monitoring the progress of the IT and business planCommunication betweenIS and BusinessExecutives- access/proximity of the senior IS manager to seniormanagement- top management support for IS- status/reporting level of the senior IS manager- presence, composition and usage of the IS SteeringCommitteeImplementation ofPrevious IT Plans- level of success in implementing the IS planThe Characteristics of IT in the Organization group exists primarily to allowresearchers to classify organizations as to the importance and the incidence of IT withinthem. Maturity, diffusion and infusion of IT, and perceived strategic grid position are thecharacteristics which have been shown to be significant in previous studies .The Connections between IT and Business Planning Systems focuses quitenarrowly on those activities which take place during strategic business and/or ITplanning. Not all organizations will engage in these activities; many organizations willdo strategic planning only sporadically, in reaction to significant events. This group27encompasses the facets of business and IT strategic planning processes which serve to"connect" them, namely their timing, the availability of shared information, andinvolvement of specific individuals in both processes.The Communication between Business and IS Executives factor group containsthose factors which, taken together, would describe the level of everyday interactionbetween IS and senior business executives. These include physical proximity, reportingrelationships, and involvement on committees and task forces. The "steering committee"factor is included in this group, although it might also be considered part of the"connections between the IT and business planning system" group. It was placed under"communication" based on our belief that steering committees are primarily used asvenues for communication and mutual education, not for strategic planning. In otherwords, steering committees may exist in an organization which did no strategic planning.A factor which was ill-defined by previous research and therefore difficult to placewas "top management support for IS". It was placed under "communication", since wefelt that the primary indicators of top management support, in addition to a largeoperating and capital budgets, would be increased communication with IS executives andhigher reporting relationships for IS executives.Although the "IS knowledgeable line managers" factor was identified by only oneempirical study, it seemed important since our model defined linkage as "mutualunderstanding" . Although it might be important enough to stand alone (i.e. as a"characteristics of executives in the organization" group), it has been placed in the"communications" group due to our belief that its influence would be primarily on thecommunication between IS and business executives.The Implementation of Previous IT Plans factor group represents the set oforganizational memories and perceptions about the success of IT within the business.Specifically, it would measure the satisfaction with IS, both from a project perspective28(i.e. did projects go in on time and on budget?) and from an organizational contributionperspective (i.e. did projects contribute to business success?).In the next section, these five factor groups are used to create the integratedresearch model and propositions concerning their influence are developed.C. An Integrated Model to Guide Research into Linkage1. The ModelA model incorporating potential factors identified in the broad survey of ITliterature and the linkage dimensions developed in the previous section is shown in FigureII.1.Although organizational factors often are recursive in their causality(Jang,1989), 16 there are several factor groups which logically would precede particularbehaviours that might be observed in an organization. Recognizing the existence of thislag between events and outcomes, we have organized the factors into the two-stage modelshown in Figure 11.2 - separating Antecedents from Current Practices. One of the itemsin the Communications factor group, IS-knowledgeable line managers, has been separatedout as an Antecedent. Observed linkage is the dependent variable. In general, Antecedentsare expected to affect linkage indirectly, by affecting Current Practices. Current Practicesare expected to directly affect Linkage. The expected contribution of each of the factor16 In the only reported test of causality in this area, Jang (1989) used path analysis to show the bi-directional nature of causality. In his data set, while external factors (IS maturity, volatility of theenvironment, size of the organization) and internal factors (style of top management, status of ISmanager, IS access to senior executives) influenced the extent of business strategy in the IS planningprocess, the IS planning process also significantly influenced the internal IS factors.EXTERNAL INFLUENCESIT and business environment)CHARACTERISTICS OF ITIN THE ORGANIZATIONCOMMUNICATION BETWEENIS AND SENIOR EXECUTIVESCONNECTIONS BETWEEN IT AND [.._BUSINESS PLANNING SYSTEMSDOLBMIDITIAION OF THEIT PLANLINKAGE BETWEEN IT ANDBUSINESS MISSION,OBJECTIVES, AND PLANS.1.MUTUAL UNDERSTANDINGBETWEEN IS AND BUSINESSEXECUTIVES.(current, future)2.HIGH QUALITY SET OF ITAND BUSINESS MISSION,OBJECTIVES AND PLANS.(internal, external)Figure 11.1A Proposed Model of the Factors Influencing Linkagegroups is expressed as a set of researchable propositions as follows. The next chaptercontains the model which guided this research.a) External InfluencesIn general, external influences such as environmental turbulence and complexityare expected to affect linkage indirectly by influencing Current Practices.29CHARACTERISTICSOF rr iN THEORGANIZATIONCOMMUNICATIONBETWEEN BUSINESSAND IS EXECUTIVESCONNECTIONSBETWEEN BUSINESSAND IT PLANNINGCURRENTPRACTICESOBSERVEDLINKAGEHIGH LEVEL OFMUTUALUNDERSTANDINGCOMPLEMENTARY,SET OF BUSINESSAND IT OBJECTIVES30ANTECEDENTSFigure 11.2A Two-Stage Model of Linkage and the Factors Influencing LinkageIn a turbulent IT' and business environment, one might expect that companies(in which IT is of critical importance) 18 will:Proposition I: have their IT and business planning systems tightly connected (e.g.through timing, exchange of documents, use of certain executives as members of both17 In this discussion, there are several levels at which the abbreviation "IT" is used. First, there is an ITindustry or environment, composed of suppliers and customers of information technology and services.Secondly, there is the internal IT function of an organization, composed of the IS personnel andinformation technologies. One of the systems in an organization is the IT planning system, composed ofthe people and procedures used when an organization creates its plans for the IT function.18 We expect that "perceived strategic grid position" would moderate the relationships described inProposition I and II.31processes), although both may be short term in nature, andProposition II: exhibit high levels of communication between IS and businessexecutives in order to keep each other informed of changes in their respectiveenvironments which impact their objectives.b) Characteristics of IT within the OrganizationAs the prevalence and importance of IT grows within an organization, planningand communication practices are affected. For example, it has been found that mature ISfunctions (measured in number of years) have more formal and therefore morestrategically focused IT planning processes (Doll and Torkzadeh, 1987; Raghunathan andRaghunathan, 1989). We expect that factors such as IS maturity, diffusion of ITthroughout the organization, and the perceived strategic grid position will influencelinkage by affecting Current Practices.Proposition III: Organizations in which IT is mature, highly diffused, andperceived to be central to organizational objectives would exhibit higher levels ofconnection between their business and IT planning systems.Proposition IV: Organizations exhibiting high levels of diffusion or infusion of ITwithin the organization will exhibit high levels of communication between IS and businessexecutives.Within organizations, barriers between functional areas (e.g. marketing, IT) andbetween roles (e.g. staff, line) can inhibit understanding and cooperation. An importantfactor in explaining ISP success was found (Vitale et al., 1986) to be the presence of IS-knowledgeable line managers. If line executives within an organization have experiencein IT management and IS executives have experience in line management, we believe thiswill lead to a higher levels of communication.Proposition V: Organizations in which there is a high level of shared experience32among the senior business and IS executives will lead to high levels of communicationbetween them.c) Success in the Implementation of Previous IT PlansA good, but unimplemented, IT plan obviously contributes little to an organization.Subsequent planning efforts may be weakened since organizational members have lowexpectations of the outcome and therefore contribute few creative ideas. At this stage inIT research, these ideas (the effect of history on current behaviours) have not beenempirically tested. However, we believe that success (or failure) in implementingprevious IT plans will have an indirect effect on linkage by affecting other componentsin the model. Previous success is conceptualized as having at least two levels: operational(i.e. IT projects delivered on time and on budget) and strategic (i.e. IT projects seen tocontribute to important business objectives).Proposition VI: A recent history of strategic IT success will lead to tighterconnections between IT and business planning.Proposition VII: A recent history of operational IT success will lead to increasedlevels of communication between IS and business executives. Conversely, a history ofoperational IT failure will lead to reduced levels of communication between IS andBusiness executives.d) Communication between IS and Business ExecutivesWhen members in an organization meet frequently, in both informal and formalsettings, barriers between functions and roles can be broken down, leading to higherlevels of mutual understanding and trust. Executives will be more inclined to share theirplans and strategies with executives they know well and have frequent contact with.Proposition VIII: High levels of communication will lead to high levels of33connection between the IT and business planning systems.'Proposition IX: High levels of communication will lead to high levels of mutualunderstanding of objectives and similar visions of the future role of IT.e) Connections between the IT and Business Planning SystemsCertain IT and business planning processes and behaviours (e.g. participation ofIS in business planning, participation of senior management in IT planning, interchangeof planning documents) will increase the volume of information shared between IS andbusiness executives and thus will directly influence linkage.Proposition X: High levels of connection between business and IT planningsystems will lead to a high quality, complementary set of IT and business plans.Proposition XI: High levels of connection between business and IT planningsystems will lead to high levels of mutual understanding and a shared vision for ITbetween IS and business executives.2. Literature CritiqueFrom our investigation into the literature pertaining to linkage, we concluded thatresearch into linkage suffers from some important problems. Most of these can becategorized as 1) lack of theoretical foundation, or 2) inappropriate research designs.Each will be briefly discussed below.The lack of theory or even a theoretical framework has meant that research intolinkage lacks focus. Certain variables of interest (e.g. rank of the CIO, presence of thesteering committee) have been investigated but these have not been identified as part ofa comprehensive theory. The lack of theory has made it difficult 1) to interpret results19 We also believe that the "perceived" position on the strategic grid will moderate this relationship. Themore important that IT is perceived to be by business executives, the stronger the relationship will be.34which are contradictory to earlier findings and 2) to hypothesize backwards from thestatistically significant findings. For example, if the rank of the CIO is positivelyassociated with higher levels of strategic IS planning, what influences the rank of theCIO? In addition, the lack of research interest in the linkage construct itself has resultedin the absence of a clear definition and a theory of measurement.The preponderance of current research into IS planning in general and linkage, inparticular, has been done via large survey studies or smaller interview-based studies.Many of the research designs have lacked the reliability and validity components whichare important when conducting organizational level research. The most obviousshortcomings are: 1) use of single respondents to investigate organizational variables',2) exclusive use of the IS manager as a survey respondent as opposed to senior linemanagers, and 3) use of single item scales to investigate linkage.Other studies have included a mix of organizations, without testing for theperceived strategic value of IT in each. This oversight makes interpretation of linkagestatistics meaningless if one believes that the strategic grid theory has some validity.Another important weakness in the linkage research is the lack of longitudinal studies.For practitioners, as well as researchers, causality is important to understand and thecross-sectional nature of the current work makes it impossible to discern. For example,does the presence of a highly ranked CIO follow or precede high levels of strategic ISplanning? The findings to date give little direction to a manager who wonders whatactions to take.This research project has been conceived as a first step in filling some of the gapsin our knowledge about linkage and about the factors which influence it. By examiningseveral organizations in depth (paying attention to their history and the multiple points of2° The one notable exception to this generalization is Pyburn's (1983) case research which provided uswith a rich perspective on organizational life.35view within them), and creating "organizational stories" about their attempts at linkage,we hope to show the interrelationships between the factors in these specific cases and toidentify additional ways in which linkage can be defined and measured. By comparing the"stories" and drawing out commonalities among them, we hope to create grounded theoryconcerning the factors which influence linkage.In the next chapter contains a discussion of the way we chose to study linkagewithin organizations and will identify the strengths and weaknesses of this approach.36III. RESEARCH METHODOLOGYA. The Research ModelIn Chapter II we proposed a two-stage model to guide investigation into the linkageconstruct. This research project has operationalized parts of this model. We have focusedon measurement of the "social process" dimension of linkage and identification of thefactors which influence it. We have chosen a sample which allows us to hold constantseveral of the factors identified in the two-stage model and thereby to simplify theinvestigation. Within this limited research scope, we have tried to be as comprehensiveand holistic as possible with our data collection practices and analysis.Our decision to focus on the social process (understanding and commitment ofmanagers) dimension of linkage stemmed from our belief that too much attention hadbeen paid to the dissection of the other dimension (methods of planning and contents ofplans) and not enough to the very real problem of ensuring that managers understood andagreed upon a common strategy and vision for IT. Calhoun and Lederer's (1990) researchsupports this view, identifying communication of the business plan, rather than the contentof the business plan, as influencing the comprehensiveness of the IT planAnother reason to focus on the social process was the relative immaturity of theresearch into the typology of IT strategies and missions (i.e. the intellectual contentdimension). The exception is Zviran's study (1990) which reported the presence of 16 ISobjectives in the 131 companies which responded to his survey. As yet, no attempt hasbeen made to investigate whether the list is complete or if the items are mutuallyexclusive. No work has been done to match these IS "objectives" with IS "missions" andno empirical study has yet matched different sets of IT objectives to various business37objectives'.Our sampling strategy was chosen to minimize variance on two of the factorspotentially influencing linkage: External Influences on the Organization, andCharacteristics of IT in the Organization'. The resulting model which guided our studyis shown in Figure III.1. This model, since it was created from a wide body of ISresearch, should represent a comprehensive approach to the internal organizational forceswhich have the potential to affect linkage. In the next sections, the plans for investigatingthe groups of factors in the model will be discussed.Bonoma (1985) suggests that phenomena which meet the dual conditions of littletheoretical knowledge and high complexity should be suited to the application ofqualitative research methods. As discussed previously, little theory exists about linkagein organizations. It can also be argued that linkage is inherently a complex phenomenasince it is an organization-level construct, influenced by current processes and historicalevents.The development of the model shown in Chapter II from the findings andprescriptions in a wide body of IS literature positions us at the beginning of Bonoma's"design" stage of research into linkage. Collection of data to investigate the groups offactors in the model should allow us to "assess and refine" the major areas of inquiry.However, as Bonoma cautions, we "must be willing to let further data recycle ourthinking if beginning conceptualizations do not hold up against new situations" (Bonoma,1985).This study has been designed as a case research project. The following sections21 Tavakolian (1989), however, found a strong relationship between generic business strategies (defender,reactor, prospector) and the structure of the IS function. Chan and Huff (1992) are currently investigatingthe relationship between generic business and IT strategies and business outcomes.v One element in this factor group, shared knowledge between business and IS executives, could not beheld constant and has been investigated in this research.COMMUNICATIONBETWEEN BUSINESSAND IS EXECUTIVESCONNECTIONSBETWEEN BUSINESSAND IT PLANNINGCURRENTPRACTICESOBSERVEDLINKAGEHIGH LEVEL. OFMUTUALUNDERSTANDINGANTECEDENTS 38SHARED KNOWLEDGEBETWEEN BUSINESSAND IS EXECUTIVESIMPLEMENTATIONOF PREVIOUSIT PLANSFigure III.1The Research Model for this Studydescribe the selection criteria and specific features of the research design. Following thisis a discussion of the methods used to enhance the reliability and validity of the researchand a summary of the contribution that this research approach can bring to the study oflinkage.B. The Units of AnalysisBecause our definition of linkage revolves around business and IT mission,objectives and plans, we need to carefully identify the possible locations of these artifacts.The companies in our study are multi-divisional organizations with each division havingsome responsibility for profit and loss and all divisions having responsibility to forge theirCORPORATEBUSINESSOBJECTIVES BUSINESSUNITBUSINESSOBJECTIVESCORPORATEITOBJECTIVES BUSINESSUNIT ITOBJECTIVESFigure 111.2Corporate-Level Linkage39own unique set of mission, objectives and plans within corporate guidelines. Weidentified two distinct loci of linkage within multi-divisional organizations: corporatelinkage (in which the corporate IT mission, objectives and plans need to be aligned withthe business), and business unit linkage (in which business unit (BU) IT mission,objectives and plans need to be aligned with the business). Each linkage is forged indifferent ways.We defined corporate linkage as that which is created by aligning corporate ITmission, objectives, and plans with the business mission, objectives, and plans from bothcorporate management and from each of the business units, as shown in figure 111.2.Corporate linkage, as defined, is complex. In an organization with many business units,the corporate IT mission, objectives, and plans should reflect each of their requirementsin the context of the overall corporate mission, objectives, and plans.40We defined business unit linkage as that which is created by aligning IT andbusiness mission, objectives, and plans within a single business unit. This is shown infigure 111.3. 23CORPORATEBUSINESSOBJECTIVES BUSINESSuNrrBUSINESSOBJECTIVESCORPORATErrOBJECTIVES BUSINESSUM' ITOBJECTIVESFigure 111.3Linkage at the Business Unit LevelC. Case SelectionWe selected three large Canadian life insurance companies for this researchproject. This selection of organizations was made to reflect Yin's (1989) strategy of'3 Please note: in the diagrams, we have used the word "objectives" to stand for the mission, objectivesand plans in the unit. This has been done for purposes of brevity, not to emphasize objectives over otherplanning outputs.41"literal replication" in which all cases are theoretically the same. In order to make theorganizations more comparable, the cases selected minimized differences along the twogroups of factors which were not being investigated, namely External Influences andInternal IT Characteristics.In order to minimize variance from External Influences, we chose organizationswhich were established businesses, operating for profit in a competitive marketplace.These organizations offered similar products within similar markets (across Canada).Their asset bases respectively are 1, 10 and 20 Billion $Can. They all have Canadianheadquarters and a large national network of agents and brokers. All companies sellproducts in the United States and other countries but our study investigated only theCanadian operations because of logistical problems investigating foreign operations.In each of the companies, data were collected at two levels: the company as awhole (to investigate corporate-level linkage) and within several business units (toinvestigate business unit linkage). Each business unit had responsibility to set its ownstrategic goals and plans, within a corporate framework.In selecting life insurance companies with similar products (e.g. life and healthinsurance for individuals and groups, retirement investments), we were ensuring that allcompanies and business units would theoretically be in the Strategic or Turnaroundquadrant of the Strategic Grid' model, and therefore, would have similar aspirationswith respect to linkage between business and IT mission, objectives and plans. Becausethey were in the insurance industry, all units satisfied some important criteria discussedin the strategic IS literature, including: very high levels of information intensity in boththe products and their distribution, and continual evolution of new products and new waysto use information in the industry (i.e. the information technology usage has not yet'4 According to the Strategic Grid model (McFarlan et al., 1983), companies in the Strategic andTurnaround quadrants have plans to use IT as a major enabler of their strategic business plans.42peaked).However, within different organizations, data are often interpreted in unique ways.Selecting organizations who "should" believe that IT is strategic to their future is notenough to ensure that they will in fact feel this way and therefore act to ensure thatlinkage is obtained within their organization. Without extensive interviewing, anorganization's attitude towards IT can be discerned by looking at factors such as ITbudget, composition of the IT Steering committee and the reporting relationship of thesenior IS manager. In general, the bigger the budget, the more senior the businessrepresentation on the IT steering Committee, and the higher the IS manager reports, themore likely it is that the company takes its investment in IT seriously. To ensure that the"perceived" position on the strategic grid was comparable among companies, we applieda filter during the selection process. We looked for companies which satisfied one ormore of the following criteria:a) A large IT budget (in the insurance industry, this means > 15 % of operatingexpenses).b) The President/CEO chaired the IT Steering Committee.c) An organizational structure in which the senior IS executive reports to the President.Although budgets are not easily comparable, all of our companies spend more than15 % of their operating expenses on IT. In the two companies which had IT SteeringCommittees, they are chaired by the President/CEO. In all of the companies, the seniorIS Executive reported to the Senior Vice President/Chief Financial Officer, not thePresident.Another criteria which indicated to us that the company believed in the strategicvalue of IT was its willingness to participate in this study. We were asking for access tostrategic business and IT documents and interviews with most of the senior executives inthe company. By agreeing to participate, we felt that each of our companies exhibitedbeliefs that IT was strategic to their success.43In order to minimize variance on the Internal Characteristics of IT within theOrganizations, we chose companies which had been stable on a macro level (mergers,acquisitions, major expansion within Canada) for at least five years. However, at abusiness unit level, such stability was hard to find and several of our business units haverecently been formed or have been restructured. These changes will be noted in ourdescriptions and analysis.The IT units within the organizations ranged from being relatively mature (ISemployees with 5 years of company experience) to very mature (IS employees with 20years of company experience). However, all senior IS executives had been in the industryfor at least 10 years.Another goal of our case selection strategy was to maximize the variance inattained linkage between the companies and business units. This was difficult to do a-priori since we did not yet have sophisticated instruments to measure linkage. The firstcompany was found through personal connections of the primary researcher. The secondwas suggested by a colleague. After interviews with the IS manager and the ChiefFinancial Officer in both companies, we felt that they would exhibit low-moderate levelsof attained linkage. The third company was selected after extensive probing of industrycolleagues to determine which companies were considered to be exemplars in the use ofIT. We invited three of these companies to participate and when one accepted, we weresatisfied that we had a set of units which would exhibit a variance on the linkageconstruct which was available in Canadian industry.The life insurance industry, with respect to its use of IT, is quite mature. Most ofthe back office (policy enrolment, underwriting, actuarial, claims) is fully supported byinformation systems. The front office (sales and product management) is more unevenlysupported but agents have had presentation software to aid their sales efforts for severalyears. Most life insurance companies are operating in an extremely competitive44marketplace where IT is used both to reduce unit costs and to create new and innovativeproducts and marketing support environments.Within these three Canadian life insurance companies, we selected 10 businessunits for analysis. They were; 3 units selling individual life insurance, 2 selling groupinsurance, 2 offering retirement assets, 1 selling reinsurance services, 1 doing internalinvestments and 1 selling life insurance for automotive sales. In order to keep ourpromise of confidentiality to these companies, we will not be identifying any of thebusiness units by name, nor will we be associating any business unit with any corporateunit.The companies who participated in our study will be named company A, B and C.the business units have been mixed into random order and numbered from 1 to 10.D. Data Collected1. InformantsBecause the data to be collected was qualitative and largely post-hoc, it wasimperative to collect different types of data to explore the topic and verify any finding(Jick, 1979). There were two data collection methods used in the study: gathering ofwritten archival data and interviews with informants. In this section, we identify thepeople who were interviewed and the questions they were asked; in the next section welist the archival data collected.It was important for two reasons to include a wide constituency of informants.First, by interviewing a wide range of informants, we were able to gather data about allaspects of an event or a process, thus increasing the reliability of our data. Second,having multiple informants at the same level in the unit of analysis increased our abilityto identify differences of opinion within organizations and to investigate any outcomes ofit. Earlier studies which used single respondents did not have this opportunity.45The research model included three factors which potentially influenced the levelof attained linkage: the history of IT implementation, communication between IS and lineexecutives, and the connections between business and IT planning. In order to gather dataabout these factors, we identified the roles which would have data pertinent to ourinvestigation. These included: the senior business executives, the senior IS executives,the members (if any) of the IT Steering Committees, and any technological "gatekeepers" .To the best of our ability, we included people from all these roles in our interviewschedule.A list of the interviewees is presented in Table 111.1 for the corporate level analysisand in Table 111.2 for the Business Unit analysis. In total, we conducted 57 interviewswith 45 informants. To investigate corporate level linkage, we interviewed both corporateand business unit executives. Because Senior Vice Presidents have responsibilities atcorporate and at business unit levels, each interview with these individuals was done froma corporate as well as a business unit perspective. In each unit except BU 4, we were ableto include at least three informants who had relevant information about the factors fromour model.With most executives, one long interview (approximately two hours) wasconducted. Some short follow-up interviews were also held as necessary. The Seniorcorporate IS executives were interviewed between three and five times since they werethe people most knowledgeable about the IT history within the organization.A generic set of interview guides is presented in Appendix A. As shown, separateguides were prepared for the President/CEO, the IS managers, the business unitexecutives and other participants in IT. The actual interview guides used were customizedto use correct organizational titles and to ask about specific events which the author hadidentified previous to the interview. This customization process is discussed in the nextsection.46Table III.1Interviewees within each of the Corporate UnitsUnit ofAnalysisCorporate Business Executives Corporate IS Executives Business UnitBusinessExecutivesCompanyA- President/CEO- Senior Vice President (SVP)/Chief Financial Officer- SVP/Chief Investment Officer- Vice President (VP)Information Services (IS),- 2 Assistant Vice Presidents(AVP), IS,- 2 Directors, IS- 2 SVPs ofBusinessUnitsCompanyB- President/CEO- SVP, Corporate Services- VP/Controller- Director, Corporate Planning- SVP (retired) Corporate Services- VP, Corporate IS- VP, Corporate Operations- 2 Directors, Corporate IS- 2 SVPs ofBusinessUnitsCompanyC- President/CEO- SVP, Corporate Services- VP, IS - 3 SVPs ofBusinessUnits2. Archival DataArchival data was used in three ways during the study. Before any interviewingcommenced, planning procedure manuals, organization charts and annual reports wereused to give the researcher background knowledge concerning significant organizationalbusiness and IT events. They were also used to customize the interview guides withcorrect organizational titles and to guide the selection of interviewees.During the process of interviewing (which took several weeks), minutes frommeetings, strategy documents, and consultants reports were used to add specific questionsto the interviews concerning important events and outcomes and to enlarge the set ofinterviewees. This allowed the researcher to collect holistic data about organizational"stories" with contributions from many sources.47Table 111.2Interviewees within each the of the Business UnitsUnit ofAnalysisBusiness UnitBusiness ExecutivesBusiness UnitIS ExecutivesBusiness Unit1- SVP, Business Unit- VP, Marketing- VP, Admin. and IS- Director, Admin. SystemsBusiness Unit2- SVP, Business Unit- VP, Marketing andAdministration- VP, Finance- VP, Admin. and IS- Director, ISBusiness Unit3- VP, Finance- Director, Administration- Director, Marketing- Director, IS- Manager, Systems DevelopmentBusiness Unit4- VP, Operations - Director, ISBusiness Unit5- SVP, Business Unit- Manager, MarketingAdministration- Manager, Systems DevelopmentBusiness Unit6- SVP, Business Unit- Director, Finance- Manager, Admin. and ISBusiness Unit7- VP, Administration- AVP, Marketing- AVP, ISBusiness Unit8- SVP, 2 Business Units- VP, Business Unit- AVP, ISBusiness Unit9- SVP, 2 Business Units- VP, Business Unit- AVP and Controller- AVP, ISBusiness Unit10- SVP, 2 Business Units- VP, Business Unit- AVP, IS- Manager, ISAfter the interviewing was complete, strategic plans were analyzed to assess one dimensions oflinkage "the extent to which written IS and business plans reference each other". The minutes fromSteering Committee meetings and other strategic documents were used to assess the workings of these48liaison groups such as the level of actual participation, and the strategic level of the meetings. Minutesfrom weekly and monthly managers meetings provided us with data to corroborate interview data oncommunications between IS and business executives. For example, we could make rough calculationsconcerning the frequency with which IS concerns were addressed at these meetings and we could identifythe amount of time that an IS manager was exposed to other executive's views and concerns and viceversa.A detailed list of the archival data collected for each unit of analysis is presented in Table 111.3.We were successful in obtaining IT and business plans from each unit in the study (if they had preparedthem). In some cases, we collected plans from previous years since the current year plans were in draftform or otherwise incomplete.Table 111.3Archival Data Collected within each Unit of AnalysisUnit ofAnalysisBack-groundBusiness Plans IT Plans Steering Committee (SC)Minutes etc.Other Strategy DocumentsCompany A - 5 Annual - Strategic 5 yr - 1991 IT Plan - SC mandates - 1991 Planning Processreports plan - 1990 IT Plan - 10 sets of minutes (senior Document (2 levels)- 1991 Annual - Strategic management SC) - PC Acquisition PolicyBusiness Plan Technologydirection paper- 12 sets of minutes(technical SC)- Authorities &Responsibilities manual re IT- Presentation to SC on datamanagement, and rapidapplication development- IS Chargeback PolicyCompany B - 2 Annual - 5 year planning - 1991 IT Plan - SC mandate statement - minutes from IT planningreports directive 1991- - 2 years of SC minutes meetings1995- Five year plan(technical SC) - reports from a reorganizationof corporate IS units1990-1994 - letter from President to ISdirectorCompany C - 4 AnnualReports- Not produced - 1991 IT Plan - 2 years of SC minutes(senior management SC)- 1989 IS Strategic Planningretreat minutes- 1986 IS plan- 1985 IS Direction StatementsTable 111.3Archival Data Collected within each Unit of AnalysisUnit ofAnalysisBack-groundBusiness Plans IT Plans Steering Committee (SC)Minutes etc.Other Strategy DocumentsBusiness Unit1- AnnualReports- 5 yr Strategy1991-1995- 5 yearOperating Plan1991-1995- 1991 Plan- no separate 5 or 1yr IT planproduced.- projectsembedded in 1991Plan.- unpublished 5 yr.technologystrategies.- no BU steering committee - minutes from 2representative monthlyplanning meetings of the BUexecutiveBusiness Unit2- AnnualReports- Strategic Plan1990-1995- 1990 BusinessPlan- 1991 IT Plan - no BU Steering Committee - Various Reports on SystemsStrategy (1984, 1987, 1989,1991).Business Unit3- AnnualReports-Backgroundsections,BusinessStrategy, 1991-1995- 1991 Plans- no 5 yr or 1 yr ITplan compiled.- projectsembedded in the1991 Plan.- no BU Steering Committee - 2 memos re systemsprioritiesBusiness Unit4- AnnualReports- excerpts from1991 StrategicPlan (notfinished)- 1990 BusinessPlan- IT Planningdocuments, 1989,1990- no 1991 IT Plan- no 5 yr IT plan- no BU Steering Committee - Mission, goals, objectives forIS, 1989- Strategic Planning output,1990Table 111.3Archival Data Collected within each Unit of AnalysisUnit ofAnalysisBack-groundBusiness Plans IT Plans Steering Committee (SC)Minutes etc.Other Strategy DocumentsBusiness Unit - Annual - Strategic - no unified IT - no BU steering Committee5 Reports Marketing Plan strategy1991 - 1991 ProjectsBusiness Unit - Annual - Administration - IT Plan, 1991 - no BU Steering Committee6 Reports Plan, 1991Business Unit - Annual - Strategic Plan - IT Strategy 1990- 6 months SC minutes - "Impact of IT on Insurance7 Reports 1991-1995. 1995 1991-1995" document- 1991 Business - 1991 IT PlansPlan - 1991 ProjectsBusiness Unit - Annual - Strategic Plan - 1991 Planning - 5 months SC minutes - 1991 IS Staff Review8 Reports 1991-1995 Summary- Operating Plan - 1991 Key1991 ProgramsBusiness Unit - Annual - Strategic Plans - IT Plan, 1991 - 3 months SC minutes - Technology Directions, 19909 Reports 1990-1995,1991-1996- 1991 PlanBusiness Unit10- AnnualReports- Strategic Plan1991-1995- Business Plan,1991- no 1991 IT Plan- projects imbeddedin 1991 BusinessPlan.- 2 months SC minutes - Systems Strategy Report,1988c5%52E. Operationalizing the Research Model1. Observed LinkageTo operationalize the linkage construct (defined as the "degree of mutualunderstanding between IS and business executives about the IT and business mission,objectives and plans"), we needed to consider: 1) whether to capture perceptions orobjective measures or both, 2) whether to use quantitative or qualitative measures of the"degree" of linkage, and 3) how to investigate both the "current" and "future" dimensionsof linkage. These questions will be discussed below.a) Perceptions vs. Objective MeasuresOne of the purposes of this research is to discover whether there are any objective,outcome-based ways to measure linkage. However, because much of organizational realityis constructed by individuals and groups, it did not seem reasonable for us to createpurely objective measures without verifying that the perceptions of organizationalparticipants would be congruent with them.For example, a researcher might categorize an industry as being in the "Strategic"quadrant of the model developed by McFarlan et. al (1983). However, if influentialorganizational participants do not perceive themselves as being a "Strategic Quadrant"company, their actions may not be congruent with the prescriptions for that kind ofcompany. In this case the perceptions and objective measures would show no congruence.Neither measure is right or wrong, they measure two different aspects of the sameconstruct, the actual and the perceived. In this project, we are attempting to developobjective measures which can be reconciled with perceptions and, therefore, we53investigated linkage using both objective and perceptual measures.'b) Quantitative or Qualitative Measures of LinkageSince there may be multiple IS and business objectives in a given business unit,we can construct both quantitative and qualitative measures of the degree of linkage. Forexample, we might say that 7 out of 10 written IT objectives reference businessobjectives and use this statistic in a comparative analysis to rate this organization as more"linked" that an organization in which only 4 out of 8 IT objectives referenced businessobjectives. However, without a theory which identifies the nature of IT objectives and thepercentage that we would expect to reference business objectives, this use of intervalmeasurement scales is meaningless. Therefore, our preliminary measures of linkage wereordinal. Their construction is described below and the scales are contained in AppendixC.c) Measuring Current and Future LinkageBecause no study had looked comprehensively at the linkage construct, we createdseveral measures of the construct, expecting that others would emerge and that some orall of our pre-designed measures would be impractical or redundant. The measures usedincluded:1) Mutual understanding of current objectives, measured by assessing whether ISmanagers could articulate the business objectives and whether business managers couldarticulate the current IT objectives. The data was gathered during interviews and analyzedAs Lee (1991) suggests, there are three objectives which a researcher may pursue during data collectionand analysis: - to identify the "constructed reality" of the participants, to make qualitative interpretationsguided by scales developed by the researcher, or to provide quantitative assessment from instruments suchas surveys or experiments. In this study, we spent considerable effort to ascertain the participants"reality" and to make our interpretations based on this reality and on our own independent assessments.54using a simple scale as shown in Appendix C.i.e. does the IS manager understand the near-term business objectives are? Do thesenior managers know what the IT objectives are?2) Cross-references in written plans, measured by assessing the degree to which IT andBusiness plans reference each other. Two measures of cross-referencing were taken - onefrom the one year plans and one from the 5 year plans. An ordinal scale (see AppendixC) was created for this measure.i.e. do the IT plans reference specific parts of the business plans, such as businessobjectives or current problems? Do the business plans indicate how IT might helpin achieving organizational objectives?3) Shared vision for the future role of IT, measured by comparing the open-ended visionstatements obtained in interviews from IS and business executives and assessing theircongruence.i.e. do senior managers and the IS manager share the same vision of IT in thefuture, specifically with respect to:- the usage of Information Technology (i.e. processors, communicationtechnology, database technology, expert systems..) in the business- the use of Information Systems to automate certain business functions- who will make IT-related decisions- the future structure and size of the IS function4) Subjective ratings of linkage, taken from all interviewees. Each manager interviewedwas asked to rate linkage (at the business unit or corporate level) as high, moderate orlow and to give rationale and evidence supporting the rating (see Appendix C). This datawas used as the "perceptions" measure and also allowed us to explore their subjectiveunderstanding of what linkage was.552. Shared Knowledge between Business and IS ExecutivesIn organizations where managers have a deep understanding of each other'sposition, communication is likely to be more efficient. Less communication may beneeded to accomplish organizational goals than in organizations in which basicmisunderstandings cause differences of opinions and take time away from discussion ofcontent issues. Understanding is fostered, in part, when people have similar workexperiences (e.g. by being in the same industry, the same organization or in similarroles). Each of these similarities gives the participants a "shared language"; a shorthandmethod of describing situations and objectives.The primary way organizations create shared work experiences is throughtemporary and permanent transfers, such as job rotation and special assignments. Sinceparticipants in each case selected for study were already in the same industry andorganization, evidence of similarities in experience were made by gathering data abouttemporary and permanent transfers from general management to IT management and viceversa within the organization. Evidence of three types of similar work experience werecollected: line management transfers'', project management roles, and membership inliaison roles. A nominal scale was developed (Appendix D) to reflect the amount ofmanagerial experience each executive had in the opposite domain. Specifically, data wascollected to identify:a) How many of the current executives have:- held direct responsibility for the IS function (i.e. was the senior ISmanager),- chaired an IS steering committee/ ISP project/large Development project,- acted as a member of an IS Steering Committee/ISP project/largedevelopment project.This measure reflects the permanent lateral transfers which have occurred in the organization.According to Galbraith (1977), lateral transfers increase the amount of direct communication betweenvarious functions. So this variable may be a predictor of the direct contact component of communication.56b) Whether or not the senior IS manager has:- held direct responsibility for a non-IS function,- managed a non-IS project,- been a member on a non-IS project.3. Implementation of Previous IT Plans.Since past behaviours are hypothesised to influence current levels of linkage, thispart of the investigation was designed to identify evidence of long-standing strategiccontrol behaviours, not intentions. Previous studies have indicated that factors such asreview and approval of the IT plan, monitoring of the implementation of the IT plan, andthe actual level of implementation of the IT plan are important influences on planningoutcomes. Evidence of implementation success was investigated by gathering descriptivedata about any:a) Formal procedures for monitoring and communicating the progress of the projectsin the IT plan to senior management.b) Formal procedures for changing the contents of the IT plan because of changingexternal/internal conditions.c)^Incentive system which rewards participants for successful implementation of theISP projects.In addition, the organization's success with respect to implementation of IT Plans wasassessed by both IS and line managers.'4. Communication Between IS and Senior ExecutivesIt is our belief that mutual understanding (i.e. linkage) comes about through contactbetween individuals during the everyday course of events - either face to face or writtencontact, in groups or in pairs, formal or informal. Therefore, we gathered data to assess27 The level of implementation of previous IT plans will not be objectively assessed since no theoreticalwork has been done which would suggest a method or a measure for such an assessment. However,perceptions of organizational participants concerning implementation was gathered.57the differences between organizations on this dimension. This data was collected fromindividuals in interviews and corroborated, where possible, with written documents (e.g.minutes from meetings) and with interview data from other executives in thecommunication network.Because we wished to investigate communication between different functions in thecompany (the management function and the IS function), we needed to employ aninvestigative framework which emphasises relationships between distinct groups'. Thebest known framework was that of Galbraith (1977), whose typology of lateral relationswithin organizations contained seven items. Galbraith (1977) used the typology to itemizeways to reduce information overload at the top of an organization by increasing thecommunication between groups at lower levels of management. However, the overall goalof these techniques was to "maintain integration" (p 116) which is very similar to ourpurposes of establishing "linkage" . Apart from the last level (i.e., the matrix form, whichwe found to be irrelevant since his model assumes that the two groups are hierarchicallysimilar), the typology can be used to capture many of the ways in which IS and seniormanagement interact. The typology plus examples of how it was investigated are shownin Table 111.4.Data gathered using the Galbraith typology was analyzed to see whether thecommunication is primarily IT-related (i.e. focused) or general business-related (i.e.diverse) and an assessment was made of its frequency, relative to other units in the study.Zs For the purposes of this study, the senior IT manager will be considered as a member of the IT grouprather than as a member of senior management. The proposed typology of communication will detect anyoverlap in roles.58Table 111.4Galbraith's Lateral Relations Typology with examples of Communication betweenIS and Senior ExecutivesRelation ExamplesDirect contactBetween ISmanagement andsenior management.- Informal interaction e.g. ad-hoc meetings, business lunch/dinnerengagements, social engagements.- Formal internal interaction e.g. scheduled meetings, presentations.- Formal external interaction - e.g. joint attendance by IS and Executivesat events- Written contact - memos, reports, electronic mail, etc.Liaison role - A senior V.P. named as a contact person between IS and seniormanagers.- Creation of specific IT-related roles (i.e. marriage broker, rich uncle)(Vitale et al., 1986) to foster strategic systems.Temporary TaskForces- A strategic IS planning project with senior management participation.PermanentTeams/Committees- A permanent part-time IS Steering Committee staffed with seniormanagement.Integrating Roles - A permanent full-time department staffed with people from other partsof the business to address IS issues.Managerial Linkingrole- IS managers in roles as "general managers"- Senior IS manager on the Executive Committee.5. Connections between Business and IT PlanningMuch of the literature on linkage identifies the IT planning process as the crucialtime during which linkage is forged. In order to assess the amount of connection betweenthe business and IT planning processes, we gathered data about the most recent IT andbusiness planning cycles and looked for evidence that IT planning was done inconjunction with business strategy setting.To create a "level of connection" scale, we investigated the IT planning literature.There have been numerous attempts to categorize IT planning methodologies. Theoriginal "top-down" vs "bottom-up" dichotomy (Rapanos, 1985) was expanded to include59"middle-out" (Henderson and Sifonis, 1988). Huff and Munro (1985) developed a modelof IT adoption and assimilation which suggested that IT opportunities could be surfacedthrough issue-driven, technology-driven, or opportunistic organizational processes.Recently, a number of descriptive and prescriptive typologies (Kottemann andKonsynsld, 1984; Galliers, 1987A; Jang, 1989; Henderson and Venkatraman, 1990;) havebegun to converge on five generic types of IT planning and can be used to develop atypology of planning styles based on the degree of connection between the IT andbusiness planning process. Table 111.5 contains a summary of the various typologies.With respect to their potential to create linkage, the types of IT planning depictedin table 111.4 theoretically represent a continuum, from low to high. For example, anArchitected planning system can be hypothesized to be easier to link with businessobjectives than is an Isolated planning system since the overall blueprint for futureapplications and technology is known. A Derived IT planning system ensures a higherlevel of linkage than an Architected (but unprioritized with respect to strategy) plan. Abetter situation is the Integrated model in which both business and IS planning is done atthe same time, with interplay between the systems. The fifth type of planning, labelled"Proactive", might be considered superior since the IT opportunities are guaranteed tobe considered when business strategy is formulated. Only companies which incorporatethe latest technology developments in their business objectives would find this planningmethod to be useful. For those which do, this model would ensure a higher level oflinkage than the other four since IT would receive separate advance treatment and wouldhave more potential to influence business strategy.60Table 111.5A Comparison of IT Planning TypologiesDescription of the Type of IT PlanningKottemannandKonsynski(1984)Galliers(1987)Jang(1989)HendersonandVenkatraman (1990)1ISOLATEDITP is performed inresponse to known IT orbusiness process issues, abottom-up approach.isolated isolated pre-planningtechnologyimplementation2ARCHI-TECTEDITP creates architectureswhich would support a top-down approach, but the ITplans are created fromknown IS or businessprocess issues.integrated(architected)separate strategyimplementation3DERIVEDITP occurs after businessplanning and the contents ofthe IT plan are derivedfrom the business plan.Items in the It plan arebased on businessobjectives.derived derived linked technologyleverage4INTEG-RATEDITP is done at the sametime as business planningand each plan can influencethe other.participative"ISP ispart andparcelof BP"integrated5PROAC-TIVEIT opportunities areidentified in advance ofbusiness planning. Thenintegrated IT and businessplanning is done."ISP istheimpetusfor BP"technologyexploitationThe data collection was organized to capture the inputs, the process and the outputsof the latest IT plan. This comprehensive approach allowed us to 1) describe the planningprocess, 2) to reveal the presence of factors which have been previously identified asbeing influential (e.g. availability/communication of the business plan, timing between6 1IS and business planning, involvement of senior management & business planners in ISP,and particular ISP methodologies which create linkage) and thereby to 3) classify the ITPprocess based on our typology in Table I11.5.Specific data was collected concerning:Inputs to ITP: format and type (written/verbal, formal/informal), sources of information(industry, executives, middle management, IS managers).Process of ITP: participants (individuals, committees), timing (with respect to businessplanning), methodology used, eventsOutputs of ITP (intermediate and final): format and type of outputs, communication ofthe results.F. Data Gathering and Analysis1. Data Gathering ProtocolIn this research project, the doctoral student, as "primary researcher", conductedall interviews and performed the data analysis. Her advisor, acting as "secondresearcher", reviewed the site reports, delving into the raw data as necessary to ensureconsistency across the findings.The data collection process in each organization proceeded such that the leastintrusive methods are first and the most intrusive methods last. There were 7 steps to thedata gathering process:a) Preliminary interviews were held with the key informant to assess the suitabilityof the site and its willingness to participate. Evidence of strategic reliance on ITand general maturity of the IT function was gathered.b) Written documents, including annual reports, procedure manuals, organizationcharts were gathered as background material for the interviews.c) Interviews were conducted. Audio tapes were created during interviews.d) Other written documents (strategic plans, minutes of meetings) were gathered and62analyzed to verify the interview data.e)^Any discrepancies between the interview data and the written documentation wasexamined in follow-up interviews or in additional interviews.0^After the within-site data analysis, site reports was sent to all 3 corporate units and8 of the business units and the key informants in each were asked to verify orchallenge the descriptive and interpreted data.g)^The primary researcher gave presentations at 2 sites (to 2 corporate and 4 businessunits) to summarize the findings and gain additional verification that theinterpretation had validity for the informants.2. Within-Site Data AnalysisAfter the interview and written documents were collected, the researcher had thetapes transcribed. From these data, a long (20-30 single-spaced page) site report wascreated by the first researcher which described the unit using the factors from the modelas headings. After each descriptive passage, the researcher assessed the factors usingordinal scales. A description of the ordinal scales is contained in Appendix D.At the end of the report, the researcher summarized the factors which seemed, byaccounts from participants and from interpretations by the researcher, to be influentialwith respect to linkage. Several examples follow:a) " A history of late, over-budget, implementations coupled with an inability to capturebenefits from the systems once they are implemented."b) " A low level of communication between IS and line managers due to the infrequencyof management meetings and the lack of a strong informal network tying the IS managerto the executives."c) " A high level of connection between IT and business planning - planning is done forall projects simultaneously."The draft of each site report was read by the second researcher to identify anyweak spots in the analysis and to ensure that the analysis was done to the same level of63completeness and consistency for each unit. Changes were suggested and made by theprimary researcher.The updated report was circulated (minus any direct quotes taken from theinterviews) back to the units for their verification. Half of the reports were returned backto the researcher with comments from the informant. In all but one case, the informantsapproved the analysis and the reasoning. In the one case where there was a discrepancybetween the report and the informant's understanding of events, the researcher returnedto the site and interviewed the informant, who had not been available at the time of thefirst visit.3. Across-Site Data AnalysisThe inputs for this phase of analysis were the corrected site reports for each of theunits of analysis. These site reports are not contained within this dissertation document.The first step was to take each report and separate out the analysis of linkage fromthe analysis of the factors which had influenced it. This rather drastic action was taken,rather than going directly to a holistic look at the factors which influence linkage:a) to assure ourselves that the scales for each linkage measure and each factor had beenapplied consistently. This was important in linkage, especially, since we had altered thelinkage measures during the analysis and had added a new measure.b) to be able to rank the sites on the dependent variable - linkage - without fear that thefactors data would confound the ranking.c) To assess the importance of each factor, both hypothesized and emergent.This step, which consumed at least four months of time, would not have beennecessary had we started with reliable scales. However, this was an exploratory study andwe had no illusions that the scales we created from literature studies and from experiencewould stay completely intact when confronted with real data. Furthermore, one of the64primary research questions was to identify appropriate ways to measure linkage.Four new reports were created and incorporated in this dissertation:a) corporate level linkage in the 3 corporate units (Chapter 4),b) influential factors in each of the corporate units (Chapter 5),c) business unit linkage found in each of the business units (Chapter 6),d) influential factors in each of the business units (Chapter 7)Each report contained one section for each unit of analysis, a section summarizingthe across-site findings, and a section ranking each unit on the relevant scales. Chapter5 and 7 contained revised models based on the data from the study.The last step (results are shown in Chapter VIII) was to combine the findings ata higher level of abstraction and to put them into the context of current research andpractice.G. Issues in Reliability and ValiditySeveral steps were taken in a deliberate attempt to increase the reliability andvalidity of this study. These will be discussed below.1. ReliabilityFour aspects of the research design directly affect the reliability of the datacollected and the interpretation of this data. They will be outlined below.For each organizational unit being analyzed, multiple respondents were used toprovide reliable data about the factors. Having multiple informants allowed the researcherto learn about a process or event in the first interview and to follow it up and verify itin subsequent interviews. Also, multiple perspectives (SVP, line manager, IS manager)were identified on the topics.65Data from multiple sources, documents and interviews, was collected. Asmentioned earlier, written data was used to customize the interviews and use the timemuch more efficiently. It was also used to verify other details such as the exactattendance at meetings and the frequency of meetings. For example, the IT SteeringCommittee of one unit was designated to meet monthly but in fact met only three timesover the course of an eight month period.After the site report was drafted by the primary researcher, the second researcherread and annotated it and looked for inconsistencies in the application of the model or theassessment of the factors. These reports were filled with anecdotes, quotes from writtendocuments, and interpretive notes in order that the researcher was exposed to as muchraw data as practical and could form an unbiased conclusions (Yin, 1989; Sviokla, 1986).After comments were received, the report was subsequently modified to eliminateinconsistencies.The modified report was sent to the main informant of each unit in order to checkfor correctness and completeness of the data. The informant was asked to annotate andreturn the descriptive analysis. Eleven of the informants returned their write-ups and onlyone requested significant changes based on new information. A follow-up phone interviewand meeting was held to correct this situation.A very important step in improving the reliability of the research occurred afterthe site reports were completed and verified. In order to ensure that scales wereconsistently applied across each site, the reports were taken apart and each factor wasresealed across all units. This resulted in a few adjustments in ratings, expansions to twoof the scales (i.e. vision for IT and congruence in written documents) and a higher levelof confidence that future researchers would be able to follow the reasoning and wouldgenerally support the interpretation.A more indirect influence on the reliability of the interpretation was the history of66two decades of IT work, including five years in strategic IT planning, brought to theproject by the primary researcher. The SVP in one unit remarked that, at the end of theproject, "she knew more about us than we did ourselves'.2. Internal ValidityContent validity in the overall study was addressed by 1) making a wide survey ofthe IT literature, 2) by using the strategic management literature to expand on the ITperspective, and 3) by drawing on the experience of the researcher, who has conductedseveral strategic IT planning projects.Construct validity is more difficult to demonstrate, since one of the objectives ofthe study was to develop and define the linkage construct. However, the linkage constructwas given a-priori dimensions which did not bias against companies which do not conductformal business planning. We separated the outcomes of linkage from its antecedents andused information from the respondents to reconcile our theoretically based measures withdocuments and perceptions from the field.A weak form of predictive validity was a goal of this research design. This wasa difficult objective for interpretive research, but several steps were taken towards it. AsEmory (1985) points out, causal hypotheses can be created only if covariation andtemporal relationships are shown and alternative explanations are refuted. Our designpermitted the creation of "stories" which clearly outlined the sequence of events in eachunit of analysis. Across units, we identified patterns of covariation between the factorsand the dimensions of linkage. Being able to refute alternative hypotheses was a matterof serendipity, however, since it depended on the presence of certain patterns of factorsin different units and different linkage outcomes.29 This is a paraphrase of comments made by both the IS manager and the President in one of the sites.673. External ValidityThe objectives of case research study differ significantly from survey-basedapproaches. For each unit of analysis, we tried to understand "why" and "how" linkagewas enabled and blocked. By aggregating our findings over a few selected organizationalunits, we have created a preliminary theory about the influence of certain factors. We areusing our data to "generalize to theory" (Yin, 1989), rather than to a specific population.Subsequent studies will be required to test the relevance of the theory in different settingsand to identify additional elements in the framework.However, we believe that the findings will be relatively robust over the populationsof multi-divisional organizations which believe and act as if IT is crucial to the attainmentof their goals and which are operating in a turbulent, competitive environment.H. Potential Contribution of the Research MethodologyThere are many research approaches which have the potential to shed light on thelinkage question. Our design - a small sample of theoretically homogeneous units -coupled with a comprehensive data collection protocol, enabled us to produce some newinsights. A few examples are:a) Because of its holistic approach to data gathering, the research project surfacedadditional factors of interest to researchers and practitioners.b) Because it involved participants from several roles (CEO, Executives, IS manager),the project identified and documented conflicting sets of objectives and beliefs whichinfluence the attainment of linkage. The existence of these conflicts had been suggestedin previous literature (Galliers, 1987B) but not confirmed.c) Collecting data covering a two year time period enabled us to construct causal modelsat the organizational and business unit level. Previous cross-sectional approaches did nothave this potential.68IV. FINDINGS CONCERNING LINKAGEand the MEASUREMENT OF LINKAGE at the CORPORATE LEVELTo summarize the discussion on linkage from chapter III, this project measured thesocial dimensions of linkage at two levels (the corporate and the business unit), using theexecutives' subjective and the researcher's interpretive assessments based on data from1) interviews, and 2) written strategic business and IT plans. The social dimension oflinkage was defined as "the mutual understanding between IS and business executivesregarding business and IT mission, objectives and plans" .We identified four possible ways to measure the social dimension of linkage: 1)the degree to which written IT and business plans referenced each other, 2) the degreeto which IS and business executives could articulate each others' current objectives, 3)the degree of congruence between the visions for IT from IS and business executives and4) the subjective assessments of linkage by IS and business executives.An objective in this research project was to explore the effectiveness of thesemeasures of linkage and to identify others as the study proceeded. We captured the fourmeasures throughout the project and then examined the convergence/divergence betweenthem during the data analysis phase of the study.This chapter reports the results of our assessment of corporate level linkage at thethree organizations. The next chapter reports on the factors which were hypothesized toaffect this linkage.At the corporate level, we focused on the Corporate IT objectives and assessedwhether or not they were "linked" with the organizational objectives. These corporate ITobjectives are the responsibility of the Corporate IS departments, whose role within theorganization typically includes provision of technology infrastructures (processors,networks, system software), specialized technical services for application developers, and69support services for end-users. Because the companies in our study had several product-oriented business units, we expected that Corporate IT objectives would be created inconjunction with both corporate business objectives and business unit (BU) businessobjectives. Therefore, we interviewed both corporate executives (CEO; SVP, Finance;SVP, Investments) and business unit executives (SVP, Individual Insurance; SVP, GroupInsurance, etc .)3° .Several of the linkage measures are portrayed in a diagram of "ideal" corporatelinkage in Figure IV.1. Because corporate IT objectives include goals and strategies thatdirectly impact both corporate departments and business units, "corporate linkage" hasbeen interpreted within this project to mean that the Corporate IT objectives areunderstood by senior company executives' (as shown by lines 1, 2 and 3 in thediagram), and that Corporate IS executives understand both corporate and BU objectives(as represented by lines 4 and 5). We also expected to find a high degree of crossreferencing between corporate IT plans and both corporate business plans and BUbusiness plans (as represented by lines 6 and 7). The other two measures, presence of ashared vision, and subjective assessments, are not shown in this diagram.Our data collection measures and analysis procedures were designed to test thedefinition of corporate linkage and to assess the success of the three companies inachieving it. As mentioned earlier, these companies were chosen because they haddemonstrated that IT was a critical success factor for their businesses, so the achievement' In this context, "Corporate Executives" include the CEO/President, the Chief Financial Officer and theChief Investment Officer and "Business Unit executives" are the Senior Vice Presidents of each of theproduct-oriented business units. In all three companies, the senior management team included both"corporate" and "business unit" executives.31 We also included business unit IS executives in our interviewing since early results indicated thatbusiness unit executives often delegated the interface with corporate IS to their own IS people. Insteadof automatically rating linkage as LOW in these cases, we widened our interviewing to gain as manyperspectives as possible on the subject of corporate level linkage.UnitExecutivesBusiness UnitObjectives 51-Corporate ITObjectivesCorporateISExecutives"IDEAL"CORPORATE LINKAGEcross referencing between.^ understanding ofCEOCorporateBusinessObjectives6.otherCorporateExecutives•70Figure IV.1Some aspects of "Ideal" Corporate Linkageof linkage was an important goal for each of them.In the next three sections, the findings associated with corporate level linkage forthe three companies are discussed. Then two summary sections are presented - onedescribing the across-site findings on linkage and the other discussing the issues inmeasuring linkage. Appendix C contains the scales used to rate the various aspects oflinkage.71A. Company AThe Corporate IS department within Company A had embarked on a number ofinitiatives in the past two years (1989, 1990). They had upgraded the mainframecomputer, identified a preferred mainframe technology platform (hardware and software),and coordinated a comprehensive study into rapid application development using CASEtools. They had established many corporate standards and guidelines on hardware andsoftware in an effort to streamline the technology offerings. They had reduced operatingexpenses (and chargeout rates) in the previous year and were committed to a further 5 %reduction in each of the next few years. A more complete description of Company A andits corporate IT initiatives is contained in Appendix B.1. Cross-references in Written ObjectivesIn Company A, one and five year business plans were available to peruse. Therewere several IT objectives mentioned in the Key Programs section of the StrategicBusiness Plan, 1991-1995:1) ensuring that management participates fully in systems decision makingsimilar to their involvement in other business issues,2) reducing the range and age of technologies managed,3) examination and implementation of rapid application developmenttechniques to redevelop application systems and to standardize developmentmethodologies.In the Annual Business Plan, 1991, it is stated that:Special emphasis will be placed on systems enhancements whereproductivity gains and cost reductions can be realized. We will, through ourSystems Steering Committee structure, continue to develop operationalmanagement who understand and can use information technology to achievebusiness goals. Similarly we will continue to develop information technologymanagement with requisite understanding and knowledge of businessenvironment.. to work with operational management as a team...72These plans suggest that corporate management is committed to improving theunderstanding of business managers so they can make more effective IT decisions andimproving the business knowledge of IS managers. It is also directing business units touse IT to control costs and to control the costs of IT by standardizing technology.There was no separate, written, 5 year IT plan within Company A. The 1991 Planfor Information Services is organized in such as way that all of the initiatives are listedunder Key Programs headings taken from the business plan. There are IT projects (e.g.responsible involved management, technology leadership, rapid application development)for all the key corporate objectives and the two major corporate thrusts (i.e. value formoney, quality first). Therefore, linkage as exhibited by the written corporate businessand IT plans is rated as HIGH.There was no mention made of any business unit objectives in the corporate ITplan. In the BU business plans, there was no mention of the corporate IT strategy ofusing Steering Committees to raise the knowledge and skills of business and IS managers.There was no explicit recognition of the corporate emphasis on using technology only tolower costs. The two current IT projects, rapid application development, and eliminationof old technology, were mentioned in several of the BU plans. Overall, there was LOW -MODERATE linkage exhibited between the written corporate IT and business unitobjectives.2. Mutual Understanding of ObjectivesIn Company A, the CEO and four SVPs - two from corporate divisions, and twofrom the business units were interviewed.Most of these executives could identify the corporate IT objectives. For example,the SVP of Finance remarked that the thrust was to "shorten the train" (i.e. rewrite oldsystems) and to "narrow the path" (i.e. standardize on a subset of technology for all73users). The SVP of one of the business units correctly identified the corporate ITobjectives as "reduction of the technology set, control of the mainframe costs, andevaluation of the rapid application development concept".Three corporate IS executives were interviewed. The VP of Corporate IS couldarticulate the corporate business objectives, however, neither he nor his two Directorsknew the specific objectives of the business units.On this measure of linkage, the mutual understanding of objectives, Company Aranked HIGH on most aspects: senior executives (corporate and business unit) understoodcorporate IT objectives and corporate IS executives understood corporate objectives.However, the company ranks LOW in the corporate IS understanding of business unitobjectives.3. Congruence in Vision for ITTo assess the level of congruence in the vision of IT, the CEO, the SVP ofFinance and the VP, Corporate IS were interviewed. The CEO stated:"I am not terribly excited about great innovative roles (for systems). I'mvery anxious to have low cost and I'm very anxious to have efficientservices. I think that efficiency is more important than anything else. I'mtelling my people to concentrate on those things we do with the computerwhere you can cut costs."The VP of IS was more specific in his thinking about corporate IT, suggestingthat his staff will be halved in the years ahead as users did more of the applicationsdevelopment and only a core of highly technical people remained. The role of central ISwould be to "optimize an expensive resource" and "act as a purchasing agent" forinformation technology within the company. "The real linkage and synergy will occurthrough the application system areas in the business units".The SVP of Finance was focused more on the short term objectives of IT in74reducing costs and questioned whether the decentralization of IS personnel was supportiveof that goal. He suggested that a reorganization to centralize IS people might occur in thefuture.Each executive had his own view of the future of IT in Company A reflecting hisindividual perspective'. For example, the CEO was focused on the issue of overallcorporate viability and the VP of IS on the structure and size of his staff. Therefore,congruence in their articulated statements of vision was rated as LOW.4. Subjective Assessment of LinkageTwo corporate executives rated linkage between corporate business and corporateIT objectives as HIGH since corporate IS was aggressively pursuing cost reductions andefficiencies. The VP of Corporate IS rated linkage with corporate objectives asMODERATE to HIGH. He had no specific examples of poor linkage but felt that hemight contribute more to the company if he reported to the CEO. The rating on thismeasure is HIGH, to favour the assessments of the non-IS executives.The VP of Corporate IS rated the linkage between corporate IT and the businessunits as HIGH..."In terms of what business unit management want from a central IS, I think thelinkage is there, because I listen to what they want and I'm delivering what theywant.. in the sense of low cost, good service and application development.... we'recontinually working towards delivering that and as long as we keep going in theright direction, then all we're talking about is the rate".However, the business unit executives did not agree with his assessment. Onebusiness unit SVP rated linkage with corporate IT objectives as MODERATE or lower.His perception was that a healthy tension existed between corporate IS people and32 ^finding might be an artifact of our interviewing methods. This will be discussed further in theAcross-Site Findings section.75business unit IS people and that this was not only good but effectively limited the linkagethat could ever be attained. Another SVP rated linkage with Corporate IT as LOWbecause the corporate IS group did not play a strategic role with respect to his businessunits."you have a central IS department acting more in the nature of a controldepartment.. in terms of leadership and moving people's IS strategies incoordination with strategic goals... I would really look to my own systemspeople for that.. "Another business unit executive rated linkage as LOW or MODERATE becausethe specific technology that she perceived to be critical to her unit's success was notsupported by corporate IS and it might take a two year study to get it included on thepreferred technology list. Another VP wondered what connection the rapid applicationdevelopment initiative had with his business unit. He rated linkage as UNKNOWN orLOW.The IS executives in the business units were slightly more positive in theirassessment of linkage between the BU objectives and Corporate IT objectives. Onereported "we make sure that corporate IS programs are aligned with our needs" .However, she felt that corporate IS was abdicating responsibility for vision and directionin favour of standards and procedures and rated linkage as MODERATE. For example,the business units were exploring the ideas of cooperative processing, PC-based expertsystems and LANs at the same time as corporate IS, having just upgraded the mainframe,was trying to fully recover the mainframe costs and was not interested in alternativesources of processing cycles."' In Company A, the president asked to have the decision to upgrade the mainframe justified to him. Thiswas one of the first acts of the newly created Steering Committee. The justification for more computingpower was obviously compelling but what may have been missing was an examination of alternative waysto provide that power and/or an understanding that upgrading the mainframe would preclude the businessunits from obtaining other sources of computing power in the foreseeable future. This decision was76Another BU IS executive noted that the limit of $1000 to purchase non-standardtechnology without corporate IS approval was unnecessarily limiting. "The thousanddollar limit is a real bone of contention with me.. My budget is $6 million."The average of the subjective assessments of linkage between corporate IS and thebusiness units is LOW - MODERATE. The important point to note here is that the BUexecutives indicated that corporate IS may not be able to contribute strategically to theirobjectives, that the best they could do was to support them. They complained, however,that corporate IS often "got in the way", rather than supported them. Corporate IS, onthe other hand, is marching to the "cost reduction" tune played by the CEO and SVP ofFinance. They have concentrated on a few technologies and put many standards andprocedures in place to restrict individual initiatives.5. SummaryA summary of the corporate level linkage found in Company A is contained inFigure IV.2. In Company A, we see that linkage between corporate IT and corporatebusiness objectives is very HIGH, revolving around efficiency and standardization.However, there is no shared long-term vision for IT within the company.The linkage between corporate IT and business unit objectives is LOW orMODERATE based on 1) subjective assessments by BU executives, 2) lack ofunderstanding by corporate IS of BU objectives, and 3) lack of cross-references betweenwritten corporate IT and BU plans. Corporate IS is seen to be bureaucratic and slowmoving by the business units.In summary, Corporate IT objectives are strongly based on corporate objectivesand weakly based on BU objectives.treated "tactically" rather than "strategically".BusinessUnitExecutives^1xCEOBusiness UnitObjectivesOther Findings:No shared Visionfor IT.CORPORATE LINKAGEIN COMPANY Across references between^ understanding ofnot present in this sitepartially presentCorporateBusinessObjectivesotherCorporateExecutivesis \CorporateIsExecutives (-Corporate IT---Objectives }77Figure IV.2Corporate Linkage Findings in Company AB. Company BWithin Company B, Corporate IS has developed three initiatives in the last 18months. First, the corporate IS unit has been reorganized to make it more responsive andaccountable to the business units. Second, two studies have been undertaken - one todefine a common Distributed Computing Environment and the other to define a commonApplication Development Environment. If adopted, both would significantly influence theway in which applications are developed and delivered within the business units. Third,corporate IS is committed to a 5% reduction in its operating expenses each year for the78foreseeable future. A more comprehensive discussion of Company B is contained inAppendix B.1. Cross-references in Written ObjectivesIn company B, there is no written corporate-level strategic plan other than a setof financial targets. Business units and corporate departments create separate five yearplans based on their own projections and assumptions.The Corporate five year IT plan contains a set of critical success factors for thebusiness and a set of technological objectives. This document clearly outlines the waysin which corporate IT will support what it perceives to be the highest priority businessobjective - the reduction in unit costs - by encouraging system units to limit variations intechnology, improve productivity, and recognize the 80/20 rule in systemdevelopment. This document exhibits high internal consistency (i.e. the technologicalobjectives fit with the analysis of the critical success factors and the business objectives).However, because there is no written corporate plan and the IT plan references corporateobjectives, this written dimension of linkage is assessed as MODERATE.The Corporate five year IT plan does not contain any references to the businessobjectives of the business units, thus the rating for this aspect of linkage is LOW.2. Mutual Understanding of ObjectivesWithin Company B, the CEO and the SVP of Corporate Services understood thatcorporate IS was pursuing a cost containment goal by proposing common technologyplatforms and common application development approaches. They both mentioned theneed for corporate IS to bring the business unit IS groups "in line". The CEO remarked:' Known as "Pareto principle", this rule of thumb suggests that 80% of the benefits result from 20% ofthe work effort. After this point, the marginal benefit of additional effort declines rapidly.79They (corporate IS) know their role... they are the policemen... to keepeverything running on the same track.. There are major disadvantages inallowing people to diverge, even though it might be the right answer for abusiness area and the cheapest way to go. We have to say "no, you can'tdo that" because there are other disadvantages down the road.. "In the interviews, the business unit executives could identify the corporate ISmission, namely technical leadership, research, efficient use of the CPU, and efficientsupport for user departments. They could not identify the strategies of corporate IS butrecognized that they were committed to reducing the chargeout rates. No businessexecutives mentioned any of the three recent corporate IT initiatives when asked aboutcorporate IT direction. This suggests strongly that the issue of commonality in aspects ofapplication development, which is being resisted by most of the business unit ISdepartments, has not been raised to the senior business unit management level in thecompany. It is still being negotiated between the corporate and business unit ISdepartments. In effect, the President has unofficially mandated the corporate IS group to"be the policeman" but has not signalled this intention to his business unit executives.This aspect of linkage had mixed ratings: corporate executives had a HIGHunderstanding of corporate IT objectives and business unit executives had a MODERATEunderstanding of corporate IT objectives. The rating of corporate IS understanding ofcorporate objectives is rated as UNKNOWN because there are no corporate objectives,and corporate IS had a LOW understanding of business unit objectives.3. Congruence in Vision for ITWithin Corporate IS, the vision for the future as expressed by the Corporate ISDirector was"an open systems, standards-based approach which is quite a radicaldeparture for an insurance company. We will be going with UNIX and RISC80processors for our distributed computing pla tform. "The corporate IS Director had been trying to give the business units control over researchprojects.The CEO's goal for Corporate IT is to build up the strength of the role so that itcan effectively keep the business units "in line". He favoured this strategy over thealternate one of disbanding the unit as their U.S. company has done.The VP of Corporate IS sees the role of the Corporate IS group as diminishingover time as the business units can provide their own services. In the future, they wouldbe limited to the provision of the communications network and mainframe computing.There was very little in common between these vision statements and this measureof congruence in vision is rated as LOW.4. Subjective Assessment of LinkageTwo corporate IS executives were interviewed: the VP of Corporate IS and oneof his direct reports, the IS Director.The VP of Corporate IS remarked:"I'd rate it (linkage with corporate objectives) as GOOD ENOUGH. I viewthe real decision making at the line of business level. Unless I'm missingsome real problems arising out of corporate thrusts, I don't care (aboutlinking with corporate thrusts) any more. What I'm much more concernedwith is getting ourselves more aligned properly across the divisions"."I would rate our linkage (with the business units) as MEDIUM. I wouldreally like to find better ways of getting a more effective appreciation of thebusiness (unit) objectives."Because the VP of Corporate IS does not have strong corporate objectives to tiethe unit to, he is more interested in identifying and linking with business unit objectives.In another interview, the VP of Corporate IS made the statement that "linkage happens8 1basically around projects. Either those we initiate or those that are business applicationsthat are being developed." He believed that the "technical support" provided to clientswas as important as the "IT leadership".The Director of IS rated linkage as LOW because there were no corporate businessstrategies and no way for corporate IT strategies to link to the general problems that areknown. He rated the linkage with the business units as varying, depending on the unit.Linkage with two of the units were rated LOW and two were rated MODERATE-HIGH.The CEO had another view - that the linkage was HIGH because Corporate IS"must" be highly linked to the organization. In his opinion, there are no Corporate ITobjectives ("I don't see that we have system objectives - any more than we have telephoneobjectives"), the Corporate IS group exists only to serve their clients (by providing scarceresources as a reasonable cost) and acting in a watchdog role, to protect the corporateinterests.One of the senior business unit executives rated linkage as UNKNOWN, havingdelegated all contact with corporate IS to his IS managers. Another executive suggeststhat linkage may be LOW since he hasn't seen any evidence of leadership in technologyin the past five years. He discussed the potential difference between the objectives ofCorporate IS, which is trying to charge out all of their mainframe and people costs, anda business unit which is trying to find the cheapest way to get things done. Another SVPrated linkage as MEDIUM and remarked that:"I don't think that the Corporate systems area is leading. They are more ofa watchdog as opposed to a leader. There are a lot of funny people runningaround doing a lot of funny things ... what business sense does any of thisstuff have? So I've got some credibility problems with that area."The business unit IS managers had varying assessments of linkage. One ratedlinkage as HIGH - he said that Corporate IS knows what he is trying to achieve in thebusiness unit. In general, however, the initiatives from Corporate IS towards82standardization of technology and common application development environments had notbeen well received by the business units. It appeared to one IS manager in a business unitthat corporate IS was trying to create a "one best way" approach and thus rated linkageas LOW. Another VP of Corporate IS rated linkage as only MODERATE becausecorporate IS people "have a religious commitment to new technology. They think theyknow the business and that is not true."5. SummaryA summary of the linkage findings is presented in Figure IV.3 . Linkage betweencorporate business and corporate IT objectives is difficult to rate. Two measures, crossreferences in written documents and IS understanding of corporate and objectives, areinfluenced by the lack of corporate objectives within Company B and rate UNKNOWN.Two other measures, executive understanding of IT objectives and a shared vision for ITwere rated HIGH and LOW, respectively. Company B's CEO rated linkage as HIGH andits corporate IS executives rated it as MEDIUM and LOW, respectively. These measuresdo not show convergence and we do not assign any overall rating for linkage betweencorporate business and corporate IT objectives. The issue of a lack of corporate objectivesis discussed further in chapter 5.Linkage between the business units and Corporate IS is rated as LOW -MODERATE. Most business unit executives are not involved in any way with CorporateIS, having delegated contact with corporate IS to their internal IS people. However,corporate IS does not have a detailed knowledge of the business unit objectives and isconcentrating on producing generic computing platforms and methodologies, which donot satisfy the IS managers in the business units.otherCorporateExecutivesBusinessUnitExecutivesCEOCorporateBusmessObjectives------------^----CORPORATE LINKAGEIN COMPANY Bcross references betweenunderstanding ofnot present in this sitepartially present Other Findings:No shared Visionfor IT.Business UnitObjectivesCorporateISExecutives CCorporateObjectives83Figure IV.3Corporate Linkage Findings in Company BC. Company CIn Company C, the corporate IS department is concentrating its efforts during thelast two years in helping the last two business units convert and upgrade their applicationsto the new IBM mainframe. It has been investigating new software and an upgrade to themainframe, but has not proposed or taken any strategic initiatives recently. Appendix Bcontains a more complete discussion of Company C.1. Cross-references in Written ObjectivesIn Company C, there were no written corporate plans, either long-term or annual.84There were only five year and one year budgets. The Corporate IS department createda written 1991 plan which contained several IT objectives, including administrativeefficiency through computerization, displacement of consultants, and a conversion fromone computing platform to another. None of these related to any specific businessobjectives. The rating on this aspect of linkage is LOW.The corporate IT plan did not reference any of the business unit objectives.Therefore, rating on this aspect of linkage is also LOW.2. Mutual Understanding of ObjectivesIn interviews, the senior IS manager was able to articulate several of the overallcompany objectives (growth, presence in the U.S.) but his knowledge of the business unitplans was out of date by at least one year. For example, through discussions withbusiness unit executives, it was learned that several of the business units were planninga major expense control and profitability focus for the next few years, a distinct departurefrom the "larger market share" approach taken in the latter half of the 80's. TheCorporate IS manager was only aware of the previous strategies of growth and marketshare and had not factored these new strategies into corporate IT objectives.The senior executives, both corporate and business unit, could articulate severalof the current IT objectives (better programmer productivity, higher levels of businessknowledge in IS people, changing technology platforms). Mutual understanding ofobjectives was rated as follows:a) Corporate IS understanding of Corporate objectives - UNKNOWN since therewas no corporate objectivesb) Corporate IS understanding of BU objectives - LOW.c) Corporate and BU executives' understanding of Corporate IS objectives - HIGH.853. Congruence in Vision for ITCompany C was engaged in an internal debate concerning decentralizingresponsibility for programmers and analysts. Many of the visions for IT expressed bybusiness unit executives revolved around this topic."I see all layers going.. all programmers might report to the business unitwho needed them. Programming done by people who understand thebusiness. The analogy I draw for Corporate IS is the personnel function -interview people, train them."These managers expressed no views about how IT would be deployed in the company infuture years. Other comments included: "I don't think Corporate IS has a long termvision", "I don't see the company's total IS plan yet", "IT objectives are not reallyclear". One SVP said"I don't see IS taking a leadership role in the company.... Ideas are alwaysgoing to be developed by the business unit and then acted upon by the ISgroup."The CEO, on the other hand, stated his vision as"To use IS technology in every possible area which is cost effective andhelps marketing goals. You can't treat IS as a little department. To me thereare three areas of the company - the investment area, the marketing areaand the administration area. And administration is the computer."There is LOW linkage as measured by congruence in vision.4. Subjective Assessment of LinkageThere was agreement that linkage was LOW or MODERATE at best. Somecomments included:"We are not connected to business decisions. The IT objectives were notdeveloped in response to business strategies. They represent an attempt toget IS cleaned up internally during the next two years. After that is done,we can focus on the users. Our current linkage strategy is to do whateverthe users want to do."86"At the moment, we (corporate IS and the business units) don't talk thesame language to any depth...I don't know if we know our corporatetechnology direction.""Some of the problem is that IS is separated from the company.""If you looked into the Corporate IT plan, I don't think you could see thebusiness unit plans."Most business unit executives expressed frustration that they were not able toinfluence the Corporate IT plans and did not feel that these plans would benefit theirdivisions. The average rating on this aspect of linkage is LOW.5. SummaryA summary of the linkage found in Company C is presented in Figure IV.4. InCompany C, there are no written corporate objectives, resulting in UNKNOWN linkagedimensions which relate to corporate objectives. Linkage with Business units wasgenerally LOW, although the BU executives did understand corporate IT objectives.BusinessUnitExecutivesotherCorporateExecutivesCORPORATE LINKAGEDI COMPANY Ccross references between^ understanding ofx^ not present in this siteCEOCorporateBusinessObjectives }Other Findings:No shared Visionfor IT..." .. .. .. ..... .... --X ......^... .Business UnitObjectives• ......... • ........ •CorporateISExecutives (Corporate ITObjectives87Figure W.4Corporate Linkage Findings in Company CD. Across-Site FindingsThis research project set out to assess the linkage that had been attained betweencorporate IT objectives and business objectives in three organizations. The first findingis that there are two loci of corporate-level linkage within these multi-divisionalorganizations: linkage between corporate IT objectives and corporate business objectivesand linkage between corporate IT objectives and business unit objectives. The across-site88findings will be presented in three parts: findings relating to linkage between corporateIT and corporate business objectives, findings relating to linkage between corporate ITand BU objectives, and overall conclusions.The discussion of corporate level linkage is continued in Chapter V, where the dataon factors which influence linkage are presented and conclusions are drawn concerningthe enablers and inhibiters of linkage.1. Linkage between Corporate IT and Corporate Business ObjectivesTable IV.1 displays linkage measures for the three companies. The findings oneach measure are discussed below.TABLE IV.ISummary of Linkage Ratings between Corporate IT and Corporate Business ObjectivesMeasures Company A Company B Company CCross-references in Written Documents- between corporate business and corporateIT objectivesHIGH MODERATE LOWMutual Understanding of Objectives- Corporate executives understanding ofCorporate IT objectives- Corporate IS understanding of CorporateObjectivesHIGHHIGHHIGHUNKNOWNHIGHUNKNOWNShared Vision for the role of IT LOW LOW LOWSubjective Assessment of linkage- by Corporate IS- by corporate executivesMODERATEHIGHLOW-MODHIGHLOWLOWOVERALL LINKAGE RATING- between Corporate IT and CorporateBusiness ObjectivesHIGH ??? LOW89a) Linkage as exhibited by cross-references in written documents exhibited widedifferences among the companies. Company A had clearly written corporate objectivesand the Corporate IT plan had identified the connections between IT projects and thecorporate objectives. Company C had no written corporate objectives and Corporate ITobjectives were included based on perceived needs of individual departments andtechnological innovation. In Company B, the corporate IS group had "developed"corporate objectives and tied the IT objectives to them in their planning document. Thelinkage as exhibited in these documents was a good predictor of overall linkage in thethree companies.b) Linkage measures relating to mutual understanding of objectives revealed a one-way understanding. Business executives in all three companies had a high level ofunderstanding of Corporate IT objectives, but corporate IS executives in company B andC did not understand corporate objectives, largely because these objectives were notformulated and communicated within these companies.c) No company had created a vision for the future role of IT which was shared bybusiness and IS executives.d) Subjective assessments of linkage revealed a difference between ratings given byCorporate IS executives and corporate business executives in two companies. ISexecutives were more dissatisfied with their linkage achievements than were theirsuperiors. They felt relatively powerless in influencing corporate direction and wanted tocontribute more directly and more effectively.Chapter V will explore the effect that various factors had on producing these90findings. However, one conclusion can be drawn from this measurement exercise- thepresence of corporate business objectives is important in the creation of linkage.Company A outperformed companies B and C on three indices of linkage: mutualunderstanding, written cross references and subjective assessments. One importantdifference which set Company A apart from the other two was the presence of clearlydefined corporate business objectives. Company B and C had deliberately chosen not tocreate corporate objectives - they preferred to let the business units independently developgoals and strategies which satisfied corporate financial and regulatory requirements."Why Company A decided to formulate corporate objectives and how they wentabout creating and communicating them is discussed in Chapter V. However, the veryfact that they had them made high levels of this aspect of linkage possible to achieve.2. Linkage between Corporate IT and Business Unit ObjectivesTable IV.2 summarizes the linkage findings for each company. They are discussedbelow.a) There was very little cross-referencing between written corporate IT and business unitplans. One could not identify, by reading the Corporate IT plans, what lines of businessor what strategic business units were being supported. Discussions revolved aroundtechnology, both installed and new, rather than around business unit objectives. The factthat different lines of business required different technology platforms or were at differentstages of growth in their use of IT was not discussed.' The VP of IS from Company B wrote to the researcher: "What is a Corporate Objective (Business orIT)?" The only non-financial corporate business objective at Company B was the Canadian/U.S.organizational separation.91TABLE IV.2Summary of Linkage Ratings between Corporate IT and Business Unit ObjectivesMeasures Company A Company B Company CCross-references in Written Documents- between BU and corporate IT objectivesLOW -MODERATELOW LOWMutual Understanding of Objectives- BU executives understanding of Corporate ITobjectives- Corporate IS understanding of BU objectivesHIGHLOWMODERATELOWHIGHLOWSubjective Assessment of linkage- by BU executives- by corporate IS executivesLOW -MODERATEHIGHLOW -MODERATELOW - HIGHLOWLOWOVERALL LINKAGE RATING- between corporate IT and BU objectivesLOW-MODERATELOW-MODERATE LOWb) Mutual Understanding of Objectives was markedly different between BU executivesand corporate IS executives. In companies A and C, business unit executives exhibitedhigh levels of understanding of corporate IT objectives, with company B executivesexhibiting a moderate understanding. In all companies, corporate IS executives exhibiteda low level of understanding of BU objectives. For some reasons, which will be exploredin chapter V, corporate IS executives had no access to information or were not interestedin learning about the objectives of individual business units.c) Unlike the finding of the previous section, Corporate IS executives rated linkage higherthan the BU executives did in Company A and B. They were, for the most part, satisfiedwith their support of business units. Business unit executives demurred, rating linkage92much lower. Although the business unit executives in all companies understood thecorporate IT objectives, they rated linkage as being low or moderate.Overall, none of the companies achieved a very high level of linkage between theircorporate IT and business unit objectives. Our measures provide two clues which wereexplored. The first indicator of problems was the lack of understanding of BU objectivesby IS executives. In Chapter V, the factors data is used to explain this finding.The second indicator is the difference in subjective assessments between IS and BUexecutives. Our examination of the data collected during the interview questions onsubjective assessment revealed a difference of opinion between IS and business unitexecutives about the role of the corporate IS department. In interviews, corporate ISexecutives and business unit executives, discussed two aspects of linkage: support andleadership. Support encompassed activities such as provision of technical specialists,writing generic programs, and participation in business unit projects. Leadership includedidentifying and researching new technologies of interest to the business units, anddesigning technology platforms and methodologies to support the strategies of the businessunits.The Corporate IS executives acknowledged that they provided better support thanleadership, but they favoured the support activities in assessing their level of linkage. Forexample, the VP of IS in Company A summarized: "in the sense of low cost, goodservice and application development.... we're continually working towards delivering thatand as long as we keep going in the right direction, then all we're talking about is therate" In fact, the "rate" was a dominant theme in Company A - the fact that thecompany was required to buy IBM PCs was seen by BU executives as an anachronismsince many other good quality, lower-priced, products had been available for severalyears and the company was trying to significantly cut costs in all areas. However,93Corporate IS was still studying the possibility of buying PCs from other manufacturers.This exact situation also existed in Company C - where only IBM PCs were allowed tobe purchased although each business unit was a profit centre and wished to controlexpenses. Corporate IS executives seemed to be rating themselves on criteria they setrather than understanding where the business units needed leadership.Business unit executives first mentioned the good support they were receiving fromcorporate IS but they followed this praise with criticism, either that1) corporate IT initiatives lacked relevance to their business units (e.g. the 4GLproject in Company C) or2) corporate IT initiatives were counter-productive in their business units (e.g. theRFP proposal and the application development environment in company B) or3) corporate IS was showing no leadership (e.g. no new technologies in companyC).The support dimension is not the type of linkage this project was investigating.Linkage as defined in this project revolved around "objectives", which are a combinationof mission, goals and strategies. Support is at a lower level, i.e. the operational andtactical level. Furthermore, commenting positively on the support dimension seemed tobe a tactic that the business executives used to balance their criticism of corporate ISinitiatives. Therefore, other than to note the difference in meanings about linkage, we willnot rate the quality of corporate IT "support" as an aspect of linkage.This finding of a difference in perception about roles led us to a speculation thata complex "catch-22 36 " affects the linkage between corporate IT and business unitobjectives in the companies, precluding the corporate IS department from achieving ahigh level of linkage at this location. This is discussed in Chapter VIII.' This term is taken from the novel "Catch-22" by Joseph Heller. Broadly speaking, it is a situation inwhich you one cannot achieve a goal because the act of moving towards it causes it to recede.943. Overall ConclusionsWe concluded that the simultaneous achievement of linkage between 1) corporateIT and corporate objectives, and 2) corporate IT and business unit objectives was difficultand, possibly, unattainable. Several executives opined that it could never be achievedsince the tension between independent business units and the desire for standardizationby corporate IT would reduce the attainable linkage. Furthermore, they believed that thistension was healthy and productive.For example, company A exhibited very high levels of linkage of Corporate ITobjectives with corporate business objectives. However, their linkage with the BUobjectives was only low to moderate. The very acts that they performed to bringcorporate IT into line with corporate objectives (e.g. standardizing technology, loweringdiscretionary IT spending limits) brought them into conflict with the business units. Thesame goal, cost reduction, could be pursued by both levels in different ways. CorporateIS was achieving cost reduction goals and full allocation of mainframe costs, but thebusiness units wanted to reduce costs by buying cheaper and more flexible technologyplatforms. In this case, corporate IT was strongly in line with corporate goals, but wasseen to be out of line with BU goals.The tension between the corporate culture (standardization) and business unitculture (individuality) was very noticeable, especially in Company A and B, in whichcorporate IS departments had initiated programs to reduce diversity of technology andapplication development methodologies.While there may be better measures left to define, there is a strong indication thateffecting corporate level linkage is very problematic, given its multiple dimensions(leadership vs. support, corporate vs. BU-centred), and the tensions between them. It maybe that achieving high linkage simultaneously on both dimensions is impossible since highachievement on one may preclude high achievement on the other. Perhaps the "perfect"95corporate linkage that may be attained in a multi-divisional organization should be definedas high linkage on one dimension and moderate linkage on the other.96V. FINDINGS CONCERNING FACTORS WHICH POTENTIALLYINFLUENCE LINKAGE at the CORPORATE LEVELIn chapter IV, the level of corporate-level linkage attained by the three companiesin the sample was discussed. In this chapter, data on the factors hypothesized to affectlinkage are examined.' In addition, through interpretation and by asking informants,additional factors which seemed to be influencing the attainment of linkage in theirorganizations were identified. This chapter reports on the findings of this investigationfor each site and then draws across-site conclusions.In each within-site analysis, the intention is to create causal inferences from thedata although admittedly, all criteria required for causality (Cook and Campbell, 1979,p.31)38 cannot be satisfied. Although the collection of organizational histories allows usto show temporal precedence, the small sample precludes there being much covariationin the data and the possibility to explore and discount alternative explanations in anyrigorous way. However, we did verify our reasoning with those people who were ableto confirm or deny it by sending the writeups back to the key informant in each of thesites to solicit their feedback on the analysis and interpretation of events. Their commentsare reflected in the data presented in this chapter.To summarize the discussion on factors from chapter III, this project investigatedfour major factors which were hypothesized to influence linkage: 1) shared knowledgebetween business and IS executives, 2) implementation of previous IT plans, 3)communication between business and IS executives, and 4) connections between business37 Details of the scales used to measure the levels of these factors are included in Appendix D.38 "From J.S. Mill we take three important criteria for inferring cause: (I) covariation between thepresumed cause and effect, (2) the temporal precedence of the cause, and (3) the need to ... rule outalternative interpretations for a possible cause and effect connection."97ANTECEDENTSCURRENTPRACTICESOBSERVEDLINKAGEHIGH LEVEL OFMUTUALUNDERSTANDINGISHARED KNOWLEDGBETWEEN BUSINESSAND IS EXECUTIVESCOMMUNICATIONBETWEEN BUSINESSAND IS EXECUTIVE;IMPLEMENTATIONOF PREVIOUSIT PLANSCONNECTIONSBETWEEN BUSINESSAND IT PLANNINGFigure V.1Expected Relationships Between the Factors and Linkageand IT planning.When preparing a model to guide this study, we developed propositions (ChapterII contains the full set) concerning the expected relationships between these factors andlinkage. The connections expected for corporate level linkage are depicted in Figure V.1.Findings from the corporate level data are discussed in light of these propositions in thischapter. In examining the effects of these factors, we distinguish the two loci of linkageat the corporate level: 1) the linkage between corporate business and corporate ITobjectives and 2) the linkage between business unit and corporate IT objectives.98A. Factors - Company AFive business executives and five IS executives were interviewed to gathercorporate level data. Company A communicates, to use Pyburn's (1983) typology, informal, written ways, thus a large number of written documents were made available tous. We examined strategic business and IT plans, minutes from IS steering committeesand the technology committee meetings, IT policy documents, and copies ofpresentations. Details on interviewees and documents were presented in Tables 111.2 and111.3. A summary of the company characteristics is contained in Appendix B.1. Implementation of Previous IT PlansThe VP of corporate IS has been in his current position since 1988. He has beenconsistent in his efforts to establish more controls in the IS department and to identifyways to improve productivity in the business unit IS departments. He began to rationalizecorporate IS by introducing a program called "Value for Money" which was subsequentlyimplemented corporate-wide by the CEO. He also introduced spending guidelines for thecorporate IS and the business unit IS people and created mechanisms whereby his peopletracked their time against identified projects. Other strategies pursued included: 1) gainingcontrol of and eventually repatriating the data centre, and 2) simplifying the main systemsoftware to lower the cost of supporting it. During this process of tightening the controlson corporate IS, several of his senior managers quit, perhaps because they preferred theprevious approach which allowed them considerably more autonomy.In 1989, corporate IS initiated a project investigating ways to achievestandardization in applications development which would facilitate the movement of ISanalysts and programmers, minimize support staff and minimize the cost of developingand maintaining application software. The result of this initiative was a projectinvestigating rapid application development technologies. The proposal for a pilot project99was being discussed by the Senior IS Steering Committee when data for this research wasbeing gathered.One of the goals of the corporation has been to drastically reduce discretionaryexpenses. Corporate IS has reduced its budget by more than 5% in each of the last twoyears. No new positions have been added in corporate IS for the last several years; it hasbeen downsized through attrition. This reduction has also had the effect of reducing thechargeout rates for users, thereby making computing power cheaper for the businessunits.Company A is very conservative in its approach to technology selection, relyingon IBM or IBM-clone hardware and software for all of its mainframe computing needs.It is also conservative with respect to its use of PC and mid-range technology. No mini-computers are in use in Company A and only five local area networks are installed. Mostof these were purchased under the previous IS administration; they might not have beenapproved under the current guidelines. Company A buys only IBM PCs - a projectunderway was assessing the feasibility of buying clones. No advanced technologies ( systems, imaging, RISC processors, distributed databases) are under study withincorporate IS.In summary, corporate IS has been very successful in bringing their technologyand, therefore, their costs under control. They exhibit low levels of discretion andinnovation within corporate IS and within the business units.2. Shared Knowledge of IS and Business Executivesa) Business ExecutivesThe CEO had some computer project management experience as a young managerbut he has spent the last 20 years in very senior positions within various insurancecompanies. He was rated as having high levels of insurance knowledge,low levels of IT100project managment experience and awareness of new technology, and moderate companyknowledge.The CFO has an MBA and a background which includes being a business systemsanalyst with the a consulting company. As the CFO, he helped to computerize a smallcompany and the last 15 years of his career has been in a financial capacity with a bankand then with Company A. He was rated as having moderate levels of IT projectmanagement experience, new technology awareness, and company knowledge and lowlevels of insurance knowledge.The SVP in charge of two of the business units has had no hands on managementexperience with IS projects. He was responsible for the IS function within Company Afor a year whereby he became aware of some of the problems of managing large ISprojects. His rating was high on company and insurance experience and low in IT projectmanagement experience and awareness of new technology.The SVP of another business unit has little experience with or understanding ofinformation technology. His rating were similar to the previous SVP.In summary, with the exception of the CFO, there are no senior managers atCompany A who have had any significant hands-on experience in managing large ISprojects or who exhibit a high interest in technology. None of them have sponsoredprojects which make use of leading edge technology within the insurance industry.b) IS ExecutivesThe VP of corporate IS has been with Company A almost 30 years. He first joinedCompany A in the insurance administration area in 1962 and left it to join the IS area asa junior programmer in 1967. In 1987, when application systems was decentralized intothe business units, he remained in corporate IS as the Assistant VP, reporting to the Chief101Financial Officer. He holds an insurance designation, the FLMI". Rating: HIGH inCompany A and IS Project Management knowledge, MODERATE in insuranceknowledge and new technology awareness.The Director of Technology Planning and Development, Information Services,joined Company A in 1973 and has been a technical expert and a manager of othertechnical experts for all of his career. Rating: HIGH in IT awareness, IS projectmanagement, Company A knowledge, LOW in insurance knowledge.The Director of the IS Support Centre, has spent 14 years at Company A in IS,but has eight years of experience developing applications within the Group business unit.He has completed two courses towards the FLMI designation. Rating: HIGH in CompanyA experience, Technology Awareness, IS Project management, MODERATE in insuranceknowledge.In summary, some of the central Information Services managers have completedinsurance courses and others have worked extensively on application systems for businessunits, but none of them have had any management experience outside the IS area.At Company A, there is a combination of senior executives with limited ISexperience and corporate IS managers with limited business experience. We might predictinfrequent, IT-oriented communication between these two groups of people and a lowlevel of IT innovativeness in Company A.3. Communication between business and IS executivesAlthough the previous section resulted in a prediction of low levels ofcommunication between business executives and corporate IS, in reality there are many" The FLMI, or Fellow of the Life Management Institute, is obtained by studying and writing tenexams on insurance concepts. The courses are taken by young administrative insurance personnel sothat they can learn about the insurance business and achieve promotions.102channels of communication, most of them recently initiated by the CEO in response tohis concern about the lack of value received from IT investments. Using the Galbraithtypology, there are two types of lateral relations in Company A - direct contact andpermanent teams. While direct contact is infrequently used, there are many examples ofpermanent teams: the Senior IS Steering Committee, other corporate committees, and theIS Steering Committees within the business units. There is also another permanent teamwhich indirectly affects the communication between corporate IS and business executives,the Technology Committee. Each lateral relations device is discussed below.a) Direct ContactCompany A is an organization in which hierarchical position is a very importantdeterminant of one's communication opportunities. The company culture is such that thereis very little direct contact between senior corporate IS executives and senior businessexecutives.The VP of IS interacts very infrequently with the CEO. Every week or two, hebooks a meeting time to talk with his boss, the CFO, about particular issues. Apart fromthe committees to which they both belong, they do not see each other outside thesemeetings. The VP has no meetings with the business unit SVPs other than his committeeresponsibilities. The IS Directors have little or no direct contact with any companyexecutives.The senior management team (CEO and SVPs) meets twice a month for four hoursbut the VP of IS does not attend and there are no minutes published. His boss, the CFO,holds a management meeting with his direct reports every two weeks to pass along theimportant items from this senior management meeting.103b) Permanent TeamsSenior IS Steering CommitteeIn May of 1990, the CEO mandated the creation of a Senior IS SteeringCommittee which included the CEO, the SVPs, and the VP of corporate IS. One of thestated objectives of the committee was to "achieve mutual understanding between businessand IS". The committee mandate was to: 1) review and approve IS strategy, plans andbudget of the business units, 2) review and approve of IS policies and principles, and 3)review and approve technology platforms.An analysis of the minutes from the Senior IS Steering Committee producedevidence that the senior management of Company A was becoming actively involved incorporate IT issues because they: 1) requested justification for the CPU upgrade, 2)requested a policy on PC vs mainframe usage, and 3) discussed the extent of dataredundancy in Company A. Evidence that cross business unit communication was beingfostered by the committee includes: 1) the discussion about a shared system to administerpremiums, and 2) the discussion of a joint evaluation of periodic payment systems.While all respondents felt that the committee was still in its infancy and was notcapable of making IT decisions, the discussions being held there reflected a willingnessto tackle issues over and above the current development projects.The Senior IS Steering Committee has had some effect, even though it was newlyformed. The CEO remarked "Even the first few meetings we have had in the Senior ISSteering Committee, it's just astonishing what people realized right away. For example,the SVP of BU X hadn't realized that his lovely US pension system was really screwingup corporate's goal of putting stuff together. When I opened his mind was when I gavehim Canada pensions and he suddenly discovered that, try as he might, the two systemscouldn't talk to each other. They were incompatible. Once we sat around the table andstarted talking about it... it's a start." We see in this comment how the CEO is104deliberately using the Senior IS Steering Committee to link corporate objectives (loweroverall technology cost) to the objectives of the business units.Other Corporate CommitteesThe VP of corporate IS sits on the Quality Steering Committee and the HumanResources Policy Committee. These are company-level groups which have been initiatedby the CEO to support corporate programs. Both are chaired by the CEO. The ChiefFinancial Officer, to whom the VP of IS reports, also sits on these committees. They givethe VP of IS an excellent opportunity to fully understand corporate-wide programs beingpursued within Company A.IS Steering Committees within the Business UnitsIn each of the four business units, the CEO has mandated that the seniormanagement will meet together with their IS manager as a IS Steering Committee. He hasalso appointed a Director from corporate IS as a member on each of these committees.For example, the VP of corporate IS is the "corporate IS member" on the Individual ISSteering Committee. This committee structure is designed to increase internal business'unit linkage and to keep corporate IS informed of the objectives of the business units.These committees have had mixed success to date; meetings are often not heldwhen the company is under stress and Company A has experienced a lot of stress lately.Also, there were regular management meetings already in place in the business units40^The CEO's objective for the steering committees: "so that every issue in systems is discussedin a group at the senior level. So that they don't get this situation where the marketing guy and theproduct actuary guy work out a product and go and sell it to the boss and then somebody says, by theway we'll need a system. And then the system guy gets called in and by then its too late for him toadd any influence whatsoever because this thing has been set in concrete. I don't want anybody tohand the systems people a bunch of specs and go away. I want a process where they sit and talkabout it - talk about costs and manhours and all those other things."105before the steering committees were mandated and they often displace the IS SteeringCommittee meeting. During interviews with business unit executives, many complaintswere raised about the redundancy of these committees and also the thought was expressedthat committees which only dealt with IT issues actually reduced the linkage that thebusiness units have created with cross-functional management meetings. What was seenby the CEO as a mechanism to increase linkage is seen by the business units to becounter-productive'.Thus far, these steering committees have not been incorporated into the regularcommunication patterns within the business units.Technology Steering CommitteeOne of the older traditions within Company A is a permanent team which includescorporate IS managers and business unit IS managers. Some of corporate IS'scommunication with senior business unit executives is facilitated through this group. Forexample, in response to a question of how he managed to persuade the one SVP to giveup his highly successful, but nonstandard, application software and move to the preferredtechnology platform, the VP of IS remarked: " is really through the TechnologySteering group.. we present background position papers and discuss them at our quarterlyplanning session.. and really it is more through peer pressure of that group in terms ofcoming to an agreement as to what is the best overall strategy for the company. In termsof initiating that migration I haven't done anything specific. All we have done is identifyour standard production platform for future development."The VP of IS has strengthened the mandate of the Technology Committee (it isnow the sponsor of all projects and ratifies the corporate IT plan) while weakening the' Another explanation for the resistance is a power struggle between corporate control and BUcontrol. This is discussed in the across-site findings.106powers of individual IS managers within his own organization and the business units. Theonly way to get anything done or any budget approved is to go through this committee.By discussing all IT actions at this one forum, the committee keeps the BU IT activitiesin line with the corporate IT direction; it also keeps the corporate IS people in line withthe realities of the BU environment. The VP of IS wields the most power within thiscommittee since he is the most senior person in rank.c) SummaryIn the last few years, the number of communication channels between seniormanagers and corporate IS have dramatically increased as the CEO implemented anumber of permanent committees. These committees embody the "linking pin" strategy(Likert, 1967) of multiple, overlapping, group-interactions. For example, the VP of ISsits on several cross-organizational, cross-functional committees which link him upwardsto corporate management and laterally to business unit executives and business unit ISexecutives.The VP of IS's communication with his direct superiors, the CEO and CFO, israted as moderately frequent and diverse. With the SVPs of the business units, it is ratedas infrequent and focused (on IT issues). His strongest communication channel seems tobe with his peers in the business units via the Technology Steering Committee. Thiscommunication is rated as frequent and focused on IT.4. Connections between business and IT planninga) Between corporate business and corporate ITWhen the CEO was being recruited, he was given a copy of Company A'sbusiness plan. He decided that "It would never work because ... the whole plan waswritten by one person and there was no sign of tension, everything tied together. It ought107to look like a candleholder and this one looked like racehorse... much too good to bepractical. Business planning is a messy process. But if you don't get at least 100managers involved, you haven't got a good business plan".He revised the planning process and it is now a comprehensive top-down, bottom-up exercise. "In January, I write down what I think the objectives should be for thecoming year and what our major programs need to be to achieve those. Then I writedown a laundry list of all the issues I think we may have to address if we are going to dothat and them I circulate that to all my vice presidents (there are about 30 of them) andask them to tell me whether they agree, disagree, and want to add or delete things fromthe list. These are corporate issues such as profitability, persistence, human resources.On a single sheet of paper, they are asked to identify what they would delete, add andwhere would they disagree. Then I sit with the SVPs and we look at every single idea andtry to understand it and debate it. That's the first bottom-up process. When we're throughwith that process, I rewrite the overview. This one goes to all unit heads and their seniorpeople and they sit down to write unit overviews for their business unit. That has to tieinto the major programs of the company that we've already agreed upon. That createsmore changes to the corporate programs. Then after the unit plans have been reviewedand adjusted, all second line managers are given the corporate overview and businessunit overviews and asked to create their program plans and estimate resources. Then wemassage that data.. The whole process takes all year, its a continuous process. We haveour first strategic plan in July."After taking over his position in 1988, the VP of IS prepared a strategic plan forhis department. This plan was the source of strategies such as repatriation of the datacentre and narrowing the technology path. From 1989 to 1991, they have beenimplementing the previous strategies and no long range strategic plan was prepared. Hesays "we haven't gone back to rethink. There are a number of components and they show108up each year in our major programs, and basically we build on them.The VP of IS does his annual planning using Company A's top-down, bottom upbusiness planning process. He gets overviews from the CFO and from the CEO. He thenworks out his unit overview with his managers and sends it upwards and discusses it withthe Technology Committee.Another connection between corporate IT and business planning processes occursthree times per year at the variance review meetings chaired by the CEO. At thesevariance reviews, the CEO and the SVPs ask the VP of IS to explain various initiativesthat are happening inside his group and throughout the company and "you're on the hotseat and the other people are sitting there throwing in their questions.. its just from thatinteraction that you sort of assess what their expectations are. And my interpretation ofthe expectation in these areas is that I'm supposed to be controlling and managing costs."In summary, the corporate IT planning process is strongly connected with thecorporate business planning process through exchange of documents, in timing, andthrough variance reviews. Using the typology of planning connections developed inchapter III (and summarized in Appendix D), the planning processes are rated as beingintegrated.b) Between BU business and corporate IT planningThe connection between corporate IT and business unit planning is weaker than theone between corporate IT and corporate business planning. The VP remarks "Normallywe don't get access to business plans." , meaning that he does not see the contents of theplans created by the business units. He does not participate in any way in the creation ofthese plans. However, the heads of the business units, as participants on the Senior ISSteering Committee, do have access to corporate IT plans and these are discussed duringthe meetings.109There is a one-way connection between the corporate IS and business unit planningprocesses because the business unit executives (on the Senior IS Steering Committee andthe Technology Council) review the corporate IS plan but corporate IS does not receiveor review their plans. The mechanism which might allow corporate IS to view businessunit plans, the business unit IS Steering Committees, are not yet effective. Becausecorporate IS presents its plans both to the Senior IS Steering Committee and to theTechnology Committee, the connection between the planning processes is rated as beingnegotiated.B. Summary and Analysis - Company AIn Table V.1, the findings are summarized for Company A on factors, discussedin this chapter, and linkage, discussed in Chapter IV. In the analysis following,explanations of the linkage results are presented. First, the relationship betweenAntecedents (shared knowledge and previous implementation of IT plans) and CurrentPractices (Communication and Connections in Planning) are discussed. Then therelationship between Current Practices and Linkage are discussed. Alternativeexplanations are also examined and the question of why there is no IT vision is explored.Table V.1A Summary of the Factors and Linkage Ratings for Company AFACTORS/LINKAGELinkage between Corporate Business andCorporate IT ObjectivesLinkage Between Business Unit and Corporate IT ObjectivesHistory ofImplementationof IT PlansVery good progress was made in reducing ITcosts, which is a key corporate goal.Previous implementations are judged to besuccessful by Corporate executives.Corporate IT initiatives have resulted in a low level of innovationand autonomy for the BU IS people.Corporate IT rates were reduced, which supports BU objectivesto cut costs.BU executives view corporate IS as having had mixed success intheir initiatives over the last two years.SharedKnowledgebetween IS andbusinessexecutivesCorporate IS executives have no linemanagement experience.Corporate executives have some IT managementexperience.No executive has been in a company withexemplary use of IT.Low - Moderate level of shared knowledge.Some corporate IS executives have developed applications forbusiness units but none has had line management experiencewithin a business unit.BU executives have very limited IT experience.Low level of shared knowledge..OTable V.1A Summary of the Factors and Linkage Ratings for Company AFACTORS/LINKAGELinkage between Corporate Business andCorporate IT ObjectivesLinkage Between Business Unit and Corporate IT ObjectivesCommunicationbetween IS andbusinessExecutivesDirect contact between Corporate IS andcorporate executives is infrequent.The CEO, the SVP, Finance, and the VP, IS siton two corporate committees.(permanent team)The Senior IS Steering Committee has met for10 months. It contains CEO, all SVPs and VP,IS. (permanent team)Overall communication is moderately frequentand diverse.Direct contact between corporate IS and business unit executivesis infrequent.The SVPs of the business units sit on the Senior IS SteeringCommittee. (permanent team)Corporate IS execs sit on the BU IS Steering Committees, butthey have not met regularly yet. (permanent team)The BU IS and corporate IS execs sit together on TechnologyCommittee. (permanent team)Overall communication is moderately frequent and focused.Connectionsbetween IT andbusinessplanningTop-down, bottom-up process connects corporateIT and corporate business plans.Planning processes are rated as integrated.BU execs review the corporate IT objectives through the SeniorIS Steering Committee.BU IS execs review the corporate IT objectives through theTechnology Committee.Corporate IS does not participate in planning with or see BUplans.Planning processes are rated as negotiated.Table V.1A Summary of the Factors and Linkage Ratings for Company AFACTORS/LINKAGELinkage between Corporate Business andCorporate IT ObjectivesLinkage Between Business Unit and Corporate IT ObjectivesOther Factors The CEO is committed to improvingcommunication and decision making about IT inthe company.The company produces corporate objectiveswhich are widely communicated.LINKAGE- in writtendocuments- understandingof objectives(EXECS/IS)- Shared Vision- SubjectiveAssessment(EXECS/IS)OVERALLLINKAGEHIGHHIGH/HIGHLOWHIGH/MODERATEHIGHLOW - MODERATEHIGH/LOWnot measuredLOW/HIGHLOW-MODERATEWOO11131. Linkage between corporate business and corporate IT objectivesa) Stage One: The Relationship between Antecedents and Current PracticesIn company A, IT implementation success began in 1988 when the VP of IS,working with the SVP of Finance, began to tighten controls over IS decisions. Theirprograms were strengthened and supported by the new CEO when he joined the companyin 1989. Although there was no more than a moderate level of shared knowledge amongthese three executives, there was consensus on an important point: that centralized controlwas the way to restore the company to financial health. The CEO and the VP of ISworked together to create the IT committee structure, initiated in 1990, whichsignificantly increased the communication between IS and business executives. Thus, thecombination of IT implementation success and shared values contributed to improvedcommunication.Although shared values and IT success seemed influential, the most importantreason for the high levels of communication and connection in planning was the qualitiesthe CEO brought to his position which caused him to promote linkage specifically withinthe IT function. The CEO brought to Company A a long history of personal frustrationwith IT based on experiences in several organizations: "There's one part of my careerthat's been a total failure and that is systems. I've been trying for years and haven'tsucceeded". This statement led us to believe that this CEO was unique in the intensity ofhis approach to establishing IT linkage.'Under his direction, Company A adopted a disciplined top-down, bottom upplanning process which tied all corporate departmental plans (including the IT plan) tothe corporate objectives. This action resulted in the planning processes being rated as42 The CEO deliberately set out to increase mutual understanding of IT issues between IS and businessexecutives. His reason for establishing the steering committees was so that "every issue in systems isdiscussed in a group at the senior level".114"integrated", which is the highest level of connection we had expected to find ininsurance companies.b) Stage Two: The Relationship between Current Practices and Observed LinkageIn Company A, ratification of the corporate IT objectives, as contained in the ITPlan, is a specific responsibility of the Senior IS Steering Committee. Every SVP in thecompany is required to attend the meetings of this committee. During seven of the tenmeetings for which minutes were made available to us, IT Objectives and Corporate ISservices were discussed. The communication in these meetings undoubtedly led to thehigh levels of understanding of corporate IT objectives which were measured in thisproject, especially since outside these meetings, there was only a low level of direct (i.e.person to person) contact between IS and corporate executives.Corporate IS executives exhibited a high level of understanding of corporateobjectives. We attribute this finding to the fact that 1) Company A created corporate levelobjectives (in contrast to Company B and C, who did not) and 2) the planning processrequired all managers to plan their own objectives in support of the stated corporateobjectives. This second factor, plus the culture in Company A of producing writtencorporate level plans, explains the high level of linkage exhibited in the written plans.Therefore, communication about IT objectives in steering committee meetings andconnections in planning appeared to have a direct, positive influence on the linkagefindings.c) An Alternative ExplanationAnother possible explanation of the high levels of linkage exhibited in CompanyA is the existence of a macro-level influence, survival of the company. Company A isunder significant financial stress and it could be argued that this alone accounts for the115changes and for the outcomes, since many companies under stress establish controls andcentralize decision making. There is some support for this argument, since other actionswere taken which do not specifically relate to IT. For example, the controllers weremoved from under the business unit SVPs to report to the corporate SVP of Finance.Managerial discretion in approving purchases was reduced throughout the entirecompany.There are many ways that a CEO can consolidate control and rescue a companyin distress. The fact that the CEO in Company A chose to concentrate a significantamount of his and his senior executives time on IT seems unusual and we trace it backto his previous unsatisfactory experiences with IT.' This made him unique in hisapproach and we conclude that the CEO is the major factor in the high levels of linkageattained by company A.Figure V.2 shows the influences among factors and between factors and linkage."d) The Absence of a Shared VisionOne measure of linkage was rated LOW in Company A - shared vision for IT.There are several plausible explanations for the lack of an IT vision:1. the company is under stress and is intent on controlling costs. The CEO hasstated that he wishes only to focus on cost-cutting uses of technology. The Corporate ISgroup is taking a "managerial" rather than a "visionary " approach to its mandatesince the VP of IS perceives (correctly, in our view) that this is what the CEO requires43 ^IT represents a significant percentage of the discretionary expenses in company A,expenditures were not higher than in the other companies we studied and were not considered by theexecutives to be abnormal." The relationships between the factors and linkage are shown by arrows. The size of the arrowdenotes the importance of the influence. Factors which emerged from the data are shown insidedotted-line boxes.LINKAGECommunicationbetweenBusiness andIS ExecutivesC.onnectionsbetweenBusiness andIT PlanningProcessesShared Know-ledge betweenBusiness and ISExecutivesImplementationof PreviousIT Plans..,\\ /Shared ValuesbetweenCorporate and ISExecutivesCharacteristicsof the CEO116Figure V.2Corporate IT to Corporate Business Linkage:Causal Relationships in Company Aof him. Therefore, the "climate" does not support visionary thinking or deviations fromthe strong top-down control.2. visions are not appropriate at the corporate level of a multi-divisionalorganization. The visions need only to be forged at the BU level where IT applicationsprovide realizable benefits.3. no corporate IS executive or corporate business executive has ever worked ina company which has been an exemplary user of IT. Therefore, they are unableindividually to create an IT vision.1174. The Senior IS Steering Committee, which could be the mechanism within whicha vision is created, is too new. There has not been enough shared knowledge createdwithin this group to support the creation of a vision. Therefore, they are unablecollectively to create an IT vision.All of these explanations are plausible in Company A. In the next sections oncompany B and C, we create similar conjectures and then synthesize them under theacross-site findings.2. Linkage between corporate IT and BU business objectivesa) Stage One - The Relationship between Antecedents and Current PracticesBased on the low level of shared knowledge, we hypothesized that there would bea low level of communication between corporate IS and business unit executives. Levelsof communication were higher than expected, but only because the CEO had decreed theexistence of the Senior IS Steering Committee. Outside of these meetings, there was nocommunication.Implementation of IT initiatives has had a mixed response in the business units.On one hand, the reductions in computer chargeout rates have allowed BU IS people tokeep up their level of activity for the current year while still cutting their IT budget asrequired by corporate guidelines. However the policies which allowed corporate IS toreduce chargeout rates are being resisted since they limit BU discretion. It is unclear ifthese mixed reactions have directly affected any current practices.b) Stage Two - The Relationship between Current Practices and LinkageBU executives exhibit a high level of understanding of corporate IT objectivesbecause they sit on the Senior IS Steering Committee. As mentioned previously, theyhave discussed corporate IT plans and policies at several meetings. Thus, even this118moderate amount of communication focused on corporate IT objectives led to linkage. ISSteering Committee meetings do not include a discussion of BU business or IT objectives;only the current projects within the business units are on the agenda. Corporate ISexecutives have no exposure to the planning process in the business units, nor do theyreceive the BU plans. Not unexpectedly, they exhibited no understanding of business unitobjectives in their written plans or in interviews. Therefore, communication andconnection in planning contributed to the attainment of linkage.°A factor which emerged from the data was a disagreement concerning ITobjectives between the corporate and the BU IS people. The BU IS managers chafedunder the restrictive policies regarding acquisition of technology, believing that they coulduse alternative technologies to reduce their IT costs. The CEO recognized the value indiscussing objectives and mandated IS Steering Committees in each BU which wouldcontain a member from corporate IS. These committees were not meeting regularly. TheTechnology Committee provided a good communication channel between BU IS andcorporate IS executives but it could not influence the overall corporate IT objectives,which were to standardize technology, reduce discretion, and cut costs. These goals were,of course, contrary to the wishes of individual business units, each of whom sawthemselves as needing unique solutions. This disagreement has affected the subjectivelinkage ratings of BU executives, which ranged from LOW to MODERATE.Can these conflicting objectives ever be reconciled? The CEO is hoping thateffective steering committees in the business units will help and he is still committed toenforcing their existence. They may indeed raise the awareness of corporate IS peopleabout the unique business unit situations which may be addressed by non-standard"This CEO has not been successful at achieving one of the stated objectives of the Senior IS SteeringCommittee, which was to "review and approve the IS strategy, plans and budgets of the businessunits". If they had achieved this objective, the understanding of the corporate IS people about BUobjectives would have been much higher.119technology. They also may raise awareness of the costs of non-standard technologyamong the BU managers. To some extent, however, the fault and, therefore, the solutionmay lie with the Senior IS Steering Committee. For example, this committee recentlyratified a decision to upgrade the mainframe. This decision was discussed by seniorexecutives because of the amount of capital it required but was not treated as a strategicdecision, meaning that a wide search for alternatives and many interested parties were notinvolved in the decision making process. For example, IS executives from the businessunits were not involved and alternative ways to provide additional computing power werenot explored. The BU executives on the Steering Committee did not realize that thecomputer upgrade would make it difficult to achieve some of their IT objectives sincethey would be prohibited from bringing in any new technology until the mainframeneeded another upgrade. Therefore, the lack of discussion of BU IT objectives lost themthe opportunity to move to other (possibly cheaper) solutions such as localized platforms,4GL languages, and cooperative processing.Figure V.3 portrays the connections between factors and linkage.'c) Why is there disagreement between corporate and BU IT objectives?We have identified four plausible explanations for this finding. The first is theexistence of a systematic factor which was hinted at by an SVP in Company A - thatdisagreements between corporate and business unit groups should be expected and viewedpositively. In his opinion, no matter how much communication, connection in planning,or goodwill exists between corporate and business units, there will always be tension andthere will never be high levels of linkage. In other words, the disagreement describedabove is predictable, useful, enduring, and not unique to Company A.46 See previous footnote for the legend.CommunicationbetweenBusiness andIS ExecutivesConnectionsbetweenBusiness andIT PlanningProcessesShared Know-ledge betweenBusiness and ISExecutivesImplementationof PreviousIT Plans\ 7,----'LINKAGE120Figure V.3Corporate IT to Business Unit Linkage:Causal Relationships in Company ADisagreement Iabout ITObjectivesbetween BU and;Corporate ITThe second explanation is an interaction effect between the task of creatingmutually acceptable IT objectives and the current people who are trying to do this. It isa complex task to create corporate IT objectives which are linked to the objectives ofmultiple business units, while adhering to a policy of fiscal restraint. Company A iscontinuing to provide standard mainframe technology - a technology that has worked wellfor them in the past but may not be appropriate in the future. The problem may lie in thecomposition of Corporate IS in which all of the senior managers have been with thecompany for more than 20 years. Their capacity to embrace new technology might be121very limited. So the combination of a complex task and a hidebound department mayexplain why no shared IT objectives have been created.The third explanation is that there is a disagreement about business strategies anda resulting power struggle between corporate executives (including the CEO, the SVP ofFinance and the VP, IS) and the business unit executives. In this disagreement, thecorporate group wishes to improve the financial health of the company by reducingexpenses and concentrating on efficiency. The business units wish to improve thefinancial health of their units by pursuing revenue and market share goals, as well as bycutting costs. The CEO has stated that IT applications which focus on reducing cost arethe only important ones; the business units disagree and are interested in uses of IT whichsupport other goals. In the power struggle, the business units are resisting the incursionof corporate IS into their management processes by not holding IS Steering Committeemeetings. The result is a low level of communication, few connections in planning, andtherefore, low linkage.The fourth explanation is a tradeoff between corporate level linkage and BU levellinkage. As stated by business unit executives, BU linkage is enhanced by having theirmanagement meetings include their IS people, discussing items of business from allfunctional areas. The CEO is asking them to separate out their IT issues and discuss themwithin an IS Steering Committee. This request results either in duplicate discussions atboth forums, which was reported by some business units, or IT issues not being discussedin regular management meetings. Both outcomes seem counterproductive to BUexecutives and they resist by not holding the IS Steering Committee meetings very often.Not surprisingly, they chose to increase BU linkage at the expense of corporate122linkage.°We do not have enough data to choose between these explanations. In fact, theymay all contribute to the findings. They are carried forward into the Across Site Findingsfor further discussion.C. Factors - Company BFive business executives, four corporate IS executives, and two SVPs of businessunits were interviewed within Company B. In addition, the five year strategic plan,minutes from the Technology Committee meetings, corporate IS reports, and a letter fromthe corporate IS Director to the CEO were reviewed. Details on interviewees anddocuments were presented in tables 111.2 and 111.3. A summary of the companycharacteristics is contained in Appendix B.1. Implementation of Previous IT PlansIn the last two years there have been four corporate IT initiatives, all spearheadedby the new Director of IS who previously was the IS executive in the largest businessunit. Two of these initiatives were organizational changes and two affected the technologyplatforms.a) The corporate technology group rearranged themselves internally to simulate a matrixstructure between themselves and the business units in order to move away from a puretechnology orientation and towards a more client-focused orientation. The group wasrenamed into a "service" group and they began direct billing for their services to businessunits. The matrix management portion of the change fits into Galbraith's communicationtypology as an example of a "liaison role" . At the time of data collection, these changes"7 Viewed a bit more widely, it is not a zero sum game. If the Technology Committee discussed BUobjectives, this could be the forum to enhance corporate level linkage, leaving the BU managementmeetings to enhance BU linkage.123had been in place for six months. Clients in the business units had mixed reviews of theresults but were in favour of the overall change.b) The mandate of the Technology Committee (which is comprised of the senior ISexecutive from each business unit and several from corporate IS) was changed frominformation sharing to direction setting. The frequency of the meetings increased and thiscommittee began to sponsor cross-business unit projects. To use Galbraith's typology, thischange is the empowering of a "permanent team". No IS managers in the business unitsremarked on this change.c) A business unit, based on the advice of a consultant, proposed that IBM be awardeda large contract. Corporate IS intervened and recommended that another vendor beselected in order to move away from an all-IBM environment. The CEO had to resolvethe issue and decided to continue the relationship with IBM. This created bad feelings inthe business unit towards corporate IS.d) The Technology Committee sponsored two projects to define platforms andmethodologies which would be shared by all business units. The proposals from theseproject groups were stalled in the Technology Committee and had not become companypolicy at the time of data collection. Using Galbraith's typology, these project groupswere examples of "temporary task forces" but did not completely conform to his modelsince they were staffed entirely by corporate IS people who solicited input from IS peoplein the business units. These projects have created tensions between corporate IS and thebusiness unit IS people.All of these changes were designed to improve linkage between corporate IS andthe business units and to increase the level of leadership that corporate IS provided withinthe company. Before that time, corporate IS was regarded as a technology-drivendepartment with little knowledge of or appreciation for the needs of its clients. Thetechnological changes have not been too successful, since they promoted a "one best way"approach which was to be followed by all business units. Therefore, the success ofcorporate IS in the last two years is seen as being "mixed" .2. Shared Knowledge between IS and Business Executives124The CEO has not had direct experience in managing a large IT project, nor is hea reader of technology literature. Rating: high Company B and insurance experience, lowIS project management experience and IT awareness.With the exception of one business unit SVP, who is known as a PC user andadvocate, none of the executive group has shown much interest in IT. Most of them havebeen senior managers for so long that they have had no direct experience in managing alarge IS project. Their ratings are the same as those for the CEO.All of the executives have a long history with Company B, often in severaldivisions of the company. With the exception of one SVP, who has been seconded tofederal jobs and has taken leaves of absence, most of the management have no significantsenior experience beyond Company B.At the time of data collection, the SVP of Corporate Services had just retired. Hehad been in charge of the corporate IS department for the last 15 years and formerly heldseveral line management roles in the business units. Rating: high Company B, insurance,and IS Project management experience, low-moderate IT awareness.The Vice President of IS has spent most of his 30 year career in the IS field. Hetook insurance courses and received the FLMI designation in the 1960's but has no linemanagement experience. Rating: high Company B and IS Project management experience,moderate IT technology awareness, low insurance experience.The Director of IS has worked in both the Individual and Group business units. Hebecame involved with technology as a user manager in 1977 and took responsibility forIT as well as claims in his business unit in 1983. He joined corporate IS in 1990 toprovide leadership in technology direction. Rating: high on all scales - Company B,insurance, and IS project management experience and IT awareness.In summary, there is almost no recent IT experience among the senior businessexecutives in Company B. The Corporate IS group has always had a senior manager with125a significant amount of line experience and the hiring of the new IS Director continuesthis tradition.3. Communication between business and IS Executivesa) Direct ContactThe VP of IS meets very infrequently with the executives in Company B other thanhis direct superior, the SVP of Corporate Services.The Director of IS meets irregularly with the SVPs of two of the business unitswhen they call him about specific technology initiatives within their division. He isinvited to participate in the management meetings within another BU. He does not meetoften with the SVP of the fourth BU. He has had several meetings with the president onspecific technology issues.b) Permanent TeamsThe CEO and SVPs meet regularly as the Executive Committee. The SVP ofCorporate Services sits on this committee and represents Corporate IS. No minutes arepublished from this meeting.With the exception of the executive committee, there are no cross-functionalcommittees at the executive level within company B, such as an IS Steering Committee.Therefore, there is no forum within which the VP of IS can interact with the business unitexecutives.The Technology Committee has been in existence since 1974. It brings togetherthe head of IS from each business unit and the VP and his Directors within Corporate IS.In the past, they met every quarter and have recently changed the frequency to bi-monthly. This forum existed previously to exchange information and is now taking amore proactive role in setting direction for IT within Company B.126In summary, direct communication between Corporate IS and top management(both corporate and BU executives) was very infrequent before 1990. Since the new ISDirector joined corporate IS, the SVPs and the President have shown that they value hisadvice concerning their projects and communication has increased somewhat. There is nouse of liaison roles or permanent teams to foster communication between businessexecutives and corporate IS. Communication between Corporate IS and businessexecutives is infrequent and focused on IT issues.4. Connections between Business and IT PlanningThe corporate five year planning process usually occurs in three rounds betweenApril and August of each year. The package prepared for the process by CorporatePlanning includes economic forecasts, valuation and taxation policies and a summary ofspecial issues. All of the material is financial in nature. Business units formulate theirbusiness directions, translate them into budgets and forecasts and feed the numbers intoCorporate Planning. Only financial information is collected corporately. There is nocompilation of a "strategic plan" at the corporate level: strategies are written up by thebusiness units in their preferred format and used internally by them.The corporate one year planning process, which occurs after the five year process,results in detailed plans and budgets for the following year, using the first year of the fiveyear plan as a starting point. These are presented to the Executive Committee by eachbusiness unit.The Corporate IT planning processes involves Corporate IS and its clients. Aftera talk from the CEO, corporate IS internally discuss the issues they see as important tothe company and the VP of IS drafts a document for discussion. After more talks (up totwo or three days in some years), they finalize this draft document and schedule meetingswith each of their major clients for discussions of their five year plans. The top four127people in the Corporate IS group will meet with the SVP, the IS Director and otherexecutives of each of the business units and corporate departments and review theCorporate IS direction document. In a given year, they may have up to 10 of thesemeetings to ratify their direction.After this process, they prepare tactical one-year plans and strategies identifyingthe services they are providing for each client and what technology projects they areundertaking. They ratify this document at a single long meeting with all their clients.The attendees at this meeting are usually at the Director and manager level in the businessunits. Therefore, many executives and managers within the business units have exposureto the corporate IT plan during one or more meetings each year.The Technology Committee is not involved as a group in ratifying the CorporateIT one or five year plans.Business unit planning is occurring at the same time as the Corporate IT planning,and there is no crossover of strategic plans from business units to Corporate IS. IfCorporate IS does receive a copy of these plans, it is usually late in the process of thefive year planning process. Therefore, the plans that Corporate IT formulates are notbased on written business unit objectives, they are based on information from their IS andexecutive contacts in the business units.In summary, the information being fed into the corporate IT planning process fromthe CEO is entirely financial. This input does not give any direction to corporate IS onthe technology required or the vision for the company.The Corporate IS group formulates plans internally and then presents them tocorporate executives and to the business units in a series of meetings. Corporate IS donot have the benefit of seeing an BU plans or in participating in BU planning before orduring this process. We rated the connections between the planning processes as"negotiated", reflecting the fact that although the plan is produced in isolation from the128planning done in the business units, it is subsequently presented to them for approvaland/or modification.5. Other Factors within Company BSince the Director of IS joined the corporate IS group, he has been very aggressivein trying to influence the overall technology direction of the company to rationalize it andcut costs. He has the private backing of the CEO, who selected him for the job.We examined the content of a letter written from the IS Director to the CEO. Itspurpose was to summarize a recent meeting between them. In it, the IS Director outlinedhis understanding of the "business issues", the "technology issues" and the expectationof the CEO for him in his new role. On the latter point, he mentioned "standardizedtechnology environment..., not too much technical risk..., enterprise-wide approach tomanaging information technology". He asked the CEO to "communicate that messageclearly to all executives" . However, the CEO has not issued any statement regarding IT.Therefore, the IS Director has no "official" mandate from the CEO to make the changeshe is initiating. This situation detracts from his effectiveness because business units feelthey have the right to "opt-out" of any change they do not agree with.D. Summary and Analysis - Company BTable V.2 contains a summary of the findings on the factors and linkage.Relationships between the factors and linkage are discussed after the table.Table V.2A Summary of the Factors and Linkage Ratings for Company BFACTORS/LINKAGELinkage between Corporate Business andCorporate IT ObjectivesLinkage between Business Unit and Corporate ITObjectivesHistory ofImplementationof IT PlansAchieved reduction in budget of 5% last year, whichwas in line with the company fmancial objective.Previous implementation was judged to be successful.The new IS Director has made changes to link corporate ISmore tightly to the objectives of the business units.The two technology projects have not been accepted widelywithin the business units.The organizational changes are viewed favourably, but it istoo early to judge the results.Mixed results.SharedknowledgeCorporate executives have no recent experience in IT.Corporate IS has a moderate level of businessexperience.Several corporate executives (CEO, SVP, CorporateResources, IS Director, Controller) all come from thesame BU.BU executives have very limited experience with IT.The new IS Director Corporate has a high level of businessexperience and comes from one of the business units.Communicationbetween IS andbusinessExecutivesNo regularly scheduled meetings between corporate ISand corporate business executives.Communication is infrequent, and focused^on IT.Very little direct contact and no permanent team to facilitatecommunication between corporate IS and BU executives.BU IS and corporate IS execs meet bi-monthly onTechnology Committee. (permanent committee)Communication is infrequent and focused on IT.N-0Table V.2A Summary of the Factors and Linkage Ratings for Company BFACTORS/LINKAGELinkage between Corporate Business andCorporate IT ObjectivesLinkage between Business Unit and Corporate ITObjectivesConnectionsbetween IT andbusinessplanningThree rounds of planning produce plans for corporatedepartments and business units. There are no overallcorporate objectives.Corp IT objectives are negotiated in meetings withCorporate executives.BU execs review Corporate IT plans in individual meetings.Corporate IS does not participate in planning with or seeBU plans.Planning processes are rated as negotiated.Other Factors The IS Director has a mandate from the CEO. The mandate from the CEO to the IS Director is not publicand, therefore, is not enforceable.LINKAGEa) writtenb) underst'g (EXECS/IS)c) Visiond) Subj. Ass.(EXECS/IS)OVERALLMODERATEHIGH/UNKNOWNLOWHIGH/UNKNOWN???LOWMODERATE/LOWnot measuredMODERATEMODERATE1311. Linkage between corporate objectives and corporate IT objectives.In Chapter IV, we concluded that the mixed ratings on the linkage measures madeit difficult to assign an overall linkage rating for Company B. An examination of therelationships between factors and linkage resulted in a re-examination of these findingsand a modified linkage rating. Both are discussed below.a) Stage One: The Relationship between Antecedents and Current PracticesThe new Director of IS has increased the amount of shared knowledge withinCompany B. This factor alone might have marginally increased the communication fromprevious levels. However, there is another connection between many of the corporateexecutives, which we believe has had a similar, but stronger affect on current practices.The CEO has moved members of his previous business unit to positions of influencebeneath him. These moves include the SVP of Corporate Resources (who is in charge ofcorporate IS), the IS Director, and the Controller. These people have all worked togetherin a highly successful business unit and we hypothesize that they have high levels of trustwhich effect linkage in two important ways. First, the CEO is talking directly to the ISDirector, which is unusual since they are separated by two levels of hierarchy. Althoughtheir communication is infrequent, they are able to discuss strategic issues and have cometo a shared understanding of IT objectives (the CEO says "they know their role... theyare the policemen".)Second, although there are no written corporate objectives, the SVP of Corporateresources acts as a "linking pin" between the executive committee and corporate IS.Because this person is closely aligned with the CEO, he assists the VP of IS in puttingIT objectives together which meet corporate approval. Therefore, the shared experienceof the corporate executives seemed to influence both the effectiveness of the132communication and the connections between planning. As Pyburn (1983) showed, in acompany without written objectives, the IS executive needs to have proximity to thecorporate executives to be judged successful.b) Stage Two: The Relationship between Current Practices and Observed LinkageCompany B lacks two of the hypothesized prerequisites for linkage:corporate objectives and frequent communication. The corporate IT planning process doesnot receive any direction from corporate executives. However, the executives rated HIGHon their understanding of IT objectives and their rating of linkage. This is due to theplanning process developed by the VP of IS. Each year, the VP of IS, with the ISDirectors, created statements of "corporate objectives" as part of the IT planning processand developed corporate IT objectives which were derived from to these statements. TheIT plans were then reviewed by the Corporate SVP, who, as mentioned earlier, is closelyconnected to the CEO.The VP of IS created a connection in planning when he took this plan "on theroad" to the executives so they could correct his assumptions about corporate objectivesand alter the IT objectives as required. The IT plans were reviewed by the controller andother corporate departments. In this way, the planning process transcended the lack ofcorporate objectives and influenced linkage.We have re-rated the linkage between corporate objectives and IT objectives asMODERATE to HIGH" based on this analysis and suggest that Company B is evidencethat, as we suspected, the lack of formally stated corporate objectives does not precludelinkage.48 The reason it is not HIGH is that the CEO stated that he could not see why corporate IT objectivesare needed "any more than we need objectives for the telephone." This opinion may be one reason fora lack of corporate IT vision.CommunicationbetweenBusiness andIS Executives• SharedExperience inthe sameBusiness UnitShared Know-ledge betweenBusiness and ISExecutivesImplementationof PreviousIT PlansConnectionsbetweenBusiness andIT PlanningProcesses/LINKAGE133Figure V.4 shows the relationships discovered between the factors and linkage.Figure V.4Corporate IT to Corporate Business Linkage:Causal Relationships in Company B2. Linkage between BU objectives and corporate IT objectivesa) Stage One: The Relationship between Antecedents and Current PracticesIn many interviews with BU executives, they mentioned the IS Director, whojoined corporate IS from one of the business units eighteen months ago. His wealth ofline experience has increased the level of shared knowledge and has improved the134communication between the business units and corporate IS. He has been asked for adviceby two of the SVPs and a third extends an open invitation to him to attend projectmeetings.On the negative side, many executives in the business units still believe that mostof the people in corporate IS are technology focused, not business focused. This makesthe business units reluctant to include them in any meaningful way in their planning.Corporate IS is thus isolated from business unit planning.b) Stage Two: The Relationship between Current Practices and Observed LinkageThere are no formal communication channels between corporate IS and BUexecutives and very little direct contact between them. Most communication betweencorporate IS and business units happens at the IS level, between corporate IS and BU ISpeople. They meet informally based on old friendships and formally in the TechnologyCommittee. Unfortunately, no BU objectives are discussed in this forum and so this richcommunication channel has no effect on increasing the connections in planning processes.The lack of involvement in BU planning has resulted in low levels ofunderstanding of BU objectives exhibited by Corporate IS executives. Because theCorporate IT planning process is isolated from BU planning, the corporate IT planconsists of generalities and is not of much interest to the business unit executives.The annual process of taking the corporate IT plans to the business units forratification has resulted in some understanding of IT objectives exhibited by BUexecutives although, as noted above, not much excitement.Another factor which emerged from the data is a disagreement about corporate ITobjectives. As mentioned in the detailed writeup, the IS Director has introduced severalinitiatives designed to standardize technology and methodology within the company.Implementationof PreviousIT PlansLINKAGEShared Know-ledge betweenBusiness and ISExecutivesCommunicationbetweenBusiness andIS ExecutivesConnectionsbetweenBusiness andIT PlanningProcesses135These initiatives are being resisted by BU IS executives and have resulted in low ratingson linkage from them. This disagreement is occurring at the IS level in the company andhas not yet been raised to the executive level, which is why the BU executives, ininterviews, exhibited no understanding of the corporate IT projects.Figure V.5 shows the connection between factors and linkage.Figure V.5Corporate IT to Business Unit Objectives:Causal Relationships in Company BDisagreementabout ITObjectivesbetween BU and?Corporate !1136c) Why is there disagreement about Corporate IT objectives?This disagreement is similar to that discussed in Company A, although the BU ITobjectives in Company B are not as clearly defined. There are several explanations forthis disagreement.First, this is a power struggle between corporate and BU IS groups. There are twolarge business units in Company B, each with over 100 people in them. They havesignificant technical and applications development expertise and have been free for thepast several years to create their own environments. Both groups naturally resententerprise-wide corporate IT initiatives which limit their freedom to act.Second, this is rivalry between two business units. The IS Director, who ismasterminding the IT initiatives, comes from one of the big business units. It is the VPof IS from the other unit who is resiting the initiatives most strenuously. In interviews,he explained that the new IS Director came from the other business unit and thereforedoes not have an understanding of their needs and cannot create enterprise-wide solutions.In practise, he has made little attempt to influence the direction of the initiatives (e.g. herefused to put his people on the project teams), and relied on his considerableorganizational power to combat them after they were introduced. There is a long historyof rivalry between these two business units, possibly exacerbated by the actions of theCEO to surround himself with people from his previous business unit.The third explanation is an interaction between a lack of mandate for corporate ISand a power struggle between business units and corporate IS. The IS Director is actingas if he has the right to create enterprise-wide solutions. The business unit IS people donot agree that there are any "shared" problems. The CEO has provided the IS Directorwith a private mandate to make the necessary changes promoting organizational efficiencybut has not signalled this intention to any other executives. Therefore, the BU IS peoplefeel justified in resisting these changes. This explanation is consistent with the finding that137mandate from the CEO is an important factor in attaining linkage (Lederer andMendelow, 1989).These explanations will be analyzed with those from Company A in the Across SiteFindings section of this chapter.d) Why is there no vision for IT within Company B?In company B, there was a vision for IT expressed by the IS Director" but notshared by others. There were no other strong visions expressed by the executives or byother corporate IS executives. The reasons for this might include:1) There were no corporate objectives to tie a vision to.2) The IS Director's vision was very technology-based, and did not capture theimagination of the executives.3) There was a low level of recent IT experience or IT awareness at the senior level inCompany B and the executives were not capable individually of creating an IT vision forthe company.4) There were no lateral relationships (e.g. an IS steering committee) in place withinwhich executives in Company B could communicate among themselves and create an ITvision.These explanations will be compared with those from Company A in the AcrossSite Findings.E. Factors - Company CTwo corporate business executives, two corporate IS executives and three SVPsof business units were interviewed within Company C. In addition, the annual IT Plan,49 "an open systems, standards-based approach ..."138minutes from the Senior IS Steering Committee meetings, and corporate IS reports werereviewed. Details on interviewees and documents were presented in tables 111.2 and 111.3.A summary of the company characteristics is contained in Appendix B.1. Implementation of Previous IT PlansIn the mid 1980's, Company C aggressively purchased personal computers andinstalled an organization-wide office automation package which is widely used andconsidered to be very integral to the functioning of the company. It decided to consolidatethe two mainframes in 1986 and began the migration to a new mainframe in 1988. Sincethen, no written IT Plans have been created. Programmers and analysts have moved fromproject to project as one business unit after converted their applications to the newmainframe.The opinion of senior management at Company C is that IT activity over the pastfew years has not been very productive. As the CEO said "after we installed OfficeAutomation in 1986, we have done nothing important. We have essentially stood still withrespect to new technologies because we have been implementing mainframe softwarepackages and transferring to the IBM." The SVP of Corporate Services agrees "I thinkwe were better positioned three years ago. We were very aggressive in introducing PCsin to the company. Then in 1988, we took delivery of the IBM and spent the last few yearsdoing this absolutely useless work of conversion." .2. Shared Knowledge of IS and Business ExecutivesBecause the managers who influence IT objectives within Company C are not allSVPs, table V.3 has been created to show the level of shared experience of eachexecutive who sits on the Senior IS Steering Committee or who is an SVP of a BU. SeeAppendix D for an explanation of the ratings.139The IS Director is a career IS professional and has been with Company C, first asa consultant, and then as an IS employee, for two years. He has held no non-IT positionswithin Company C and has no other insurance business experience. We classify hisknowledge of the insurance business in general and of Company C specifically as "low".The SVP of Corporate Services, an actuary, has no formal IT management trainingor experience beyond actuarial programming. He has, however, had the IT functionreporting to him since 1985 and wrote the major IT direction report after the merger in1986. His knowledge of the general insurance business and Company C's business inparticular is extensive. His understanding of IT management is rated as "moderate" andhis knowledge of current technology as "moderate".The CEO, an actuary, has no formal training in IT other than a course in BASICand has no line experience managing the IT function or an IT project. He is however,interested in computers "from both a strategic and an efficiency perspective" . He is asophisticated self-taught user of several PC packages and regularly reads PC periodicalsas well as articles on technology in the general business literature. The CEO is anenthusiastic supporter of OA and in a recent usage survey, was rated as the most frequentuser. His understanding of IT management is rated as "low" and his knowledge ofcurrent technology as "high".The SVP of one of the business units has a Masters degree in Computer Scienceand has spent several years with previous employers as an IT professional - a systemsanalyst, project leader and manager of the IS function. At Company C, he acted as theexecutive sponsor for the major conversion project. His understanding of IT managementis rated as "high" and his knowledge of current technology as "moderate".The SVPs of the other two business units have had no previous IT experience.Neither of these executives sit on the IS Steering Committee. Their knowledge of bothIT management and IT technology are rated as "low" .140The Director of Finance for a business unit is a current member of the IS SteeringCommittee. He has a Computer Science degree, five years technical IS experience, 5years marketing IS products, several years experience as an IS consultant. Hisunderstanding of IT management is rated as "high" and his knowledge of currenttechnology as "high".The VP of another business unit is a current member of the IS SteeringCommittee. He has no previous experience with the management of IT. For the past year,he has been extensively involved as manager of the major conversion project. Hisunderstanding of IT management is rated as "moderate" and his knowledge of currenttechnology as "low" .Table V.3A Summary of the Shared Business and IT Experience at Company CPosition* = IS Steering CommitteeInsuranceExperienceCompany CExperienceIT ProjectManagementExperienceAwareness ofNewTechnologyCEO * High High Low HighSVP, Corporate Services*High High Moderate ModerateDirector, Corporate IS * Low Low High HighSVP, BU 1 * High High High ModerateSVP, BU 2 High High Low LowVP, BU 2 * High High Moderate LowSVP, BU 3 High High Low LowFinance Director, BU 3*Moderate Moderate High HighThere are members of the Senior IS Steering Committee who have had extensiveexperience with IT in their working careers. The VP of Corporate IS has a low level ofknowledge, both of the company and of the insurance business.1413. Communication between business and IS executivesCommunication relating to IT objectives should occur between the executives whoset business objectives (i.e. the SVPs and the CEO) and the VP of Corporate IS. As notedin the previous section, however, there are two individuals who act as agents for theSVPs at IS Steering Committee meetings. For the purposes of this analysis, the groupwith which the VP of IS should be communicating will include all SVPs, the CEO, plusthe two extra members of the IS Steering Committee - seven people in total. This logicalgroup will be called "senior management" in this section, since individually andcollectively they have the power to affect corporate IT strategy.Communication will be analyzed using two of the elements in the Galbraith lateralrelations typology: direct contact and permanent teams. None of the other elements arepresent in Company C.a) Direct ContactThere is frequent contact (two or three times per week) between the VP of IS andhis two superiors, the SVP of Corporate Services and the CEO. Discussions between theIS Director and senior management, when they occur, are largely operational in nature,revolving around projects on which IS staff are deployed. Discussions between the CEOand the VP of IS have the characteristics (future-orientation) necessary to create or evolveinto IT strategy.There is little direct communication between the IS Director and any of thebusiness unit SVPs.b) Permanent TeamsThere is no committee within which the VP of IS can communicate with allmembers of the senior management group. There are two committees which contain142various members of this group, but no committee contains the SVPs, the CEO and theVP of IS.The Executive Committee consists of the Senior Vice Presidents and the CEO. Itexcludes the VP of IS, since he does not report to the CEO. They meet to discuss itemsof corporate interest and to prepare the annual budget submission to the Board. Nominutes are taken at this meeting.The regular members of the IS Steering Committee are the CEO, two Senior VicePresidents, one Vice President, and one Finance Director. They meet monthly to discussIT issues - both future and current. These meetings are the most important vehicle formaking IT decisions, creating IT objectives and for the VP of IS to understand corporateobjectives. They are held monthly and last approximately two hours. The agenda isprepared and the meetings are minuted by the VP of corporate IS under the guidance ofthe CEO. The CEO chairs the meetings.An analysis of the minutes from the steering committees show that most of thediscussion revolves around ongoing large projects within the business units. Although thetime of the Steering Committee is largely taken up by the progress reports on theseprojects, IT issues and strategies are discussed from time to time. Not all business unitsparticipate equally at these meetings, and the representative from the business unit whichmakes the most sophisticated use of technology rarely tables any items for discussion.In summary, the existence of the IS Steering Committee and the regularity of itsmeetings are very beneficial to the communication between the VP of IS and certainexecutives within Company C. Without it, the VP would have little access to businessunits. The committee, however, tends to focus on tactical issues and avoid the strategicdirection-setting work which might lead to corporate IT objectives.Decisions made about the budget, which influences to a great extent which ITprojects are approved, are made by the Executive Committee. Two out of five of the143executives on this committee have not been directly involved in IS Steering Committeediscussions. This has caused mis-communication between the VP of corporate IS and theExecutives. Two examples illustrate this point.1. The Executive Committee cut the provision for database management softwareout of the budget in December 1989 and then had to petition the Board of Directors toput it back in in January 1990 after the project manager refused to begin the projectwithout it. The Steering Committee representative from the affected business unit doesnot sit on the Executive Committee.2. In the 1991 budget discussions at the Executive Committee, one SVP, who doesnot sit on the IS Steering Committee, supported a project to rewrite the accounting systemafter he and others had blocked the project for several years. The VP of corporate IS, notbeing privy to the discussions within the Executive Committee, did not understand thatthis change of decision signalled a significant change of direction for the whole company- from aggressive expansion to cost-cutting measures - and did not change the ITobjective accordingly.In summary, the communication between the VP of IS and corporate executivesis frequent and diverse. The communication between the VP of IS and BU executives isinfrequent and focused on IT topics.4. Connections between Business and IS PlanningBusiness planning within Company C is a process whereby the financial objectivesset by the parent company and by regulatory bodies are used by the Business Units asinput in their strategic planning process. They create product/market and administrativesupport plans and then create a budget. These budgets are then amalgamated at thecorporate level and reviewed by the Executive Committee.Company C does not create an overall corporate strategy. As the CEO remarked,"My job as far as I am concerned is to make sure that the strategy of the threeindependent units mesh together and one isn't going to draw more from the company thanwe can afford." One SVP said "There is no strategic planning at the corporate level. We144feed our numbers into the financial projections, not our strategies."The strategic direction of the company is effectively set by the business units ratherthan by corporate management.Because of the large conversion projects, Company C's efforts at IT planning overthe past five years have been primarily tactical and operational. As the time approachedfor the corporate budgeting process for 1991, the SVP of Corporate Services felt that,with the conversions being nearly completed, it was time to plan ahead. The 1991 IT plandescribes and discusses the current IT scenario and makes recommendations for the 1991fiscal year. The steps taken to formulate this plan were as follows:a) The corporate systems (e.g. claims, general ledger, personnel) and mainframecapacity were assessed by IS people and recommendations concerning the need to upgradethem were prepared.b) The idea of procuring a 4GL to improve programmer productivity was raisedby the SVP of Corporate Services and the VP of IS and recommendations to this effectwere included in the plan. This idea was not discussed with the business units.c) The business units were canvassed by the SVP and the VP of IS to determinetheir IS activities planned for next year and these were listed in the IT plan.This plan was discussed and approved at a Senior IS Steering Committee meetingwith very few comments from Company C management or the steering committee. Inpart this was because the activities suggested were tactical (hardware additions) andinternal to IS (addition of a database support person). Another reason was that the planslaid out for the business units were not seen by them to be binding in any way.Comments made during the interviews indicate that the connection between business unitplans and what was written in the 1991 IT plans was very tenuous."I don't know how all of the individual projects got into the IT plan. We haven'tdiscussed it inside the business unit yet.""We would not want to be held to what's in the current version of the IT plan, it145doesn't reflect exactly what we would like to happen.""There is not a lot of commitment from the business units to keep to this plan."Business units have their own agenda for product introduction and product changesand these plans will determine, to a large extent, the nature of the development workcarried out on their behalf in 1991. As long as the hardware platforms that they areworking on remain stable, the 1991 report is of little consequence to them. The onlyeffect on them will be the allocated costs of the IS department and these areapproximately the same as 1990.Corporate IT Planning is a combination of 1) bottom-up issue resolution forcorporate systems and the mainframe (e.g. rewriting old systems, mainframe capacity,computer security), 2) introduction of new methods of systems development and delivery(e.g. 4GL), and 3) acknowledging business unit plans. Since there are no corporateobjectives, the IT plans have nothing to link to and the corporate IT planning practicesare isolated from corporate planning. Because the corporate IS people are not involvedin any way with BU planning, corporate IT planning practices were ranked as beingisolated from the BU planning practices.5. Other FactorsThere are several other factors which seem to have affected the attainment oflinkage within Company C.a) Differences of Opinion about the Role for ITThe two people who are the direct superiors to the VP of IS at Company C differabout the future role of IT. The CEO, judging by the number of times technology wasmentioned in the Annual Report and by statements he made in interviews, believes that146IT is crucial to the success of the firm and that "total computerization" is the goal. TheSVP of Corporate Services seems to take a more fiscally conservative, low profileapproach to IT, allowing the "cost-effective" use of IT to dominate all other views. Bothexecutives have enough interest in and experience with IT to believe in the correctnessof their personal view. However, neither has worked in a company which hassuccessfully implemented IT in support of corporate strategies. Neither seems to be ableto create a plausible vision of the future and a set of strategies that will convince the otherto abandon his position.A simple difference of opinion between two senior executives would not be tooimportant except that these two people are both very influential leaders with respect tosetting the strategy for IT. The CEO chairs the IS steering committee meetings anddominates them. The SVP of Corporate Services prepared the 1991 IT Plan whichcontained proposals for major new corporate software and hardware acquisitions as wellas corporate application projects.There are two parties who could potentially mediate between these polarizedpositions and resolve the issue: the VP of IS and the IS Steering Committee. Neither hasbeen effective, however, probably because of the high rank of the two executives.b) Ineffective IS Steering CommitteeThe Steering Committee is not functioning as a forum for sharing ideas and dealingwith conflict. Potentially, it could mediate between the visions of the top two executivesand choose the IT strategy for Company C but it seems to avoid any "process" role -focusing instead on the contents of IT budgets and the progress of projects. How the ISfunction is to be managed is not discussed. Perhaps this is because the views of thebusiness units would add another dimension since they want to control systemsdevelopment and both the CEO and the SVP would not favour that approach. So the147differences do not get aired.The Steering Committee also does not impose any management control on thecorporate IS department. For example, many interviewees have suggested that the lackof a report writer is a significant problem, leaving them with virtually no managementinformation. It was raised at a March 1990 Steering Committee meeting and is minutedas follows: "A request for a report writer tool to access data on the mainframe systemshas been supported by all Divisions... IS will develop a recommendation .... for userreporting" . Twelve months later, this had not been done.F. Summary and Analysis - Company CA summary of the factors and linkage within Company C is shown in Table V.4. Ananalysis follows.Table V.4A Summary of the Factors and Linkage Ratings for Company CFACTORS/ LINKAGE Linkage between Corporate Business andCorporate IT ObjectivesLinkage Between Business Unit and Corporate ITObjectivesHistory of Implementationof IT PlansVery little IT progress achieved in the past 3years due to the conversions. Unsuccessful.The conversion projects in the last three years have beenvery unsuccessful.Shared knowledge Many corporate business executives have ITbackground and interests.The VP of IS has no business experience.Two of three SVPs have no IT experience at all.The VP of IS has no business experience.Communication between ISand business ExecutivesFrequent direct communication betweenCEO, SVP of Corporate Services and theVP of IS.IS Steering Committee meets monthly(permanent team).Communication is frequent, and diverse.There is very little direct contact between corporate ISand BU executives.Only one SVP sits on the IS Steering Committee, theother two send delegates.Communication is infrequent, focused.Connections between IT andbusiness planningThere is no corporate level strategicplanning process at all.Corporate IT plans are prepared internallyand then presented upwards.Planning processes are rated as isolated,since there are no corporate objectives.BU execs review Corporate IT plans only if they sit onthe IS Steering Committee.Corporate IS does not participate in planning with thebusiness units or see the BU plans.Corporate IT planning processes are rated as isolatedfrom BU business planning.Table V.4A Summary of the Factors and Linkage Ratings for Company CFACTORS/ LINKAGE Linkage between Corporate Business andCorporate IT ObjectivesLinkage Between Business Unit and Corporate ITObjectivesOther Factors Difference of opinion about the role of ITbetween CEO and SVP of CorporateServices.Ineffective IS Steering Committee.LINKAGEa) in written documentsb) understanding ofobjectives (EXECS/IS)c) Shared Visiond) Subjective Assessment(EXECS/IS)OVERALL LINKAGERATINGUNKNOWN - no corporate objectivesHIGH/UNKNOWNLOWLOW/LOWUNKNOWNLOWHIGH/LOWnot applicableLOWLOW1501. Linkage between corporate business and corporate IT objectivesa) Stage One: The Relationship between Antecedents and Current PracticesCompany C is smaller and less formal than company A or B. The high level ofknowledge about and interest in IT among the corporate executives has led to a high levelof communication. The VP of IS regularly talks to the CEO and the SVP of Finance andthe Steering Committee meets regularly.This high level of communication has not produced a high level of connection inplanning practices. The IT history has been very unsuccessful in the past two years, withmajor overruns in project budgets and timeliness. One major result of this history is thatthe Steering Committee spends almost all of its time discussing current projects in orderto prevent similar disasters. Therefore, the potential of this permanent team is diminishedbecause past history of IT initiatives appears to have influenced the level of thecommunication and an opportunity to turn communication into a discussion of sharedobjectives is missed.Another reason for the low level of connection between planning processes is thecomplete lack of corporate objectives. The CEO has decided to run the organization asa loose coalition of business units and no discussion of objectives takes place at theExecutive Committee.b) Stage Two: The Relationship between Current Practices and Observed LinkageAlthough communication is frequent within Company C, most of the indicators oflinkage were rated LOW. The only high one, executive understanding of the ITobjectives, is explained by the presence and the regularity of meetings of the SteeringCommittee. Both the SVP of Finance and the CEO sit on this committee.Although the CEO and the VP of IS have many discussions about the future of ITwithin the company, the CEO's vision for IT has not been implemented or even planned.151There is a serious disagreement (about the value of IT for the company) between the CEOand the SVP of Finance which has prevented the VP of IS from proceeding. The CEOand the SVP of Finance are his two superiors and he has not been able to formulate ITobjectives which will satisfy both executives simultaneously.The difference of opinion may have influenced the company's ability to producemeaningful IT objectives as reflected in the low subjective assessments of linkage.c) Why is there a difference of Opinion about the value of IT within Company C?We can interpret the difference of opinion between the CEO and the SVP ofFinance in two ways: as a power struggle or as