"Business, Sauder School of"@en . "Marketing, Division of"@en . "DSpace"@en . "UBCV"@en . "Mehra, Gaurav"@en . "2009-06-15T22:28:38Z"@en . "1999"@en . "Master of Science - MSc"@en . "University of British Columbia"@en . "This thesis was the result of a study conducted for the call-centre at the Workers'\r\nCompensation Board of British Columbia (WCB). The management at WCB wanted to\r\nunderstand the nature and pattern of calls at their newly opened call-centre. The purpose\r\nof this was to provide an efficient customer service while streamlining the flow of calls\r\ncoming to the call-centre.\r\nAn extensive data collection exercise was undertaken at the call-centre and two other\r\nunits of WCB with which the call-centre interacts. The data analysis revealed that a\r\nhigh proportion of calls were related to transfers to these departments. There were also\r\ncalls related to routine inquiries on claim payment cheques and forms that could\r\npotentially be handled by a well designed IVR system.\r\nBased on this understanding the development of an effective IVR system was proposed to\r\naddress the problems that were discovered through documenting the nature and pattern of\r\ncalls. An extensive review of literature was undertaken to design a new system according\r\nto the standard industry guidelines suggested by the best practices and customized to\r\nWCB's business needs. Two alternate scripts were developed after analysing the source\r\nand purpose of calls to WCB. One was 'person specific' and the other was 'task specific'.\r\nThe two scripts were tested on students at WCB through a computer-based IVR\r\nsimulation. The results of the student survey provided evidence that introducing\r\nadditional options and use of simple and clear instructions in the new scripts could\r\npotentially in fact address the problems discovered in the study and they were preferred\r\nover the existing WCB script. The IVR simulation is reconfigurable and can be used in\r\nfuture studies to gather further evidence in support of the results obtained in this thesis as\r\nwell as refine scripts before putting them in a production mode."@en . "https://circle.library.ubc.ca/rest/handle/2429/9198?expand=metadata"@en . "14391182 bytes"@en . "application/pdf"@en . "DEVELOPING AND TESTING AN EFFECTIVE INTERACTIVE VOICE RESPONSE (IVR) SYSTEM FOR THE WORKERS' COMPENSATION BOARD OF BRITISH COLUMBIA By G A U R A V M E H R A M B A (Marketing), Saint Mary's University, Halifax, Nova Scotia, 1996 A THESIS SUBMITTED IN PARTIAL F U L F I L L M E N T OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE (Business Administration) In THE F A C U L T Y OF G R A D U A T E STUDIES (Department of Commerce and Business Administration) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH C O L U M B I A April 1999 \u00C2\u00A9 Gaurav Mehra, 1999 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. The University of British Columbia Vancouver, Canada Date DE-6 (2/88) Abstract This thesis was the result of a study conducted for the call-centre at the Workers' Compensation Board of British Columbia (WCB). The management at WCB wanted to understand the nature and pattern of calls at their newly opened call-centre. The purpose of this was to provide an efficient customer service while streamlining the flow of calls coming to the call-centre. An extensive data collection exercise was undertaken at the call-centre and two other units of WCB with which the call-centre interacts. The data analysis revealed that a high proportion of calls were related to transfers to these departments. There were also calls related to routine inquiries on claim payment cheques and forms that could potentially be handled by a well designed IVR system. Based on this understanding the development of an effective IVR system was proposed to address the problems that were discovered through documenting the nature and pattern of calls. An extensive review of literature was undertaken to design a new system according to the standard industry guidelines suggested by the best practices and customized to WCB's business needs. Two alternate scripts were developed after analysing the source and purpose of calls to WCB. One was 'person specific' and the other was 'task specific'. The two scripts were tested on students at WCB through a computer-based IVR simulation. The results of the student survey provided evidence that introducing additional options and use of simple and clear instructions in the new scripts could potentially in fact address the problems discovered in the study and they were preferred over the existing WCB script. The IVR simulation is reconfigurable and can be used in future studies to gather further evidence in support of the results obtained in this thesis as well as refine scripts before putting them in a production mode. ii Table of Contents Abstract i i Table of Contents i i i List of Tables v List of Figures vi Acknowledgements viii CHAPTER 1. INTRODUCTION AND OVERVIEW OF RESEARCH 1 1.1 Background and Motivation 1 1.2 Call Centre Technology 1 1.3 Importance of IVR research 2 1.4 Background and Motivation for Current Research 3 CHAPTER 2. LITERATURE REVIEW 7 2.1 Previous Research on IVR 7 2.1.1 Huguenard(1993) [2] 7 2.1.2 Bradley et al. (1996) [3] 10 2.1.3 DePalma & Goodman (1995) [4] 12 2.1.4 Theory of Reasoned Action and its Variants 15 2.2 Implications for Current Research 16 CHAPTER 3. UNDERSTANDING THE NATURE AND TYPE OF CALLS AT THE WORKERS' COMPENSATION BOARD 17 3.1 Introduction 17 3.2 Pre-Test 17 3.3 Data Collection at the Call-Centre 21 3.4 Data Collection at Entitlement and Case Management 21 3.5 Problems encountered during Research 24 3.6 Results of data analysis at the Call Centre, Case Management, and Entitlement.. 25 CHAPTER 4. SCRIPT DEVELOPMENT 27 4.1 Evaluation of WCB Script against the Best Practices 27 4.2 Existing and Proposed Call Centre Architecture 29 4.3 Alternate IVR Scripts 31 4.3.1 Addressing the problems through the new scripts 32 4.4 Laboratory v/s Field 34 4.5 Computer-based simulation for the new scripts 35 4.6 Student Questionnaire 37 4.6.1 External Validity Problems 38 4.6.2 Testing the IVR Simulation 39 4.6.3 Internal Validity Problems 40 CHAPTER 5: STATISTICAL ANALYSIS OF THE QUESTIONNAIRE 44 5.1 Summary Statistics and Statistical tests 44 5.2 The Experimental Model and Hypothesis 46 5.3 Conclusions from the Student Questionnaire 61 CHAPTER 6. CONCLUSIONS AND RECOMMENDATIONS 64 6.1 Concluding comments 64 6.2 Limitations of the study 65 6.3 Suggestions for future research 66 in Bibliography Appendix I ^ Appendix II ^ Appendix III \u00E2\u0080\u00A2 ^3 Appendix IV ^ Appendix V \" Appendix V I iv List of Tables Table 1. Breakdown of different types of calls 18 Table 2. Evaluation of the current IVR system against the Best Industry Practices 28 Table 3. Objectives and Goals for an IVR System for WCB 30 Table 4. Laboratory vs. Field Method 36 Table 5. Design of experiment for Student Survey 42 Table 6. Result of student responses on IVR vs. Operator 62 Table I L L Calls suitable for an IVR 84 v List of Figures Figure 1. Screen-shot of user interface for the data collection tool at the call-centre 20 Figure 2. Screen-shot of user interface for the data collection tool at Entitlement and Case Management 23 Figure 3. Existing Call Centre Architecture 31 Figure 4. Proposed Call Centre Architecture 32 Figure 5. Screenshot of main user interface for the computer-based IVR simulation 37 Figure 6. Screenshot of configuration interface for the computer-based IVR simulation 38 Figure 7. Plot of mean intervals using the Fisher's Least Significant Difference 60 Figure II. 1 Percentage of time spent by CSRs on various activities during phone-times 72 Figure II.2 Calls by people by time of the day 73 Figure II.3 Types of calls by time of the day 74 Figure II.4 'Transfers To' by time of the day 75 Figure II.5 Calls by people by day of the week 76 Figure II.6 Type of calls by day of the week 77 Figure II.7 'Transfers To' by day of the week 78 Figure II. 8 Breakdown of Transfers 79 Figure II.9 Average length of Transfers 80 Figure 11.10 Analysis of'Wrap-up' 81 Figure II. 11 Service and Wrap-up times 82 Figure 11.12 Average Wrap-up time .83 Figure III. 1 Incoming calls by categories of people 86 Figure III.2 Breakdown of incoming calls by source 87 vi Figure III.3 Average time spent on calls with various people 88 Figure III.4 Average time spent on different types of calls with different people 88 . Figure III.5 Incoming calls by categories of people 89 Figure III.6 Incoming calls by source 90 Figure III.7 Average time spent on calls with various categories of people 91 Figure III.8 Average time spent on various types of calls with different people 92 Figure V . 1 IVR Simulator user interface .' 100 Figure V.2 File Menu 101 Figure V.3 Simulation 101 Figure V.4 Configuration Dialogue 102 Figure V.5 Output Configuration 103 Figure V.6 Input Configuration 104 Figure V.7 Error Handling 105 vii ACKNOWLEDGEMENTS I would like to acknowledge the assistance of Dr. Marty Puterman, Stephen Jones, Paul Hiom, and Lauren Gray at the Centre for Operations Excellence (COE) at the University of British Columbia for the financial assistance they provided for this study. I would forever remain indebted to my advisor Dr. Carson Woo who provided every possible resource in order for me to complete this work. His suggestions during weekly meetings at COE greatly improved the quality of not only the WCB project but also several others. My co-advisor Dr. Izak Benbasat who guided the design of the student survey and provided the research instrument developed by him for that purpose. His insightful comments during the 'Research Methods' course helped put this study on a firm ground. I would like to thank research assistants Katy, Jeff, Anoop, Art, and Leo who spent countless hours at the Workers Compensation Board of British Columbia gathering data on the phone calls at the Call Centre, Entitlement, and Case Management units. Marie-Ann Anderson, Jason, and Dorit for helping me analyze the enormous data on calls. Arash who due to his interest in my work agreed to develop the computer-based IVR simulation despite his last semester at UBC. Ednoi whose voice added a touch of professionalism to the sound files that she recorded for the IVR simulation. Colleen who frantically gathered students to participate in the IVR simulation, a day before my departure to Montreal. Lana Burnstad and Carole Michelucci at C N for facilitating my Vancouver trips in order for me to finish my study. My Manager in IT, Doug Wilson, who let me take off for a week to Vancouver to run the laboratory simulation, and also promised to take me out to dinner in one of the finest restaurants of Montreal should I make it to the May graduation deadline. My Director in Marketing, Keith Reardon, who sent me back from Toronto a day before the deadline in order to ensure that I have the thesis in the library by Friday, April 30 t h, 5:00 pm. Special thanks to my best friend Rafael who while being in Spain, was in constant touch with me via phone and email guiding the statistical analysis of the student survey. Numerous other friends, both in Vancouver and Montreal, who provided me encouragement and moral support throughout this study and at a time when I was going through perhaps one of the most turbulent phases in my personal life. Finally, I would like to thank my Mom and Dad. Especially my Dad who made me dread my phone calls home as his first question always used to be not \"How my job at C N is going?\" but \"How much progress I have made on my thesis?\" viii CHAPTER 1. INTRODUCTION AND OVERVIEW OF RESEARCH 1.1 Background and Motivation Call-centres are increasingly becoming common in many large and even medium and small-scale organizations and businesses. The growing importance of the call-centres for organizations is due to the fact that it assists in conducting the business of the organization in a smooth and efficient manner and also provides a high level of customer service. The call centre agent acts as the primary interface between the customer and organization and is thus an important element in the value chain. Establishing a call-centre in an organization is an important decision and can have a wide-ranging impact on the information flows and business processes within an organization. The call-centre also interacts with other departments in an organization and it is crucial to establish a clear role for the call-centre and its inter-relationship with other departments. 1.2 Call Centre Technology The technology behind the call-centre is fairly mature and there are various tools available to establish a leading-edge call centre. In order to provide the highest level of customer service all the components of a call centre are required, however, an organization may decide to choose only a few components due to resource constraints. The basic component of a call centre is an Automatic Call Distributor (ACD) which routes incoming calls to a free agent and is similar to a Public Branch Exchange (PBX). The number of incoming trunks into the A C D varies from call-centre to call-centre and depends on the estimated demand and size of the business. Most call-centres usually incorporate an Interactive Voice response (IVR) system which is an automated inquiry line that interacts with corporate databases to provide up-to-date information to callers. Another component of a call-centre is Computer-Telephony Integration (CTI) that co-1 ordinates information entered in the telephone by the caller with the information on the agent's computer. However, CTI technology is expensive and not all call-centres possess that, though because of its many advantages in providing a high level of customer service it is becoming increasingly important in call-centres. The call centre agents are equipped with a telephone and a computer terminal, which is connected with the company's databases. Information is retrieved from and entered into the databases by the agents. 1.3 Importance of IVR research There are many successful examples where businesses have improved their operations and increased customer satisfaction through the implementation of an IVR system. An IVR system provides users with round-the-clock access to variable rather than static information [1]. The use of a well-designed IVR system can help in reducing the number of calls that are to be answered by the Customer Service Representatives (CSR). At the same time, it increases customer satisfaction, as the customers are not restricted to the hours of operation of the business if they require routine information. They can call in 24 hours a day, seven days a week. A number of organizations today depend heavily on the IVR system for the successful operation of their business. Some of these are airline companies, banks and insurance companies, courier agencies, educational institutions, etc. Some of the traditional applications of the IVR are providing account balances, checking status of a package or claim, requesting or providing information, or allowing users to register for various activities. New and innovative applications are constantly being developed to streamline business processes. However, the selection of the right IVR system is important. It is essential to have caller-friendly and easy-to-understand voice scripts, prompts, and logical paths between 2 responses [1]. Only after a careful selection and planning of an IVR system the expected benefits will be realised. It is estimated that 80% of the costs associated with a call-centre are related to employees. Given the high personnel costs it is essential that an optimal number of call centre staff be maintained and their time optimally utilised. However, even with a good IVR system, it is usually assumed that the caller will listen to all the choices before making a selection. This, however, is not always the case. Many callers do not wait to listen to all the choices. They simply press '0 ' to talk to a live person. This is based on the assumption that pressing '0' will in fact provide them this opportunity. Due to this tendency in the user behaviour many IVR scripts are modified to give an option other than pressing '0' to users to speak to a call-centre agent. Some others require users to press '0 ' twice before connecting them to a live operator. Research conducted by some businesses has also indicated that just switching the order of choices can make a big difference in encouraging or discouraging users from choosing certain options. There is a need for research to be conducted on the effectiveness of various IVR scripts in encouraging its use by the callers. There are standard industry guidelines that assist in designing a good IVR script and avoiding some of the pitfalls. However, these standard guidelines have to be customized to suit the particular needs of the business. 1.4 Background and Motivation for Current Research This research was conducted on a real-life call-centre at the Workers Compensation Board (WCB) of British Columbia. The Workers Compensation Board of British Columbia deals with the assessment, entitlement, and adjudication of claims related to 3 injuries incurred by workers with respect to job-related duties. WCB opened a centralized call centre in April 1997 for answering all calls related to claims from the lower mainland area. The decision to implement a call-centre was facilitated by a breakthrough technological innovation implemented earlier by the WCB called the E-File (Electronic-File). For years WCB had been handling paper-based claims which required extensive manual tasks. Starting early 1997, WCB has been gradually moving towards an electronic system called the E-File. Any document related to a particular claim (doctor's report, employer report, employee form, etc.) is scanned by a central scanning and indexing department and is added to the claim file which becomes electronically available to authorized claim officers. The customer service representatives (CSR) in the call-centre access any E-file claim on their computers and are thus able to provide information to the callers. The introduction of the E-File and the opening of the Call Centre have had a profound impact on the employees. The implementation of these new information technologies has significantly affected their roles and responsibilities. This research was prompted by management's desire to understand the nature and pattern of calls received by the call-centre, and streamline call-flows through an effective IVR system based on the findings of the study. The research study was conducted in two phases. During the first phase extensive data was collected in the call centre to analyse the nature and pattern of calls coming into the call-centre. The first phase was also devoted to an understanding of the call flows in the call-centre. The findings of the first phase revealed that 20% of the calls were related to transfers to other departments within WCB, specifically Entitlement and Case Management, and approximately 7% were related to cheque inquiries. Based on this analysis it was suggested that an efficiently designed IVR system could potentially deal with some of these calls and reduce the volume of incoming calls to the 4 call-centre. An efficient IVR system is also necessary to provide a high level of customer service as it can provide information 24 hours a day, 7 days a week. The second phase of the study involved gathering data in two other units of WCB: Entitlement and Case Management. These units handle claims that are more complex or claims are passed on to these units from the CSR after a certain time is elapsed. Gathering data in these two units was also essential to complete the call-flow picture for the call-centre. After a thorough understanding of all the call flows and business processes and extensive review of best practices, two alternate IVR scripts were designed. The important research question that was tested in this study was whether the suggested guidelines for designing effective IVR systems, that were incorporated in the design of alternate scripts, would actually address some of the problems. The new IVR scripts were designed to be effective from the perspective of the organization, i.e. WCB. However, they were tested on potential users as the scripts would have been considered to be effective only if it could be demonstrated that they would actually be used by the callers. An effective IVR system would result in cost savings, and labour savings through improved productivity to the WCB. This would happen as more calls would get closed in an effective IVR without the need for a live agent. However these expected savings will not be realised unless the potential callers actually use the system. Hence, the criteria that will be measured in this thesis is the extent of usability of the new IVR scripts as it has an impact on the effectiveness of the system. The two alternate scripts consisted of a 'role specific' and 'task-specific' script that were designed based on an understanding of the nature and pattern of calls at WCB. This led to the exploration of a second research question that, whether one of the two types of scripts would be preferred over the other or there would be no difference between them. However, this question was tested in a situation where there was no clear hierarchy to 5 order the scripts. This means that the result from this test cannot be generalized to IVR system of companies where the two types (role or task specific) have a vastly different and distinct order of choices. These two alternate scripts along with the original WCB script were tested on a group of students through a computer-based simulation. The students were required to respond to a questionnaire during their interaction with the three different IVR simulations. The data that was gathered was analysed using a statistical software package called SPSS. The results of the statistical tests provided evidence that the two new alternate scripts were generally preferred by the group of students over the existing WCB script. However, between the two new scripts there was no statistical evidence to show that one was preferred over the other except for one task. The next chapter reviews relevant previous research conducted on voice response systems. Chapter 3 describes the process of understanding the nature and pattern of calls at WCB. It also summarizes the key findings from the data collection exercise undertaken at WCB. Chapter 4 describes the process of designing two alternate scripts based on the results of data collection and the best practices in the industry. It also contains the methodology for evaluating the three different IVR systems using computer-based simulation and student subjects. Chapter 5 summarizes and discusses the results of the student questionnaire. The final chapter of this thesis contains concluding remarks, limitations of the study and provides suggestions for future research. \ 6 CHAPTER 2. LITERATURE REVIEW 2.1 Previous Research on IVR There has not been much academic research done on the use of Interactive Voice Response technology. The most notable study done in this field was Brian Huguenard Ph.D. thesis at Carnegie Mellon University titled \"Working memory failure in human -computer interaction: Modeling and testing simultaneous demands for information storage and processing.\" Various consulting firms and independent IVR vendors have carried out a number of studies on the effective design and implementation of IVR systems. However, these studies have mostly been done in the U.S. The current thesis was conducted in a Canadian context by utilising a Canadian workers' insurance organization and the findings are not expected to be any different. However, there might be a chance that it might be different. This chapter reviews some of these studies and their main findings. In addition, section 2.1.4 reviews some of the literature based on the Theory of Reasoned Action and its variants. This would provide guidelines for testing the effectiveness of scripts developed in this study by predicting their likeliness of being used by callers. 2.1.1 Huguenard (1993) [2] Brian Huguenard's study was related to the design of better human-computer interfaces through the use of a computational model, which would reduce the need for experimentation to test the impact of alternative designs on user performance. He developed and tested a computational model (PBI USER) for predicting Working Memory (WM) error rates in a particular type of human-computer interaction: phone based interaction. A working version of PBI USER was developed and used to generate predictions about the impact of three factors on W M failure: structure of the interface menu hierarchy. This refers to the various steps users would have to navigate through in order to accomplish a particular task. 7 individual differences in cognitive abilities, which would affect the W M capacity. People with superior cognitive abilities would be expected to have a higher W M capacity and hence lower W M errors rate. - and task characteristics, which refers to task format and number of tasks a user is expected to perform. A complex task for which a user would have to input information or if a user wishes to perform a series of task would affect the W M errors rate. The characteristic that is most relevant for the current research is the structure of the interface menu hierarchy. The research was based on the theory of W M (Just and Carpenter, 1992) that proposes that the storage and processing of information generate demands for W M resources. The predictions of the theory were tested on human subjects by requiring them to interact with a phone-based registration system of a University (Tele-registration). The empirical results provided strong evidence for the importance of storage demands, and moderate evidence for the importance of processing demands as predictor of W M failure in PBI. The results also provided evidence for the importance of individual memory differences in W M capacity as a predictor of W M failure in PBI. Based on the results of this study, Brian suggested several guidelines for the effective design of phone based interfaces. The results of Brian's study have an important bearing on the current research and his design guidelines as well as the best practices in the industry reviewed in the following sections were utilized in the development of an effective IVR script for the WCB. The research conducted by Huguenard was based entirely on interaction with an automated phone-based system without live agent support. The nature of the current research is slightly different as the option to speak to a CSR at WCB is available during regular business hours, i.e., from 7 am to 7 pm Monday through Friday. However, 8 the study is still helpful for the part where a client interacts with an IVR before speaking to a live agent or exclusively interacts with an IVR system outside of business hours. 2.1.1.1 Structure of the Menu Hierarchy Huguenard's thesis reviewed the literature on the structure of the menu hierarchy in terms of the depth v/s breadth of the menu tree. Menu structures that emphasize depth have lesser number of choices at each menu but the number of menus the user must traverse increases. On the other hand, in broad menus the number of traversed menus decrease but there is an increase in the number of choices at each menu. Earlier research in the design of auditory menus suggest no more than three menus at any given level because of working memory limitations, thus forcing a deep and narrow menu structure. However, current theories of the role of W M favour broad menu structures as they postulate that as depth increases (and breadth decreases) the option labels especially at the top-level menus will tend to become more ambiguous. This will degrade subject's understanding of other option labels, which will produce W M errors and inhibit performance. 2.1.1.2 Results and Discussion Huguenard's study measured two types of errors generated by PBI USER: 1) rehearsal errors in which an item of task information was forgotten, and 2) navigational errors, in which an incorrect option choice was made while traversing the menu structure. Only the results of navigational errors are discussed here as they are of interest to the current study. Navigational errors were generated when the user chose an incorrect option. This happens when the option presents only a partial match to the desired goal and the user chooses that option based on the partial match even though the correct option (which has a superior match to the goal) has not yet been presented. This points towards the need for simple and easy to understand instructions in menu options. 9 The results of the study suggested that the impact of the menu structure is mainly on navigational errors. PBI-DEEP had higher navigational error rate than PBI-BROAD. This effect was because of the insufficient long-term memory knowledge for disambiguating the option labels in PBI-DEEP. Based on these findings, the current research uses the design guidelines of a broad menu structure and avoids the use of ambiguous instructions at each level. 2.1.2 Bradley et al. (1996) [3] A major study on IVR systems in American companies was done by the Enterprise Integration Group (EIG), a consulting company in the U.S. The principal author of the report, that is available in the market for approximately U.S. $3,000, was Gary Bradley who was assisted by a team of consultants from the EIG. The team developed a methodology to rate the IVR systems by compiling a list of over 3000 telephone numbers in the Directory selected at random for businesses they thought would likely have an IVR system. Researchers were trained in what features to look for and how to evaluate the IVR systems. They were then asked to call these numbers and evaluate the systems based on the pre-defined criteria. Results for each survey were closely monitored and reviewed by an IVR specialist. Each surveyor evaluated a variety of IVRs across multiple industries. During the actual evaluation process each surveyor called the telephone numbers assigned to them and recorded the total seconds required for the initial greeting and the first menu, navigated all open menus in the application recording each individual option and the key(s) required to select it. They also accessed restricted menus where they could get hold of the required account numbers and codes through the help of friends, relatives, and team members. They also made written notes of their observations and posted their evaluations to a summary sheet. 10 This study utilized a standard scoring mechanism developed by an expert at EIG and divided the observations in four categories with a maximum attainable of 25 points adding up to a total of 100 points for an IVR system. These four categories were Voice Quality, Information Delivery, User Friendliness, and Ease of Operation. A final count of 409 completed surveys was obtained as about 60% of the 3000 numbers enlisted turned out to be live answer, not IVR. Another 25% were automated attendants which just performs basic operations like directory assistance and were not targeted by the study. These figures reveal that the IVR technology is untapped by a lot of businesses and there are tremendous opportunities for cost savings and process improvements by utilizing this technology. This also points towards the need for more research into this area to help businesses realize benefits from implementing a call centre. The study came up with five general findings: 1. Most IVR applications lack a serious customer-oriented design. They are not easy to use and lack desirable features (such as letting the callers know the anticipated wait-time). 2. A large number of IVR applications offer an excessive number of menu options. This might actually inhibit the use of the IVR system as customers get frustrated with too many options. 3. State-of-the-art IVR technology (that includes features like letting the users know the wait time, automatically faxing information, etc.) is not being aggressively deployed. 4. A large number of businesses are not utilizing the CTI technology. WCB is a case in point. 5. And as a result of the preceding two points, enterprises are missing opportunities for proactive relationship management. Based on these findings the research team developed a methodology for designing and implementing an IVR system in organizations. This process is a disciplined multi-step 11 approach for either re-engineering existing applications or creating new applications. The EIG team developed this methodology iteratively, seeking input from a wide variety of subject matter experts. The suggested guidelines for a good IVR system developed by the team consist of 12 steps. These steps are mostly related to simplified design that restricts menu choices to no more than three, and use of simple and straightforward language. These suggestions were tested in this study by requiring student subjects to respond to questions on 'ease of use' and 'ease of understanding' on the student questionnaire, during their interaction with the IVR simulation developed for WCB. In their steps, the EIG team also suggests testing and refining IVR prototypes, before putting it in a production mode. Due to copyright restrictions, details on the 12 steps cannot be disclosed in this thesis. In addition to the methodology the team also suggested specific application design and menu design criteria. This included recommended introduction in seconds, commands for return and cancel, the choice of delimiter, among many others. These suggestions were reviewed and incorporated in the best practices which was utilized for the design of the IVR script for WCB in this study. 2.1.3 D e P a l m a & G o o d m a n (1995) [4] A major research on the use of Automated Response Systems (ARS) was undertaken by David DePalma and John Goodman of the Technical Assistance Research Programs (TARP), Inc. in 1995. The objective of the study were to : 1) define success measures for the use of ARS; 2) discover which ARS applications were most successful and which were not; and 3) identify any credible empirical research measuring the impact on customer satisfaction, call centre productivity, and cost. 12 The researchers sent out mail-in surveys to companies across a wide variety of industries, which included utilities, hotel chains, telecommunication, and software vendors. Some of the large corporations that participated in the study were American Express, A T & T Universal Card, Federal Express, Marriott, Microsoft, and First Union National Bank. The surveys were complemented by follow-up interviews, on-site visits to some of these companies, and analysis of data from their IVR and A C D reports. 2.1.3.1 Results and Key Findings The key measures of success were found to be Opt Out Rate (OOR) and Customer Satisfaction with the use of ARS. OOR is the percentage of callers that choose a menu option to escape the ARS in order to speak with a Customer Service Representative (CSR). The rationale for using these two factors for success was because they found that a large number of companies had installed ARS to reduce the volume of calls coming to the CSRs. Therefore, it was important that a high percentage of the calls that enter the ARS be closed successfully by the ARS without the need for human intervention. Also, customer satisfaction with the ARS system was an important measure for continued use of the ARS system. The most successful uses of the ARS based on the OOR and Customer Satisfaction were found to be: 1) Routing: where callers are directed to the proper group for assistance; 2) Provision of Basic Information such as balance inquiries; and, 3) Simple Transactions such as credit card payments and funds transfer. They also discovered that more complex transactions like bank-by-phone and collections applications used by utilities for automatic payment arrangements were highly successful given proper attention to customer education. 13 The least successful uses of ARS were: 1) Inquiries requiring timely data; 2) Provision of information that may often be an unpleasant surprise for the customer; 3) Technical assistance for complex issues that will vary by customers. The results of data analysis revealed that the simple implementation of Giving Information had the lowest OOR while for more complex applications like Transactions, over half of the callers ultimately ended with a CSR. The number of menu options also had an affect on the OOR. The results revealed that as the number of options increase, the percent of callers that opt out to a CSR goes up. The results also suggested that if the option to speak to a CSR was placed at the end of the first menu it results in more customers using the ARS system. This is because they are forced to listen to all the choices and hence more calls get closed in the ARS system. Finally, customer education was found to be a critical factor in the successful use of the ARS system. The data analysis revealed that for companies that placed usage instructions on billing statements and wallet cards had a lower OOR and a higher level of customer satisfaction than companies that do not provide such information. The study found that the use of Computer Telephony Integration (CTI) technology could provide tremendous gains to the companies in terms of higher customer satisfaction and reduced costs. Sophisticated CTI technology identifies the caller, anticipates the caller's most likely menu choices based on historical calling behaviour, and then moves the anticipated menu choices to the front of a dynamic menu. Even if the caller opts out to a CSR, it can reduce the talk time as the CSR would not have to ask the customer to repeat the information he has already entered into the ARS. 14 The study also found that companies which were involved in live ARS monitoring achieved a high degree of success with the ARS by analyzing the parts of the script where customers were opting out and then refining the script to make it more pointed. The study discovered that companies that had low Opt Out Rates and high Customer Satisfaction achieved this through: 1) Clear definition of market and customer needs; 2) Clear and pointed IVR scripting; 3) Logical structure of menus and options; and 4) Educating customers on the use of the IVR system. 2.1.4 Theory of Reasoned Action and its Variants There have been numerous studies conducted in Information Systems (IS) research literature on predicting and explaining the utilization of information technology by end-users [24]. Most of these studies have been based on the Theory of Reasoned Action (TRA) [25] and its variants, such as the \"technology acceptance model\" [26] and the Theory of Planned Behaviour [27]. The Theory of Reasoned Action is a general theory applicable to an array of behaviours, including the forces which influence the use of IT. It specifies the relationships between beliefs, attitudes, and behaviours. TRA is based on the proposition that an individual's overt behaviour is determined by the individual's intention to perform that behaviour, which is a function of his attitude towards performing the behaviour itself and his subjective norm. It further postulates that an individual's attitude towards the behaviour is determined by his belief that performing the behaviour will lead to certain outcomes. In a variant of TRA, called the Technology Acceptance Model (TAM), Davis [26] identified two general constructs, perceived ease of use and perceived usefulness of the 15 system as the beliefs hypothesized to determine attitude. Ease of use was defined as the degree to which a system is perceived as easy to use by an individual. The Theory of Reasoned Action is important to this study as based on this theory the constructs 'ease of use' and 'ease of understanding' were selected to design and test the new IVR scripts. As has been stated before, the criteria for developing an effective IVR system for WCB was to test its 'usability' by the actual users of the system. If it is perceived as 'usable' by the individual(s) it would affect his attitude towards its use and hence the actual usage as postulated by the TRA. Expressed in an equation form Usability = Intent to use (Attitude) + Actual Usage Hence, usability was tested on the student questionnaire as Usable -> Use -> Expected Benefits Arrows above indicate \"leads to\". 2.2 Implications for Current Research The research on the use of phone-based user interfaces provides valuable input for the current study. The literature review points towards the need for a thorough understanding of customers and business processes and identification of the nature and pattern of calls received by the call centre of a company. Several months were spent at W C B in order to gain an understanding of the nature and pattern of call types and the business processes. The design guidelines for IVR scripting compiled from the literature review and the best practices in the industry was utilized in formulating the most appropriate script for WCB. 16 CHAPTER 3. UNDERSTANDING THE NATURE AND TYPE OF CALLS AT THE WORKERS' COMPENSATION BOARD 3.1 Introduction The first step in designing a good IVR system is to identify the nature and pattern of calls coming into the call centre. The literature review suggested that only after gaining a deeper understanding of the call flow process the calls that are suitable for an IVR system can be pinpointed. It is essential to determine the type of calls received by the call-centre, the time spent in answering the different types of calls, the length of wrap-up associated with certain calls, and the length of pauses between answering the different kinds of calls. 3.2 Pre-Test In order to identify the type of calls coming into the call centre at WCB a representative sample of calls was tracked (approximately 60 in number) by listening in to the phone conversations and an initial classification scheme for categorizing the various types of calls was then developed. It was ascertained that the calls received by the call centre come from the claimant, doctor or employer. The nature of calls received by the call centre could be categorized as either: 1. claim or cheque inquiry, 2. inquiry about a fax number, or 3. claimants requesting a transfer. The transfer category was further sub-divided into three categories: a. clients requesting a transfer and the call ending in a transfer, b. clients requesting a transfer and the call not ending in a transfer, and finally c. the clients not requesting a transfer and the calls ending in a transfer. 17 Calls, which could not be categorized into any of the above categories, were classified as general. A preliminary survey was distributed to five CSRs to confirm these categories. Based on the feedback received from the CSRs the survey was refined to include one additional category, the doctor's billing. The survey was then rolled out to all the CSRs. Data was collected over a week by all the CSRs on approximately 5,000 calls to get a preliminary sense of the proportion of calls that fall under each category. The survey confirmed these categories. Aggregated results from the survey are presented below: Type of call Percentage Claim Inquiry 32% Cheque Inquiry 16% Wage Loss 3% Doctor's Billing 3% Not End in Transfer 7% End in Transfer 18% Not Want Transfer End in Transfer 6% General 15% Table 1. Breakdown of different types of calls Based on the results of the pre-test the WCB project team established that a suitable research instrument had to be designed in order to track the CSR's activity and record the various types of calls. This was necessary because extensive data on calls would be required in order to develop an effective IVR system for WCB, and it was unreasonable to expect CSRs to collect detailed information on calls for this purpose. Therefore, the WCB project team developed a real-time electronic tool in MS Access to track the calls coming into the call centre. The tool captures information pertaining to: 1. who is calling the call centre, 2. what type of calls are received, 3. how many of those calls get transferred, 18 4. the time taken to answer various types of calls, 5. the time spent on wrap-up, 6. the length of pauses between calls, and 7. the time spent on various activities by the CSRs during their phone-times, including making outgoing calls and coffee breaks. The tool was pretested for 3 hours by two research analysts by tapping into the phone calls at the call-centre. Based on the pretest and the feedback received from the CSRs the tool was further refined. A fourth category 'other' was added to the \"Who is Calling?\" section. To the \"Transfer To\" section two new categories were added: 'Media (including forms and posters)' and 'Disability'. 'Other' was added to each of \"Calling Who?\" and \"What type?\" in the Outgoing calls section. 'Log entry postponed' was dropped from the wrap-up category as it was learnt from the CSRs that they almost never postpone their log entry, but complete it right after the call. A screen-shot of the main user interface along with definitions of important terms used is presented below: 19 * N COE Data Collector [frm_Calls] S Fie g t t Vjew Insert Format Records Tools Window Help -1*1*1 a t i Si ^ 1 M i B -si - 51 Incoming Calls f l Claimant JJj Doctor El Employer J3 Other L~j Claim Inquiry Cheque Inquiry Doctor's Billing 3 Fax Inquiry E\"J General H Transfer To Record: '< 1 I Form View Figure 1. Screen-shot of user interface for the data collection tool at the call-centre Important definitions a. Pause: this includes the time when CSRs are plugged in and waiting for an incoming call. During this time they are working on claims or other work related tasks. b. OtherPause: this includes non-work related activities. It is further subdivided into coffee breaks and other activities such as washroom or other breaks. c. Transfer To: i f the primary purpose of the call was a request for a transfer. Notice however, the call may not be transferred. In that case the option 'nowhere' would be checked under \"Transfer to\" category. Also, there might be instances when the primary purpose of the call was a claim inquiry but it ended in a transfer to some other department. In that case appropriate options would be ticked under the respective categories. 20 3.3 Data Collection at the Call-Centre Data was collected over 3 weeks for approximately 65 hours by three research analysts at the Richmond office. Each research analyst was equipped with a laptop and filled the electronic form while listening to the phone calls with the CSRs. At two of the area offices: Kelowna and Victoria, approximately 8 hours of data was collected. The purpose was to compare and contrast the activities of the head office with the area offices. At the end of the data-gathering period the WCB project team spent one week analyzing the data. As would be discussed in the results section a striking observation was that a high proportion of calls was related to transfers and routine calls like cheque inquiries. This led to the presentation of a proposal to WCB to explore the use of a well-designed IVR system to handle some of these calls. At this time, it was also ascertained that data collection was required in other areas of the organization like Entitlement, Case Management with which the call centre interacts in order to fully understand the call centre operations. Such an analysis will shed light on how some of those operations could be streamlined through the use of a well-designed IVR system. 3.4 Data Collection at Entitlement and Case Management There are two other units of WCB that receive calls related to claims. These are the Entitlement and Case Management units. They receive the calls either directly or as transfers from the call-centre. Several interviews with Entitlement officers and Case Managers were undertaken in order to get a sense of the type and nature of calls that these units get. Based on the interviews and the understanding that had been gained so far through data collection at the call centre, it was ascertained that for the incoming calls to these units it was important to know the call source, and who was calling. The call 21 source was particularly important as that would reveal how many calls get transferred from the call centre. For outgoing calls, it was enough to know who was being called. With this knowledge, a preliminary paper based form was developed which requested information on call source, the type of caller and the nature of incoming calls. For the outgoing calls, the form measured whom the call was going to and the purpose of the call. The paper-based form was refined based on feedback received from the entitlement officers and the purpose of the call was dropped from the Outgoing calls category as it was mostly related to claims. A copy of the paper-based form is presented in Appendix I. A computerized version of the paper-based form was developed by the WCB project team similar to the one developed earlier for call centre, which permits real-time data collection. The computerized tool measured time spent on various types of calls and other activities. On the electronic form the call source was divided into: 1) Direct-in-Dial (DID) 2) Transfer from the call centre, and 3) Other The various categories of people that could call these units were categorized as Claimant, Employer, Health Care Professional, Internal, Other. Same categories were used for the outgoing calls. A screenshot of the main interface is presented below: 22 Outgoing Calls J| Claimant \u00E2\u0080\u00A2 Health Care Prof. .ll Employer J| Internal M Other ^ COE D a t a Co l l ec to r [fim_Calls] | i El Fie \u00C2\u00A3 II.4.2 Average length of transfers 0:03:36 0:02:53 E E 0:02:10 0:01:26 0:00:43 0:00:00 c ZJ o o o < c o CO o < JO CO CD CL < c CD E CO CO CD CO CO < CO CO CO b c CD E a> LU co \u00E2\u0080\u00A2o CD CD Type of transfer Figure II. 9 Average length of Transfers Note that the above graph is only for calls that specifically asked for a transfer. It shows that the average length of time spent on calls before they are transferred. 1. time spent on transferring calls to Adjudication or Entitlement is higher than the time spent on calls transferred to the other departments at WCB 2. time spent on category \"nowhere\" is the highest as it denotes that the call which initially asked for a transfer was closed successfully by the CSR. 3. minimum time spent on a transfer call is 1 minute and 24 seconds. 80 II.5 Wrap-up Analysis II.5.1 Average Service Time (Call Handle + Wrap-up Time) 0:03:36 Claim Cheque Doctor's Fax General Transfer Inquiry Inquiry Billing Inquiry To C a l l Type Figure 11.10 Analysis of 'Wrap-up' 1. four call types: Claim Inquiry, Cheque Inquiry, Doctor's Billing, and General have average service times in excess of two and a half minutes. 2. the two claim types with the longest wrap-up times are Claim Inquiry and General 3. wrap-up times are disproportionate for each call type. 81 11.5.2 Distribution of time spent on each of the two components of service time Claim Cheque Doctor's Fax General Transfer Inquiry Inquiry Billing Inquiry To Call Type Figure 11.11 Service and Wrap-up times 1. the largest proportion of wrap-up time is for claim inquiry and general, an average of 15% of total service time. 2. the smallest proportion of wrap-up time is for fax inquiries, an average of 5% of total service time. 82 II.5.3 Average Wrap-up time. 0.00:35 000:30 + 000:26 + fi 0.0022 -s \u00C2\u00A3 000:17 -00009 -000:04 -000:00 -I Claim Inquiry Cheque Inquiry Doctor's Fax Inquiry General Transfer To Billing Call Type Figure II. 12 Average Wrap-up time 1. the average time spent wrapping up claim inquiry is greater than 30 seconds, the highest of all the call types 2. Wrap-up time for calls that request transfer occurs when the CSR responds to the request without transferring the call II.6 Calls suitable for IVR A detailed analysis was then undertaken to categorize calls that would be suitable for handling by an IVR system. Calls where the research assistants had taken extensive comments were identified out of the sample of calls on which data was collected. These calls were then categorized based on the comments that were recorded by the research assistants. It was then ascertained whether these types of calls could potentially be successfully handled by an IVR system. Such an analysis was helpful in the development of the actual wording of the script which is described in the next chapter. The following table depicts this categorization: 83 TOTAL 383 %of Total IVR? COMMENTS CLAIMENT CLAIMS Status 48 12.5% Y / N Queries about status of claim. No remarks regarding actual status. A claim might be accepted and then IVR could replace CSR, it might have been pending and then CSR is needed. Information Missing 7 1.8% Y Queries about status of claim. The problem was missing information. Other 44 11.5% N Queries about claim, for instance, reopening, etc. Must talk to CSR. Claim # 16 4.2% Y Called to find out what their claim number is. Pan 1 0.3% N Called to find out P A N . PAYMENTS Inquiry 13 3.4% Y / N Payment queries. No remarks regarding actual status of payment. A cheque might be in the mail so IVR could replace CSR, i f not, CSR is needed Out 6 1.6% Y Cheque is in the mail. Information Missing 2 0.5% N Problem with payment. Missing information. Must talk to a CSR. Other 15 3.9% N Queries about payment. Must talk to CSR for various reasons. EMPLOYER 26 6.8% N A l l calls that needed to be answered by CSR.* DOCTORS PAYMENTS 4 1.0% Y Was the cheque mailed to me? When? OTHER 11 2.9% N A l l calls that needed to be answered by CSR. TRANSFERS Asked for 61 15.9% N \"May 1 speak with Nancy/My case management...\" Don't have Ext. # Didn't ask for 20 5.2% N Cases that needed to be transferred to case manager/Adjudicator Voice Mail 6 1.6% Y When the call was only to give information that could have been left on a voice mail. Actual transfers to voice mail are included in transfers (they only happened because no one answered the phone) OTHER Fax# 20 5.2% Y Also includes information about phone # for other departments. Forms 15 3.9% Y Fax-on-Demand Follow-up 12 3.1% Y \"Did you receive the fax that I just sent?\" General Info.-IVR 19 5.0% Y Address, Hours, How to open a new claim.... General Info.-Other 37 9.7% N \"Who is my case manager?\", \"Is my adjudicator on vacation?\", \"How to fill a certain form\",... \u00E2\u0080\u00A2 There were few calls from Employers that could be suitable for automation and those have been incorporated under category Claimant above. \u00E2\u0080\u00A2 At least 21.5% of calls could have been dealt with by the proposed IVR. \u00E2\u0080\u00A2 Maximum Number of calls that could have been dealt with by IVR is approximately 44%. \u00E2\u0080\u00A2 9% of calls require a Fax-on Demand system. A n additional 3% (Follow-up) might also be handled by the system. ' Table II. 1 Calls suitable for an IVR 84 The above table groups calls coming from Claimants, Doctors, and Employers based on the purpose of the calls. Transferred calls are identified separately depending on whether the caller asked for a transfer or the call ending in a transfer even if the caller had not specifically asked for it. Most of the 'other' type of calls can potentially be handled an IVR as they are requests for information. The second column in the table depicts the total number of calls in each category and the next column shows the proportion of such calls out of the sample of calls that were identified for such an analysis. The fourth column shows whether such category of calls will be suitable for automation or not and the last column contains brief comments on each type. 85 Appendix III Discussion of the results of data collection at Entitlement Incoming calls A total of 78 calls were monitored while sitting with the entitlement officers. From the graph below it can be seen that: Incoming calls by categories of people Number of incoming calls Claimant Health Employer Internal Other Care Categories of people Figure III. 1 Incoming calls by categories of people 1. 53% (41) of the calls received by the Entitlement officers are from claimants 2. Calls from Employers, Health Care Professionals, internal departments, and others are equally divided and range anywhere from 10 to 15% each. 86 Breakdown of incoming calls by source: Breakdown of incoming calls by source 14% 3% \u00E2\u0080\u00A2 DID \u00E2\u0080\u00A2 Transfers from CC \u00E2\u0080\u00A2 Other 83% Figure III. 2 Breakdown of incoming calls by source The above graph shows that: 1. 83% of the calls come on the DIDs 2. 14% of the calls received by the Entitlement officers are transfers from the call centre 3. 3% of the calls come from other sources such as internal, etc. 87 Average time spent on calls with various people Average time spent on calls with various people Time 0:00:00 Claimant \u00E2\u0080\u00A2 Health Care \u00E2\u0080\u00A2 Employer \u00E2\u0080\u00A2 Internal \u00E2\u0080\u00A2 Other Categories of people Figure III. 3 Average time spent on calls with various people The graph above shows that: 1. The longest time is spent with the claimants on the phone on an average followed by Employers and Health Care Professionals Average time spent on different type of calls with different people: A v e r a g e t i m e s p e n t o n d i f f e r e n t t y p e o f c a l l s w i t h d i f f e r e n t p e o p l e 0 :07 :1 2 0 : 0 5 : 4 6 0 :04 :1 9 T i m e 0 : 0 2 : 5 3 0:01 :26 0 : 0 0 : 0 0 iSf 0 the r C la im Inq uiry N a tu re o f c a l is IC la im Inq uiry l O t h e r C a t e g o r ie s o f p e o p le Figure III.4 Average time spent on different types of calls with different people 88 The above graph reveals that: 1. The longest time is spent with claimants on claim inquiry followed by 'other' calls from claimants 2. There were no calls related to cheque inquiries in the sample of calls that were monitored. I n c o m i n g calls Incoming calls by categories of people N u m b e r o f ca l ls Claimant Health Employer Internal Other Care C a t e g o r i e s o f p e o p l e Figure III. 5 Incoming calls by categories of people 1. 24 (52%) of the incoming calls that were monitored are from claimants 2. 9 (20%) and 6 (13%) of the incoming calls that were monitored are from Health Care Professionals and Employers respectively 3. 5(11%) are internal calls 89 Breakdown of incoming calls by source Breakdown of incoming calls by Source \u00E2\u0080\u00A2 DID \u00E2\u0080\u00A2 T s f f r o m C C \u00E2\u0080\u00A2 O t h e r 86% Figure III. 6 Incoming calls by source The above graph shows that 1. 86% of the calls come in the DIDs 2. 7% each are transfers from call centres and other sources respectively. 90 Average time spent on calls with various people Average time spent on calls with various people Figure III. 7 Average time spent on calls with various categories of people 1. The average time spent on calls with Employers is 4 minutes and 42 seconds 2. The average time spent on calls with Health Care Professionals approximately 4 minutes and with claimants is approximately 3 minutes. 91 Average time spent on different type of calls with different people: Average time spent on various type of calls with different people Time 0 : 1 4 : 2 4 0 : 1 1 : 3 1 0 : 0 8 : 3 8 0 : 0 5 : 4 6 0 : 0 2 : 5 3 0 : 0 0 : 0 0 0> .\u00E2\u0080\u0094' O t h e r r C h e q u e Inqu i r y C l a i m Inqu i r y \u00E2\u0080\u00A2 Claim Inquiry \u00E2\u0080\u00A2 Cheque Inquiry \u00E2\u0080\u00A2 Other E LU -t\u00E2\u0080\u0094' O Type of call Categories of people Figure III. 8 Average time spent on various types of calls with different people It can be seen in the above graph that 1. The average time spent with an Employer on a claim inquiry is the highest, 4 minutes and 42 seconds, followed by an average of approximately 4 minutes on a claim inquiry from a claimant. 2. The data from 'other' type of call from 'other' is skewed as there were only 2 observations and one observation was for over 20 minutes. 92 SCRIPT 1 Welcome to the Workers' Compensation Board. Please choose from one of the following options. If you have an inquiry regarding a claim press 1 now If you are calling about payment information press 2 now If you know the four digit extension number of the person you wish to speak to press 3 now If you need general information or forms press 4 now 93 1 If you are inquiring about the status of your claim press 1 now If you require more information on your claim you may press 0 at any time to speak to a Client Service Representative 1st Timeout I Please enter your 8 digit claim number now You did not enter a response. Please try again. Enter the 8 digits from your claim number now 2 n d Timeout Less or more digits or invalid characters This entry is invalid. Your claim number must have exactly eight digits You did not enter a response. Please try again. Enter the 8 digits from your claim number now Claim number xxxx. If this is correct press 1 now. To enter the number again press the star key Please standby while we retrieve your claim information The claim number that you entered does not match our records Not found Please stay on the line and your call will be answered by a Client Service Representative Found Claim information is current to Y Y : M M : D D We require more information to process this claim P,S,I,B Info only The claim has been registered but we need the following documents to process this claim: 1 Doctor's Report 2 Employer's report of injury 3 Worker's form 4 Any other A decision will be made as soon as these documents are received Disallowed/Rejected Health Care only Accepte Interim Adjud The claim has been disallowed for the following reasons: 1 2 The claim has been accepted for health care purposes only as there has been no time-loss on the claim To go back to the main menu press the star key To speak to a CSR press o or simply stay on the line If you have no further questions you may hang-up now 94 2 If you are an injured worker press 1 now If you are a Doctor press 2 now Please enter your Personal Access Number Please enter the claim number P A N xxxx. If this is correct press 1 now. To enter the number again press the star key Claim number xxxx. If this is correct press 1 now. To enter the number again press the star key To CSR The personal access number that you entered does not match our records. Please stay on the line and your call will be answered by a client service representative Not found Not found Found The claim number that you entered does not match our records. Please stay on the line and your call will be answered by a client service representative To CSR To CSR Cheque issue date is older than 30 days. Please wait and your call will be answered by a client service representative Payee1 The most recent cheque was processed on Y Y : M M : D D in the amount of $x. It was mailed on the following business day No payment has yet been issued on this claim For Worker Yes Yes The most recent cheque was processed for your employer on Y Y : M M : D D in the amount of $x. It was mailed out on the following business day 95 For Doctor Payee 3 The most recent cheque was processed on Y Y : M M : D D in the amount of $x. It was mailed on the following business day To go back to the main menu press the star key; To speak to a CSR press o or simply stay on the line; If you have no further questions you may hang-up now Please enter the four digit extension number now If you wish to speak to the person at this extension press 1 now If you wish to leave a voice mail message for this person press 2 now Please leave a voice mail message after the beep. When you are finished you may hang-up Yes Direct Connection Play voice mail message 96 If you require a form to be faxed to you press 1 now If you are a first time caller press 2 now If you require general information and information on hours of operation press 3 now If you are an employer press 1 now If you are a doctor press 2 now If you are an injured worker press 3 now r Please enter your fax number beginning with the area code and then hang-up r Fax either form 6, 7, or 8 based on the input choice If you are an injured worker you need to send us form 6 If you are a doctor you need to send us form 7 If you are an employer please mail or fax form 8 The Call Centre is open from 7 am to 7 pm Monday through Friday. You may use our automated inquiry system for claims and payment information at any time. To go back to the main menu press the star key To speak to a CSR press 0 or simply stay on the line If you have no further questions you may hang-up now 97 S C R I P T 2 Welcome to the Workers' Compensation Board. Please choose from one of the following options. If you are an injured worker press 1 now If you are a doctor press 2 now If you are an employer press 3 now If you have an inquiry regarding a claim press 1 now If you are calling about payment information press 2 now If you know the four-digit extension number of the person you wish to speak to press 3 now If you need general information or forms press 4 now If you have an inquiry regarding a claim press 1 now If you are calling about payment information press 2 now If you know the four-digit extension number of the person you wish to speak to press 3 now If you need general information or forms press 4 now Proceeds in the same way as the previous Script If you have an inquiry regarding a claim press 1 now If you know the four-digit extension number of the person you wish to speak to press 3 now If you need general information or forms press 4 now 98 Appendix V IVR Simulator By Arash Ramin IVR SIMULATOR 99 OVERVIEW 100 I N T R O D U C T I O N : 1 0 0 M E N U I T E M S 101 CONFIGURATION 102 O U T P U T 1 0 3 I N P U T 1 0 4 E R R O R H A N D L I N G 1 0 5 CODE STRUCTURE 106 T O O L S 1 0 6 C L A S S E S 1 0 6 KNOWN ISSUES 107 L I M I T A T I O N S 1 0 7 B U G S 1 0 7 CONTACT INFORMATION 107 99 Overview Introduction The IVR Simulator is capable of supporting multiple simulations, each with completely different configurations. Each simulation's parameters and states are stored in a database. The simulator is a state-based system, meaning that while running the system only is only aware of the current state. Each simulation starts off in the default 'START' state, and will make the appropriate transitions from one state to the next until the ' E N D ' state is reached. Each state consists of output (sound files), input (user touch-tone entries), and error-handling (timeouts or invalid responses) modes. When the system enters a new state, it plays any specified output sounds and then accepts input as specified by the current state. Based on that input, it can then make a transition to the next state, until finally the ' E N D ' state is reached, which marks the end of the simulation. If no input (or invalid input) is made, then the system enters the timeout states as specified by the error handling parameters of the corresponding state. Figure 1 shows the end-user interface for the IVR Simulator. The status bar at the bottom shows the current status of the simulation (the left indicates which simulation is currently open, and the right shows the state of the simulation itself - STOPPED, PAUSED, or RUNNING) When a simulation is in the running state, user input may be made by either clicking on the numeric buttons with the mouse, or by using the keyboard's keypad. I IVR S imu la to r File Simulation Help 'WCB Script 81.DAT r r r a 1 J8,' 3 4 5 6 7 8 9 * J I ' 'STOPPED Figure V. 1 IVR Simulator user interface 100 Menu items A l l menu items are context-sensitive, meaning that they appear enabled when they are valid options or disabled (greyed out) when invalid. For example, the SIMULATION -> STOP menu item is only valid when a simulation is currently running. Figure 2 shows the File Menu, which allows you to open/close existing simulations, or create new simulations altogether. ItS, IVR Simulator | File Simulation Help New Simulation Open Simulation Close Simulation Exit Figure V.2 File Menu Figure 3 shows the Simulation Menu. This allows the operator to start, stop or pause a configured simulation, or to change the configuration of the simulation. mm Help i Simulation Start Ctrl+S Configure Drl+C Figure V.3 Simulation Menu 1 0 1 Configuration Figure 4 shows the simulation's configuration window. The left pane shows the simulation's different states, and the right pane allows you to configure the output, input and error handling modes of each state. Each simulation has two default states - a START state, and an OPERATOR state. Simulation Configuration { O P E R A T O R Name Description [START j Default start state Output | Input ] Error Handling | Output 1 j NONE C BUFFER C SOUND Output 2 C NONE r BUFFER C SOUND Output 3 5| Input | Error Handling | - Output 1 1 C NONE r BUFFER SOUND : l 1 w a v \" J L J r Output 2 | | "Thesis/Dissertation"@en . "1999-05"@en . "10.14288/1.0089021"@en . "eng"@en . "Business Administration - Marketing"@en . "Vancouver : University of British Columbia Library"@en . "University of British Columbia"@en . "For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use."@en . "Graduate"@en . "Developing and testing an effective interactive voice response (IVR) system for the Workers\u00E2\u0080\u0099 Compensation Board of British Columbia"@en . "Text"@en . "http://hdl.handle.net/2429/9198"@en .