Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Milking a virtual cow? : institutional and consumptive implications of online grocery shopping Murphy, Andrew 2003

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

Item Metadata

Download

Media
831-ubc_2003-860221.pdf [ 47.85MB ]
Metadata
JSON: 831-1.0091233.json
JSON-LD: 831-1.0091233-ld.json
RDF/XML (Pretty): 831-1.0091233-rdf.xml
RDF/JSON: 831-1.0091233-rdf.json
Turtle: 831-1.0091233-turtle.txt
N-Triples: 831-1.0091233-rdf-ntriples.txt
Original Record: 831-1.0091233-source.json
Full Text
831-1.0091233-fulltext.txt
Citation
831-1.0091233.ris

Full Text

Milking a Virtual Cow? Institutional and Consumptive Implications of Online Grocery Shopping by Andrew John Murphy B A , University of Auckland, 1990 B C o m , University of Auckland, 1990 M C o m , University o f Auckland, 1995  A thesis submitted in partial fulfilment of the requirements for the degree o f  Doctor of Philosophy in The Faculty of Graduate Studies (Department of Geography)  We accept this thesis as conforming to the required standard  The University o f British Columbia October 2003 © Andrew John Murphy, 2003  In  presenting this  degree at the  thesis in  partial  University of  fulfilment  of  of  department  this thesis for or  by  his  or  scholarly purposes may be her  representatives.  permission.  of  C^~O0 * A  W  The University of British Columbia Vancouver, Canada  Date  DE-6 (2/88)  °i  C<V~(\<r<  2CTo  for  an advanced  Library shall make it  agree that permission for extensive  It  publication of this thesis for financial gain shall not  Department  requirements  British Columbia, I agree that the  freely available for reference and study. I further copying  the  is  granted  by the  understood  that  head of copying  my or  be allowed without my written  Abstract W(h)ither online grocery? In this investigation of electronically-mediated food retailing and consumption I seek to determine why and how firms and individuals become involved in ecommerce projects, and how these projects become instituted and embedded. I re-conceptualise what constitutes the firm, using a stakeholder framework to redefine the social relations within and beyond it: to portray the firm as fractured, porous and conflictual. I use qualitatively rich and 'thick' material from in-depth interviews with stakeholder-actors (managers, employees, customers and suppliers of three "e-tail" companies in Vancouver and Auckland, N e w Zealand) to provide multiple perspectives on what it means to transact for food through the internet. From these multiple perspectives a holistic interpretation of what firms are, how they operate, and for what and for whom they function is constructed. The case study firms, and the individuals who work for them and who transact with them, have a variety o f motives for their actions: technological curiosity, cost-minimisation, convenience, diversification, and a moral imperative to "build a better world through business". Electronicmediation is a. key innovation in reducing the costs of interaction between retailers and customers, and provides more targeted information about customers and their behaviour. I argue that engaging in e-commerce requires a redefinition of what a retailer does, and a re-engineering of how they operate. E-commerce requires considerable investment in trucks, warehouse or backroom space, call centres, and in branding. E-tailers seek to minimise logistical costs, and yet maximise flexibility for customers. E-tailers therefore balance competing economic and sociocultural objectives: to be the cheapest, biggest, most flexible and trustworthy food retailer. I also argue that online grocery customers are in a vanguard of e-commerce converts: committed to virtual shopping, they are large spenders, early adopters of broader-band internet access technologies, and use the mundane act of shopping for food as a springboard to other online consumptive practices.  ,  While a number of high-profile online grocers have ceased operations, others continue to expand, locally, nationally and globally. This thesis provides case study material in the processes o f innovation, institutionalization, and in the transformation of consumption, as seen through the multiperspectivist stakeholder approach.  Table of Contents Abstract  ii  List of Tables  vii  List of Figures  viii  Acknowledgements  xi  1  1  Introduction 1.1  2  O n (Revolutions and revelations: an Idea and a Thesis Or, a personal anecdote o f the  whys and wherefores of research  1  1.2  Internet and E-Commerce Growth: Feeding and Defying the Hype  9  1.3  Situating the Thesis  21  1.3.1  Retailing  22  1.3.2  Consumption  23  1.3.3  Information Technologies, Cyberspace and Culture  25  1.3.4  E-Commerce and Online Grocery  27  1.3.5  Literature Summary  28  1.4  Research Objectives and Questions  29  1.5  Structure of the Thesis  31  Online Food Retailing: Research Context  35  2.1  Introduction  35  2.2  Historical and international context  35  2.3  Case Studies  45  2.3.1  Justification for Case Studies Approach  45  2.3.2  Case Study Options  48  2.3.3  Case Study Company Histories  52  2.3.4  Case Studies Summary  61  2.4  2.5  Interviews and Data Collection: Details, details  62  2.4.1  Woolworths  63  2.4.2  Quick  65  2.4.3  SPUD  66  2.4.4  Interview and Data Collection Summary  68  Reflections on the Research Process  69  2.5.1  73  Positioning the Researcher iii  <-  2.6 3  73  Enterprising Stories: Conceptualising the Retail Firm  75  3.1  Introduction  75  3.2  Conceptualising the Firm  76  3.2.1  Towards a 'Relational Turn'  78  3.2.2  (Pre-)Conceptions and Extensions  82  3.3  Re-conceptualising the firm: A Stakeholder Approach  83  3.4  Stakeholder Research in Action  90  3.4.1  93  3.5 4  Summary  Stakeholders, Relations and Perspectives  Summary........  ,  97  Managing Innovative Change  99  4.1  Introduction  4.2  Innovations in Retailing and its Representations  100  4.3  Innovation Management in Theory: Belief, Culture, Practice  103  4.3.1  The Innovation Imperative  104  4.3.2  Innovation and Corporate Culture  106  4.3.3  Getting Innovative  108  4.3.4  Disruptive Technologies/Innovations  110  4.3.5  Innovation Theory: a summary  113  4.4  4.5  4.6  99  Booting Up: the Motives and Mechanics for Innovation  114  4.4.1  Woolworths  115  4.4.2  SPUD  121  4.4.3  Webvan  124  4.4.4  Quick  128  4.4.5  Getting started: a review of motive, capabilities and passion  132  .".  Upgrading: keeping up and moving on  134  4.5.1  Woolworths  134  4.5.2  SPUD  139  4.5.3  Quick  149  4.5.4  Growth and Consolidation: managing change  '.  151  Crashing on the information superhighway  152  4.6.1  153  Webvan: the Hindenburg of Grocers iv  4.6.2 4.7 5  155 158 161  5.1  Introduction  161  5.2  Theoretical Representations  162  5.2.1  Representing Customers in Retail Geographies  162  5.2.2  Representing Customers in Cyberspace  166  5.2.3  O n Metaphor: constructing and using virtual realities  168  5.4  5.5  7  Conclusions: C h ... ch ... Changes  Representing and Using Virtual Retail Spaces  5.3  6  Quick: model proven, but not fast enough  Profiling Online Grocery Customers: W h o are they?  172  5.3.1  Surveys  172  5.3.2  Interview Profiles  177  Adopting and Using Virtual Retail Spaces  180  5.4.1  Getting Started: Booting up  181  5.4.2  Getting Going: Shopping  186  5.4.3  Getting Around: Website Design and Metaphor  189  5.4.4  Customer Agency: responding to feedback  194  5.4.5  Customer Relations with People and Products  200  Conclusions and Discussion  203  Re-solving Space and Time: Linking Virtual Retail with Real Consumption Spaces  206  6.1  Introduction  207  6.2  Spacings and Timings: A review of retail logistics  210  6.3  Logistics in Online Grocery: the Model Contenders  213  6.3.1  217  The Infomediary and Third Party Logistics  6.4  Re-solving Internal Space and Time  219  6.5  Re-solving External Space and Time  228  6.6  Conclusions  240  Conclusions  245  7.1  The Problem and Process Restated  245  7.2  The Problem Re-solved?  248  7.2.1  Why the internet? Why now?  248  7.2.2  Why groceries?  248  7.2.3  Why here?  249 v  7.2.4 7.3  7.4  So what?  Contributions Summary  250 ;  251  7.3.1  Theoretical Contributions  251  7.3.2  Methodological Contributions  252  7.3.3  Empirical Contributions  253  Future Directions  257  References  260  Appendix A : Interview Details  303  Appendix B: Interview Question Guide  305  Appendix C: S P U D Questionnaire, July 2001  308  vi  List of Tables Table 2.1: Selected North American Online Grocers  38  Table 2.2: Venture Capital Funding, Jan 2000  39  Table 2.3: Selected online grocer sales in 2000  41  Table 2.4: North American Online Grocery Market Predictions  43  Table 2.5: S P U D Acquisitions, 1998-2002  60  Table 2.6: Case Study Company Comparison  62  Table 3.1: Example Stakeholder Group Issues  84  Table 3.2: Stakeholder Inputs, Costs and Risks  86  Table 4.1: S P U D Customer Churn, 2001 to 2002  140  Table 4.2: S P U D Customer Acquisitions, 1998-2002  140  Table 4.3: S P U D Customer Acquisitions, 1998-99 and 2000-02  141  Table 5.1: Reasons for shopping with S P U D : Feb 2000  175  Table 5.2: Percent of Grocery Expenditure with S P U D : 2000 and 2001  175  Table 5.3: Reasons for expenditure elsewhere: Feb 2000  176  Table 5.4: Customer Interviewees: Demographics  177  Table 5.5: Customer Interviewees: Income and Household Composition  179  Table 5.6: S P U D Survey 2001: Education level  180  Table 5.7: Time spent Online and with Retailer  182  Table 5.8: U S online activities by income group, 2001  184  Table 5.9: Customer access to the internet  186  Table 6.1: Picking and Delivery Policies  222  Table A l : Woolworths Staff Interviews: timing and frequency  303  Table A 2 : Quick Staff Interviews: timing and frequency  303  Table A 3 : S P U D Staff Interviews: timing and frequency  304  vii  List of Figures Figure 1.1: Forecasts for Electronic Commerce in 1997  3  Figure 1.2: Likely Effects of Electronic Commerce: sectors at risk  4  Figure 1.3: Safeway banner advertising bank kiosk  6  Figure 1.4: Mentions of "electronic commerce" in business journals and newspapers, 1991-2002.. 8 Figure 1.5: U S Households with Computer and Internet Connections  10  Figure 1.6: Different Perspectives on Internet Access and Use, U S  10  Figure 1.7: Internet adoption in U S , U K , Canada and N Z households, 1997-2002  11  Figure 1.8: Diffusion of Technologies in the U S A : broadband vs cell phones  12  Figure 1.9: U S E-Commerce and Total Retail Sales, quarterly, 1999-2003  13  Figure 1.10: U S E-Commerce as percent of Total Retail Sales, 1999-2003 :  14  Figure 1.11: U S Stock prices versus earnings, 1881-2003  14  Figure 1.12: Online Population and Sales Forecasts, November 1999  16  Figure 1.13: Online Grocery Sales Forecasts and Reasons, September 1998  17  Figure 1.14: Transformative Power of the Internet: the ladder metaphor  18  Figure 1.15: I C T Adoption in Businesses  19  Figure 1.16: Dotcomfailures.com: November 2000  20  Figure 1.17: Dotcomfailures.com: May 2003  20  Figure 2.1: Sainsbury Home Delivery, 1913  36  Figure 2.2: Webvan Stock Value, $US  ,  40  Figure 2.3 : Stongs Home Page: Peachtree Markets (November 2001)  42  Figure 2.4: US Internet Use with Income  45  Figure 2.5: Stong's Tradition and Webstore Closure  50  Figure 2.6: Organics T o Y o u Image  51  Figure 2.7: Tesco sales, 1999-2003  53  Figure 2.8: Webvan sales and losses, 1999-2001  55  Figure 2.9: Woolworths Delivery Areas, 2000 and 2002  56  Figure 2.10: Quick Partners  58  Figure 3.1: Retail Stakeholders  84  Figure 3.2: Selected Woolworths Stakeholder Relations  90  viii  Figure 3.3: Selected S P U D Stakeholder Relations  92  Figure 3.4: Selected Quick Stakeholder Relations  93  Figure 4.1: Disruptive and Sustaining Innovations  Ill  Figure 4.2: Quick Homepage  129  Figure 4.3: Quick Truck and Riser  131  Figure 4.4: Foodtown and Woolworths webpages, June 2003  139  Figure 4.5: S P U D sales, 1999-2003  148  Figure 4.6: Quick Closing Sale email, February 2003  157  Figure 4.7: www.quick.com in March 2003  157  Figure 5.1: Internet User Psychographics  167  Figure 5.2: The Information Superhighway  171  Figure 5.3: Woolworths Print Advertisement: "Let Your Mouse Get the Beer"  171  Figure 5.4: Age  173  Figure 5.5: Household Composition  173  Figure 5.6: Household Income  174  Figure 5.7: Average Weekly Spend  174  Figure 5.8: Produce Expenditure  176  Figure 5.9: Other Grocery Expenditure  176  Figure 5.10: US internet use by education level, 1997-2001  180  Figure 5.11: US internet activities, 2000, 2001: % of total US population aged 3+  183  Figure 5.12: US internet activities, 2001: % of U S internet users aged 3+  184  Figure 5.13: Woolworths Website Design, March 2000  189  Figure 5.14: Woolworths Website Redesign, June 2001  190  Figure 5.15: Website Design: Tesco, August 2000  190  Figure 5.16: Quick "House" design metaphor, January 2001  192  Figure 5.17: Quick redesign 1: Q-Market, April 2001  193  Figure 5.18: Quick redesign 2: Q-Market, June 2001  194  Figure 5.19: Woolworths Design: customer suggestion, July 2000  197  Figure 5.20: Quick design: customer education, January 2001  198  Figure 5.21: S P U D open house, March 2001  203 ix  Figure 6.1: Store Pick, Tesco  214  Figure 6.2: Tesco HomeShopping Backroom  214  Figure 6.3: Tesco Delivery  215  Figure 6.4: Webvan Warehouse  216  Figure 6.5: E m p o r i Centre, Toronto  218  Figure 6.6: HomePort Box, London  218  Figure 6.7: Shop Pick: Woolworths Personal Shopper  221  Figure 6.8: Warehouse Pick: Quick  223  Figure 6.9: Woolworths Backroom  224  Figure 6.10: Supermarket campaign against local liquor licensing laws  227  Figure 6.11: Webvan delivery slots, San Francisco, 26 June 2001  230  Figure 6.12: S P U D delivery zones  232  Figure 6.13: Auckland Customer Deliveries, December 1999  236  Figure 6.14: Woolworths Delivery  237  Figure 6.15: Quick Delivery  237  Figure 6.16: S P U D Delivery  237  Figure 6.17: Vancouver customer locations, S P U D , July 2001  239  Figure 6.18: S P U D Bicycle Delivery  242  x  Acknowledgements Six years, a few more grey hairs and many nights of zombie-zone writing later, I can finally declare to myself, friends and family, and to the relief of my supervisors: it is done! A s with any piece of prolonged work there are a number of people who are fully responsible for what appears here, and without whom I might have had a better paid career outside of academia. First and foremost, I must thank my supervisors, Trevor Barnes and K e n Denike. They gave me employment when short of cash, a chair when in need of a chat, guidance when feeling bewildered, and above all unwavering mentoring. In the generosity o f their support I have a model to live up to in the supervision of my own students. M y supervisors have played a vital role in assisting my career development, both occupationally and in reviewing journal submissions, and I hope the experience of putting this thesis to bed has not foresworn them of such invaluable assistance in the future. I have some empathy with K e r i Hulme (1984), who states in her preface to the Booker Prizewinning novel The Bone People: T o me, it's a finished shape, so finished that I don't want to have anything to do with any alteration o f it. W h i c h is why I was going to embalm the whole thing in a block o f perspex when the first three publishers turned it down on the grounds, among others, that it was too large, too unwieldy, too different when compared with the normal shape o f the novel. ... T h e editor should have ensured a uniformity? Well, I was lucky with my editors, who respected how  I feel about ... oddities. . . . Maybe  the  editors were too  gende with my  experiments and  eccentricities. Great! T h e voice o f the writer won through. (Hulme, 1984, ix).  I thank my supervisoral committee for respecting and supporting my wish to structure this thesis, complete with its "oddities", the way it is. I cannot thank enough the many interviewees who so generously gave their time and trust. I wish to give particular thanks to the managers, employees and customers of Woolworths Supermarkets N Z , Quick, and Small Potatoes Urban Delivery, without whom this thesis would have been approximately 20,000 words shorter. While I cannot do justice to any more than a fraction o f the insights and information they afforded me, I hope they benefited from the research process and find some recognition and use in this representation of their corporate and personal worlds. I owe more thanks than I can express to my friends over the years at U B C and in Vancouver. T o Wendy, Maarten, Diane and Etienne: your friendship was and is the highlight of my time there. Y o u were the best office mates, and your hospitality has always been as warm as it is boisterous. Life moves on, but let true friendship never die. T o the Cooperative Auto Network go thanks for their support, financial and emotional. I wish you all success with your database, venture and concept. T o Bob and A l i : your words of wisdom, your approach to scholarship and life and the xi  knowledge that we are all in this final struggle together have been inspirational. May we share the stage together. The last two years have seen significant changes in my personal and occupational life. Moving to Southampton opened up a new world o f resources, scholarship and collegiality that has undoubtedly influenced the work that follows. I thank N e i l Wrigley for his belief in my abilities as a retail lecturer and scholar, and for giving me a valuable boost in my career. I made good friends in my year at Southampton, and value their interest in my work and their continued support, particularly Jiirgen Essletzbichler for his regular thrashings on the squash court. While the time spent writing and delivering lectures put my thesis on hold for a year, it was an experience I do not regret for an instant. Moving to Birmingham has extended these benefits further. M y colleagues have been most generous in allowing me to prioritise my thesis over everything else, and I am grateful to Mike Taylor and Judith Petts for this. The constant reminder o f the need to earn the " D r . " on my door has paid rewards. I am simultaneously saddened and pleased to see so many of the collegial friends I made in this last year move on to brighter and better things elsewhere, but at least I managed to finish this damn tiling before they did. Getting married to Jenny, and being so welcomed and supported by her family, has been an immensely rewarding development. Jenny has sacrificed beyond the call of duty, and I hope I can now find some way to redeem this unconditional love and support. I am sure she has plans in store. My family have been unwavering in their love and support, and in their insistence that I get this thesis done. M y father Brian Murphy in particular has been an amazing help in proofreading, always there to bounce ideas off, and ever generous in accommodating us in N e w Zealand during fieldwork or pleasure. This thesis was nearly abandoned before it even began, with the sudden death of my mother in October 1997. This is the type of event that tends to force a re-evaluation of life's priorities and goals, and at the time a P h D did not seem one of them for me. M y family practically had to force me on the plane back to Vancouver, and I am grateful that they did. The last six years would have been different had my mother survived, and the world is a quieter and less joyful place without her. She would have wanted it to continue nonetheless. She would have been proud of this thesis, whatever it contained.  xii  I dedicate it in her loving memory. Erina Mary Murphy 1941 - 1 9 9 7 K i a kaha. K i a toa. K i a manawanui. (Be strong. Be brave. Be o f great heart.)  xiii  1 Introduction It used to take days to send an important letter to your best friend. Now, with e-mail it takes seconds to send trivia to people you barely know. (MacFarlane, MacFarlane and Fiddy, 1997) The Internet changes everything. That was a common refrain just a very short time ago. Almost no one seems to believe it anymore. (Liebowitz, 2002: 1)  1.1 On (Revolutions and revelations: an Idea and a Thesis Or, a personal anecdote of the whys and wherefores of research A bite to eat... A middle-aged woman picks up a bunch of bananas, checks her shopping list, and chooses a slighdy more green bunch instead. She goes on to collect more vegetables and fruit, which she tests for appropriate firmness, then some lamb chops and fresh brie; and takes her full trolley to the checkout.  This scene is no normal chore, an everyday shop by a typicalgrocery store customer. It is instead ajob, for the wom  shopperfor Stongs Markets in Vancouver. Her shopping list is a printout of a customer's order, received earlier that m  an emailfrom Stria Communications in Montreal, the managers of the website on which the order was placed. The i  the list — the firmness and colour offruit — are as per the instructions the virtual customer noted at the time of o  customer could be anywhere in Vancouver — at work, or at home — anywhere, that is, within the delivery radi  Alternatively, the customer may-choose to pick up their ready-packed groceries; either way, for this personal service a made.  While Stong's online customer base is not large (as of March it numbered around 100, compared to the more estab  500 thatfaxes and phones orders through, using the same packers and deliverers), it is the hoped-for source of futur  supermarket? Caught between the expansion limitations of a single site, and the capital and managerial requir  establishing a second store, Stong's is hoping to increase its customer base and turnover by fulfilling remote order  necessary. Its two dedicated shoppers are [said to be] "very busy", but it may be some timeyet before the store c substantial] source of income.  The move toward the world-wide web as a method of transacting business is reflected in the business strategies o  grocery independents across Canada, trading either under the Stria Communications' Peach tree network, as Stong  IGA consortium in Quebec (expansion plans to the rest of Canada are unknown). In the United States there is mo  given to independent traders such as Peapod, who initially sent their dedicated shoppers into other existing superm  now establishing their own series offulfilment warehouses; and the widely publicised, high-tech and well-funded W  This vignette is presented as first written in mid-1999, for reasons that will be made clear. Stong's is no longer part of the Peachtree Network, and Peapod is now a subsidiary of Royal Ahold's US-based chains. Sobeys Quebec's Cybermarche I G A has not expanded beyond the province, but "independent traders" have become established elsewhere, as will be described later. Interview with Cori Bonina, manager of Stong's, March 23 1999. All information on Stong's in this anecdote is from this source.  1  by the eponymous founder of bookchain Borders {Anders, 1999) using conveyor belts instead of personal shopp  Zealand the only currently operating web-based ordering system is run in the three largest cities by Woolworths,  nation-wide supermarket chains, and a subsidiary of Hong Kong-based Dairy Farm International (the holding com  supermarket chains in Hong Kong, Taiwan, Singapore, Australia and Indonesia). Rather than using dedicated shopp  the regular store, Woolworths has constructed specialised backroom packaging areas in a selected supermarket in ea  contracted with a major courier company to deliver the packages in dedicated refrigerated vans using custom-design  Less than six months in operation, Woolworths claims to be on-track to have 10,000 customers by the end of theye ... and the means to pay for it The cursor moves to the O K button, and with a click the credit card bill is paid, funds transferred from a chequing account to Visa. A few clicks later, and the phone bill and month's rent are also paid, and all the day's online transactions transferred to a financial software package for later analysis and budgeting. The computer user prints out the Visa and phone bills in hard-copy form - while she saves a dollar per month by receiving an electronic invoice rather than a mailed copy, she didn't want to leave anything to chance with the millennium rollover only a few weeks away...  While the last part of this story is pure projection — no banking institution yet presents utility bills through the inter  customersyet offered the option of mail versus the web — it may only be a matter of time, and months rather than  does occur; and bill payment systems are a staple of online banking systems. Indeed, BankAmerica is conducting e  billpresentment in California, and CheckFree is an independent US organisation dedicated to free billpayment service  banking institution to any corporate payee. Several banks, including the ASB Bank ofAuckland, New Zealand, are te means ofpresenting bills to customers and integrating this with the already-existing billpayment services.  Meanwhile, every major bank in Canada (and quite a few of the credit unions) offer web-based banking, and adverti  material sent to customers, advertisements in the media, and notices on ATM machines. The reason is simple: usi  the bank money, compared to telephone, ATM and over-the-counter based alternatives. While the banks are  acknowledging plans to shift the population over to remote interaction methods, and thus close fully-staffed bra  transactions are charged at a lower rate, and it may only be a matter of time. Meanwhile, branch closures are a ma not only in Canada, but in New Zealand, the UK, and other nations also.  I wrote the above as the introduction to my thesis proposal in May 1999. I wanted to explore through the process o f doctoral research and writing how ideas — o f and by "entrepreneur" providers and "early adopter" users — become experiments, practices and institutions. The "idea" I was most interested i n at the time was that o f electronic commerce, as a potentially revolutionary force in altering the ways retailers and their customers do business. The germ o f this idea for me stretches back to my teenage years, and the agony o f being dragged around (what seemed at the time endless) shops looking for clothes, food, you name it, with my mother. I vividly recall looking at all the (ghastly) clothes on sale racks, and food spoiling on shelves, and tliinking to myself: there has to be a better way than this! Can't we as buyers tell the manufacturers what we want before the  2  goods are made or gathered, and stop this waste? Being 1984, there was at the time no such largescale mechanism to do so, beyond the idea of the invisible hand, to coordinate supply and demand, and the telephone and catalogue for small-scale operations. Fast forward to May 1997: having applied to the P h D programme at the University of British Columbia, I was backpacking in Egypt, perusing market stalls in downtown Cairo, whereupon I came across an Economist special survey on the business implications and uses o f the internet, entitled "In Search o f the Perfect Market". (The Economist, 1997). A m o n g articles o f business-to-business markets, and customer reluctance to use credit cards, was discussion o f where "the effects o f electronic commerce are being felt most keenly - for good or i l l " (The Economist, 1997: 11): financial services, sex, retail, music, books, cars, advertising and marketing. According to The Economist (1997: 3), 'the Internet, for all the miracles it has wrought in communications, is still a horrible way to buy a shirt. But it will get better, much better.' The survey o f the state o f electronic commerce contained promising figures of expected growth of business-to-business and business-to-consumer transactions (Figure 1.1) and the retail sectors thought to be the most affected (Figure 1.2). The Economist went on to allege that 'the more tiresome a purchase is in the physical world, the more likely consumers are to try an online alternative', and thus mortgage comparison sites ought to be popular, since this task is difficult on the High Street, while online C D sellers would have to offer something physical stores cannot in order to compete.  Figure 1.1: Forecasts for Electronic Commerce in 1997 Bound to take off .. Bu4.ni",-. to bun mc.: ibn  D Consumer I M — » For rover  /  — ' V*nk«Ml Group • — Cttuv-«.'«IMBA  /  /  Group  y  As wtt>  /  mm  Source: The Economist (1997: 2)  3  yr  Figure 1.2: Likely Effects of Electronic Commerce: sectors at risk .,___„.„—.—. _ —  Where onli n e s c o r e s 4  :  Q  * 1' u'I.'«, :•  may  Ugfit  p" |  • 1 i  """"  Hi f  *m |  tm  i nli.it>') iiy  »  M B W  %:'. , ( - M. Source: The Economist (1997: 7)  The  idea  that information technology could revolutionise the way businesses and customers relate  to each other was intriguing, not the least because of the surroundings in which I was reading the article. Cairo was seemingly devoid of retail spaces as I knew them, in particular enclosed and strip malls; most selling, as far as my limited tourist gaze and experience (mostly of food) could determine, occurred in open temporary markets with a total absence of recent technological advances (and inter-actional retreats) of point-of-sale scanners and so on. Could the internet really penetrate and revolutionise food retail in consumption worlds as diverse as Cairo, Auckland or Vancouver? Groceries certainly count as a "tiresome purchase", but would customers trust the selection o f their food and toiletries to a stranger - and i f so, how? This seemed to me to be an intriguing idea worth spending more than five years to explore, analyse, and dissertate. Hence the proposal. Ideas are not worth much i f they cannot be communicated and enacted. This thesis seeks to fulfil a number of purposes, but it primarily aims to communicate the ideas and 3  actions of a number o f actors involved in the use o f the internet as a virtual retail and consumption space for food, or online grocery retailing (or "e-tailing") as it shall be referred. These ideas and actions were recorded through a series o f empirical investigations, which are outlined in detail in Chapter 2. This thesis aims to reproduce (or re-present) these ideas and actions in terms that would be recognisable to the actors involved, and the simplest way to do that  I specify these objectives in a later section. 4  is to use their words as much as is practicable, interpreted and analysed as appropriate. Hence the vignette: a hypothetical teaser for the empirically-grounded stories that are to follow. In writing this thesis I wish to communicate some of the action, the hopes and fears associated with electronic commerce, and in particular with that most mundane of grouping of products, groceries. Groceries are an interesting group of products to study for a number of reasons: everybody buys them, so finding potential customers (online or offline) to interview ought not to be difficult. While shopping for cars, houses or holidays online might involve larger sticker prices than a tomato, these activities are typically conducted in irregular but concentrated periods of time with many possible providers. Online food retailing, on the other hand, is more likely to generate longer-term "relationships" between customers and the relatively few retailers, partly because the process is quickest i f lists of previous or planned purchases are able to be transferred from one transaction to the next. For many shoppers supermarkets have supplanted the local store and the market as the prime (or even only) source o f food, and I felt it would be interesting to compare the sociality of the experiences: what does it mean for the act of consumption i f the purchaser's only interaction with the store is through a website? The purchase of food is for many a personal activity, reflecting tastes and aspirations (or "love and sacrifice", i f Miller's (1998) Theory of Shopping is correct), and an activity customers can be surprisingly passionate about. The nexus of the reality (or grounded, atom-based nature) of food and the virtuality (amorphous, bit-based culture) of the internet was enticing. There is a considerable history of academic writing on the retailing of food, 4  which permits a degree of historical comparison and provides theoretical context. Modern grocery retailing requires the coordination of tens of thousands of products from thousands o f suppliers through distribution warehouses to the thousands o f stores that are increasingly the size that malls used to be. Somewhere in this system, I thought, there might be room for electronic commerce to make life simpler for the customer, and more complicated for the retailer. If I could manage to convincingly research both demand and supply sides of the market, I thought, I would have a thesis. As the vignette above indicates, however, I was also interested in online banking. The two industries are not unrelated, o f course: many supermarkets now contain "kiosk" bank branches (Figure 1.3) or allow customer to conduct banking transactions at the checkout (Alexander and  The atom- versus bit-based nature of products is explored in more depth in Chapter 6, where these terms are more fully introduced.  5  Pollard, 2000). I figured that customers that would be interested in online grocery shopping would also be interested in online banking, for much the same reasons: time (and space) management. Online banking has indeed become ubiquitous, and the state of banking has moved on considerably to include the very trends I was speculating about. However, getting access to bankers proved to be difficult, while grocers were only too pleased to be involved in my study. The topic o f remote and online banking has also been studied quite extensively (Marshall and Richardson, 1996; Pratt, 1998; Alexander and Pollard, 2000; Leyshon and Pollard, 2000; Nath, Schrick and Parzinger, 2001; Willis, Marshall and Richardson, 2001; Cobb, 2002), whereas remote shopping remains under-theorised and under-empiricised. The research process, as is so often the case, meant a narrowing o f interest and activity to the first of the vignettes. Food became the thesis, and vice versa (in the academic sense of feeding a research career). Online grocery has not become as ubiquitous as online banking, for reasons that will be explored and explained through this thesis. However, this does not imply that online banking has undergone revolutionary change while grocery remains largely unaffected. Time has not stopped still for food retailers: innovations, both technological and institutional, have proceeded apace. O f the companies mentioned in the vignette, Stria Communications / Peachtree and Webvan have gone bankrupt, and Dairy Farm sold Woolworths to an Australasian rival. Stong's Markets still exists, and now offers an online service as a true "independent trader", but has been eclipsed by others in both scale and research activity, as described in detail in Chapter 2.  Figure 1.3: Safeway banner advertising bank kiosk  Source: photo by author of Safeway on 4 Ave, Vancouver, June 2001 th  6  The World Wide Web itself is barely recognisable four years on from that depicted in my proposal. Accessing the internet has a long history, of course, particularly in academia, but it was the media 5  frenzy that surrounded the Initial Public Offering (IPO) of Netscape Communications in 1995, and the release of Windows 95 (with the ease o f network connections that it brought to the notakeady-enlightened-with-a-Macintosh computer user), that propagated the Web and the prospect of mass electronic commerce onto the global stage. Netscape quickly became a legend of stock price inflation, followed in A p r i l 1996 by the I P O of Yahoo and subsequently by other."portals". The internet gold rush was on, and the "new" / "internet" economy seemed to promise an escape for advanced economies from the clutches o f business cycles as well as geography (Cairncross, 1997). In the meantime Amazon.com became the bookstore o f first choice, i f only for academics wishing to keep up with new releases or bibliographic information; and Google is now the search engine and information provider par excellence. Academic discourse around the internet was slow in coming, but there was a burgeoning o f media coverage, itself a schizophrenic mix of "new economy" boosterism and cynicism. For example, a search within E B S C O H o s t (an online academic database, largely indexing business-related journals and newspapers) returns 1,618 mentions of the subject "electronic commerce" in its index in 1997, following 303 in 1996 and just 32 in 1995 (and two mentions in 1991); this rose rapidly to 3,013 in 1998, 7,761 in 1999, and 13,724 in 2000 at the height of the dot-com boom (Figure 1.4). Following the market correction the "column inches" devoted to the subject fell along with stock and company valuations, with mentions declining to 7,023 in 2001 and 3,406 in 2002 (and a mere 1,211 in the first half of 2003).  6  I will not take up valuable space here with an historical review of the development of the internet and electronic commerce; Hicks (1998), Castells (2001) and Cassidy (2002) do a good enough job of that. Note that the alternative "e-commerce" was not rising in its place; in 2000 it peaked with 53 mentions. Amazon.com listed 2,868 books under the subject "electronic commerce" in July 2003, ranging from programming guides to the marketing self-help guide "MindControlMarketing.com: How Everyday People are Using Forbidden Mind Control Psychology and Ruthless Military Tactics to Make Millions Online". Clearly, with 105 customer reviews and an Amazon.com Sales Rank of 1,186 in May 2003 [ie, there are only 1,185 books that sell better at Amazon], quite a number of customers agreed with the statement that "it's all about marketing and getting people to buy without thinking about it", as one reviewer of the book put it.  7  Figure 1.4: Mentions of "electronic commerce" in business journals and newspapers, 1991-2002 16000 -r 14000 12000 10000 8000 6000 4000 2000 01991  1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002  Source: search of EBSCOHost subject index mentions of 'SU "electronic commerce'", July 2 1 , 2003  Newspapers, popular magazines, and television programmes now include plethora o f references to the internet and specific websites: on privacy concerns, unwanted and usually sexually explicit emails (SPAM), the latest losses o f internet-related companies, or the latest tribulations of search engines. Google, for example was criticised in late 2002 for "filtering" (removing links to) its content to satisfy German laws on inciting racial hatred, and was being sued by one company for manually reducing its position in the supposedly automatic "page rank" algorithm (which indexes the connectivity of webpages, rewarding well-connected websites a high ranking) because the website in question was tactically boosting the numbers of (related) pages which linked to it. One critic (and competitor) o f Google "even suggested that the wildly popular search engine be transformed into a government-controlled "public utility"" (Olsen, 2002). Also in the news at the same time was (unfounded) concern that two murdered British girls might have been abducted by someone they had met virtually while visiting internet chat rooms; that Microsoft and other software and content providers might be tracking individual users' actions through the use o f "cookies" and computer identifiers; and that CairoTbased website Islam Online used the internet to "present a positive view o f the faith to non-Muslims", and to "push for the Islamization o f societies by nonviolent means" (Wakin, 2002). Clearly the internet is having an impact on the lives and fortunes o f individuals and organisations, with these impacts encompassing economic, political, social, cultural and religious spheres. H o w far these concerns penetrated the consciousness of individuals was of great interest to me,  8  somediing I address for the customers I interviewed in Chapter 5, and to individuals within organisations in Chapter 4. Before outlining the rationale for the construction and sequencing of these chapters, however, I wish to briefly explore how the internet came to play such an obvious part in everyday lives.  1.2 Internet and E-Commerce Growth: Feeding and Defying the Hype The growth of the internet over the past decade has been extraordinary. Reliable cross-country comparisons are difficult to make, but this section will review the data that are available on internet infrastructure, usage and electronic commerce, for Canada, N e w Zealand, the United Kingdom and the United States. The U S is, o f course, the largest economy of the four, and the most advanced in terms o f internet investment and adoption by government, businesses and customers. It is perhaps "ahead of the curve" in terms of technological diffusion, and thus may point the way for the other examples; alternatively, it may be a unique case which the others may envy but not equal. It also provides a wealth of information freely, with the Department o f Commerce and the Census Bureau both commissioning studies of the internet and e-commerce and incorporating measures into broader surveys, which I draw upon in this section. For  a number  of years  the U S Department  of Commerce's Economics and Statistics  Ackninistration with the National Telecommunications and Information Administration (ESAN T I A ) has conducted surveys of the adoption of computers and the internet by American households and individuals. The most recent report (February 2002) stated that: Americans' use of information technologies grew at phenomenal rates in 2001. This past year saw a rapid increase in computer and Internet use, not only in homes, but also at the workplace, schools, and other locations. (ESAN T I A 2002:1) .  A t the time more than 50 percent of American households had access to the internet, a rapid diffusion relative even to that of the computer (Figure 1.5). However, on an individual basis the penetration o f the technology was even greater, at 53.9% (Figure 1.6), as individuals had access to the internet at the workplace or school; even more had access to the technology at home, but did not use it. Measuring the diffusion of the internet, and particularly access to it, is obviously problematic ( E S A - N T L A , 2002). A s Figure 1.6 shows, different measures o f the same concept can generate quite different results.  9  Figure 1.5: US Households with Computer and Internet Connections 100 so BO 70 60  ei-Q 'sa.s  !30 4.2  10 1  1  ,  1  f~  m  1  1964  m m  1  22J  150 If —r  7*u  24L1 me  r——i  1  1982  r-  1-  1  ,  1996  r-  .  .—  2000  Source: ESA-NTIA (2002: 3)  Figure 1.6: Different Perspectives on Internet Access and Use, US 100 so 60 70 AO  53.8  60  A4A  4«  40  M.7I  30  32.7  20 10 0 HouMhold Internet 3JMal»«hlp  PMKH»«WI Intemat AfiOMftstHem*  Penont U i h g Home Aeoau  Potvam LWng Internet Anywtwn  I Daft 1890 • Aug. 2000 DSapt 2001  Source: ESA-NTIA (2002: 6)  Thus when cross-country comparisons are made, it is difficult to be certain that like is being compared to like (Figure 1.7). However, while access and use i n the U S , U K , Canada and New Zealand now encompass a majority o f the population, this growth has largely plateaued and there are indications that a number of previously "wired" citizens i n these countries have unplugged, either involuntarily through losing access (changing or losing jobs, for example) or voluntarily 10  (from a lack of need) (Wyatt, Thomas and Terranova, 2002). There are also indications that the United States is not the leader in internet adoption, with Scandinavian countries (particularly Sweden and Denmark) and H o n g K o n g showing higher rates o f adoption.  Figure 1.7: Internet adoption in US, UK, Canada and NZ households, 1997-2002 70  US  60  UK New Zealand  50  Canada  40 30 20 10  (5 ->  ci <  -2 ^  O  (0  O. <  -5  -  t5  Q^ 6 ^ < a  6£ ^ < ^  o ^ I '  o ^  |  Source: Nua Internet Surveys from various sources, last accessed March 31 2003 [www.nua.com/survcys/how_many_onlinel  The internet has had a rapid take-up in many countries, and even the more expensive but much faster "broadband" access technology has been adopted at a more rapid rate than other recent technological advances, such as the V C R and cell phone (Figure 1.8). The diffusion o f broadband 7  is a keenly watched statistic, since it is assumed by commentators (and retailers) that with faster and "always on" access to the internet, users are more likely to turn to the internet as a retail (and entertainment) channel o f first choice, whether for information or for ordering. Substantial growth in electronic commerce sales is therefore likely to lag the take-up o f access to a more data- and visually-enriched retail experience.  7  Because of free local calls and technological and charging differences between cell phone companies the cell spread of broadband, but free local calls for "plain old telephone service" modems dampens demand  11  Figure 1.8: Diffusion of Technologies in the USA: broadband vs cell phones  1 -••  2 Broaiftnnd  3 CokrTV  4 Pagan  6 Cdl Phone*  6 VCRs  Source: ESA-NTIA (2002: 37), with X-axis depicting years from introduction. Year 5 for broadband was early 2001.  Retail sales over the internet in the United States have increased at a rapid rate i n the last four years, doubling from US$5.4b in the last three months o f 1999 to nearly US$14b i n the fourth quarter of 2002, an increase of 155 percent (Figure 1.9). In the same period total retail sales increased from US$787b to US$864b, or 9.8%. A s a proportion of total sales, therefore, e8  commerce is still a minnow, albeit growing, rising from 0.7% o f all retail sales at the end of 1999 to 1.6% at the end of 2002 (Figure 1.10). Elsewhere, where statistics are more erratically collected and reported, the general pattern seems to lag that of the United States by some months or years. Forecasts for this period were significantly more optimistic. Retailers and other organisations were interested i n the internet i n the late 1990s because it was the "thing o f the moment": part o f an assemblage of technologies that enabled the emergence of a "new economy" (Jones and Biasiotto, 1999; Williams, 2001; Beyers, 2002). According to Thrift (2001: 414) this promised: strong non-inflationary growth arising out of the increasing influence of information and communications technology and the associated restructuring of economic activity. All kinds of other features can be, and usually are, associated with this core definition - for example, the growth of small high-tech firms, the increasing importance of mobile and highly skilled talent, the rise of entrepreneurship and the centrality of venture capital. And it is almost second nature for commentators to produce grand rhetorical flourishes such as the death of the business cycle or virtually unlimited growth. What seems certain is that the new economy is both a description and, at the same time, an assumption of what constitutes a normal future.  Taking all available quarters - which for the data available means comparing the "bumper" Christmas quarter with the "lean" winter quarter that follows - e-commerce grew by 121% from December 1999 to March 2003; total retail sales, by comparison, fell by 1.9%.  12  Figure 1.9: US E-Commerce and Total Retail Sales, quarterly, 1999-2003  16,000  1,000,000 900,000  _ 14,000 ? 12,000  ^  .  -  * "  + \  w  *  #  ^  - - i /  1 10,000 co  g  800,000 700,000 600,000  8,000  500,000  E  6,000  400,000  o o  4,000  300,000  LU  2,000  0)  0  200,000  - - -Total —  I  1  1  1  1—1_ —  i  —  i  —  i  —  i  E w o  ro CO  "5 JS  01  "ro o  100,000 1—u  0  & & q? > ^ V VT0f>>VV Source US Census Bureau Monthly Retail Trade Survey, May 2003, in US$m  While Thrift's "rhetorical flourishes" abound in the popular literature, electronic commerce was heralded as the conduit for real productivity improvements through reduced transaction costs, perhaps finally refuting Solow's "productivity paradox", where the effects  of information  technology show up everywhere except in the productivity statistics (Castells, 2000; Williams, 2001; Daniels, 2003; Taylor and Murphy, 2003). The tremendous and historically unprecedented 9  increase in stock market prices from the mid 1990s (Figure 1.11) led to warnings o f "irrational exuberance" from Alan Greenspan and Robert Shiller (Shiller, 2001). A m o n g the more culturallyminded and individual-oriented, the internet also promised to unleash the spatial binds and creative powers o f citizens, with discussion of virtual communities, cyberactivism, cyberdemocracy and the rise o f "informational politics" (Crang, Crang and May, 1999; G Graham, 1999; Green and Adam, 1999; Castells, 2001; Woolgar, 2002).  However, Beyers (2002), Daniels (2003) and Pratt (2003) have noted that other periods in the twentieth century have shown equal or better growth, such as the "golden period" of 1950 to 1973. 13  Figure 1.10: US E-Commerce as percent of Total Retail Sales, 1999-2003  1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 \ &  f  Q  «S> f^  &  ^  ^  ^ « F  <V  V  &  &  -NN  ^  f  &  f &  v  ^ #  f  v N  f  cy* ^  <v ^  f #  v  v  #  v  &  v  Source: US Census Bureau Monthly Retail Trade Survey, May 2003  Figure 1.11: US Stock prices versus earnings, 1881-2003  1881  1891  1901  1911  1921  1931  1941  1951  1961  1971  1981  Source: modified from Shiller (2001), updated with data from his website [www.econ.yale.edu/~shiller/data.htm] Ratio of Standard & Poors 500 Price Index to Earnings, using previous 10 year moving average for earnings  14  1991  2001  It is worth reproducing one such boosterish account i n full, to give an indication of the "tone of the times" i n the late 1990s. According to David Batstone (1999, no pagination), writing i n one of the many new "new economy" business magazines,  there are "10 principles" of the N e w  Economy: 1.  Matter. It matters less ... Processing information is dramatically more powerful and cost-effective than moving physical products. Increasingly, the value of a company is to be found not in its tangible assets, but in intangibles: people, ideas, and the strategic aggregation of key information-driven assets.  2.  Space. Distance has vanished. The world is your customer — and your competitor. Geography has always played a key role in determining who competes with whom. Now your business can connect instantiy with customers all over the globe.  3. Time. It's collapsing. Instant interactivity is critical, and is breeding accelerated change. In a world of instantaneous connection, there is a premium on instant response and the ability to learn from and adapt to the marketplace in real time. 4. People. They're the crown jewels... and they know it. Brain power ... is the prime factor driving the New Economy. 5.  Growth. It's accelerated by the network. The Internet can dramatically boost the adoption of a product or service by "viral marketing", network-enhanced word-of-mouth.  6. Value. It rises exponentially with market share. ... a striking exception to the economic rule of scarcity. Network effects were experienced historically in the adoption of telephones and fax machines. 7. Efficiency. The middleman lives. "Infomediaries" replace intermediaries. As the amount of info-clutter grows, these infomediaries are needed to turn dumb data into usable information. 8. Markets. Buyers are gaining dramatic new power — and sellers new opportunity. Intelligent software will help buyers find the best deal. ... Businesses that have relied on physical barriers to competition will fail. 9.  Transactions. It's a one-on-one game. Suppliers will find it easier to customise products, and customers will begin to demand this sort of tailoring.  10. Impulse. Every product is available everywhere. The gap between desire and purchase has closed. ... The impulse to buy and the purchase itself used to be separated by a combination of physical and mental barriers. ... The processes for marketing, sales and fulfillment are merging. These "principles" draw on a range of "hot topics" from various time periods i n a variety of fields: time-space compression (Harvey, 1989); the "end of geography" and the "death o f distance" (O'Brien, 1992, Cairncross, 1997), and the "weightless economy" (Coyle, 1997, Quah, 1999); relational views o f firms and society; consumer empowerment, and so on. The arguments seemed compelling, and investors were duly compelled. In 1999 and 2000 the nexus of Wall Street analysts, research firms and media - which Thrift (2001) refers to as a "cultural circuit of capital" — created a frenzy of speculation about the growth o f the internet and electronic commerce (Cassidy, 2002; Pratt, 2003). Figure 1.12 is indicative of this, with both panels published i n November 1999, around the time of the Initial Public Offering of one online grocer, Webvan (on which much more in Chapters 2, 4 and 6):  15  Figure 1.12: Online Population and Sales Forecasts, November 1999 Will Online Buying Hit $185 Billion by 2004?  IDC PREDICTS 1 7 9 M I L L I O N A M E R I C A N S ONLINE I T 2 0 0 3  U . S . O N L I N E S H O P P I N G FORECASTS BT R E S E A R C H E R  U.S. Online Population Forecasts by Researcher $200  «—>Fwrester Research  200  Jupitei Communications  $150  — i -i:IT si .it on.) I Data Corp. — Yank** Group  SIM  ISO  1  SW  3  100  4  o  '01  '02  '03,  '0*  N/A  Z  40  CommtrteNei/NMseit  $7.8  *»  '00  SU IfliA T7J>" 90 2 Otter OUioft* (uoMnotM 74.5 KJO tnteflMthWMl OMt Corp. 82.0 109X1 lO&i 1K4  Cyber tMosw{*sgnrwlw)  '91  '02  N/A  103.0 90S 1)1.0 1»JI  ^ null  N/A  tilt 99.1  1302 109-4  ' ISAJO  179-0 194.7  1441  -  I ias  i B A * I spj>liy»j I N / A  SIS.* 2 S35.B MS I $60 6 $75.0 f t l l J f 'SMOj SH.t j S>fJi'"»JB\)tj$12i.6  Source: Industry Standard Media Metrics November 17 and 30, 1999 (http://ww.thestandard.com/rescarch/metrics/display/0,2799,10077,00.html) (http://www.diestandard.com/research/metrics/displa)'/0,2799J0(189,(Xl.html) »10  Almost any chart published during the "dot-com boom"'" looks pretty much the same: an inexorable rise in demand predicted for online retailers. A year earlier, the same pattern appears (Figure 1.13), this time specifically for online grocery, accompanied by some reasons for why customers (in the US) shop with online food retailers. Grocery was attractive because of its sheer size (the single largest retail industry), and its supposed inefficiencies in the number o f "intermediaries" who "touch" products (without leaving any 'finger marks of exploitation', according to Harvey, 1989: 101) along the way. The sales pitch was compelling to investors, who bid the share prices of the highlighted companies through the roof. A s one such analyst and reporter put it, As long as that cheap financing existed, no companies were safe. For example, I sold my supermarket stocks because I figured that every online delivery service that wanted to would get a billion dollars to destroy America's supermarket business. ... I figured they could borrow a page from Amazon and lose money on everything they sold until they wiped out the existing supermarkets. Seemed plausible. Who knows how long a financing fad is going to last? It didn't last long enough. N o w it is time for the bricks-and-mortar companies to strike back. A n d they will. ... I had seen so many people bet against Amazon and go out of business doing so ... Y o u just  Which is how the period has come to be known, perhaps because Netscape started it; broader "technology stocks" including telecoms were also involved and more heavily affected - witness the coverage given to Nortel's rise and later plunge in Canada and its effects on Canadian pensions, and that the collapse of Enron in 2001 probably signalled the end of the dot-com bust, as the markets have gone sideways since.  16  didn't know when the people would run out of an appetite for startups. . . . Sure was goofy while it lasted. (Cramer, 2000) Figure 1.13: Online Grocery Sales Forecasts and Reasons, September 1998 OMiNt  SAUS  9HHHHHHHHHHHBHHHHHBHHHHBHHH COMVSHIIXCE Is T H E P R I M A R Y D R I V E R  Cmiio top $6 BlUION IN 2002  Onliitt Grocery Revenue  Convenience 2H-Hmif Access Cm Shop Without inrint Howe SawTtmt Sara Money Compare Met* Wonttogel Stun Better Selection rind New Products Can Shop From Work Mite to Shop WM#oii juevFrc* L W U V  MX «•% «0% sr* *A 42%  41% J»* »%  «vrriNifT vsios.  Source: Industry Standard Media Metrics September 7, 1998 (www.thestandard.net/metncs.display/O.^SS.leze.OOhtml)  The Economist was prone to the odd apocryphal story at the time, with a germane one on Encyclopaedia Britannica, which survived a change of ownership in 1920 to Sears Roebuck, an upstart mail-order retailer, and a move of its base to Chicago. Two decades later it became the property of a foundation, the proceeds of which benefited the University of Chicago. Its door-to-door salesmen became legendary for their knack of persuading parents that an encyclopedia was a must if their children were to grow up well-educated. By 1989, sales revenues had reached a record $650m, and the firm's sales force had grown to 2,300. Yet disaster was already waiting in the wings. It began when Microsoft produced a cheap C D - R O M encyclopedia called Encarta in 1993. Britannica's bosses responded with a C D - R O M of their own, but their salesmen rebelled at seeing their hard-cover books undercut (cannibalisation, followed by channel conflict). The Internet made things infinitely worse. Sales collapsed, along with the salesforce. In 1995 the company changed hands again, for a small fraction of its book value. Today it employs fewer than 350 people. Last October the new owner decided to make the contents of Encyclopaedia Britannica available free on the Internet, in the optimistic hope of recouping the cost through advertising on the website. But, as if to question even this new strategy, the site promptly crashed. What lesson should be drawn from all this? The most pertinent one is this: if you are engaged in almost any form of commerce, from the humble market stall to the gigantic department store, the Internet is lurking, ready to blow up your business. And if you are not prepared to embrace change, to cannibalise your own sales and to disintermediate your own intermediaries, somebody else surely will. (The Economist, 2000b: 54)"  In the same article, it warned that 'many shops will have become little more than showrooms and places o f entertainment, with the actual buying and selling going on elsewhere': Only five years ago, in the pre-Amazon era, almost nothing was bought online. But extrapolate the past five years' growth, and it is easy to believe that by 2010 close to a fifth of American retail spending could be carried out over  This story was first used by Evans and Wurster (1997) and retold in Evans and Wurster (2000). Thanks go to Andrew Leyshon for pointing this out.  17  the Internet. And for things like books, music,films,toys, white goods and cars, the share seems likely to be much, much bigger.  This line o f argument suggests that organisations are likely to be significantly transformed by the internet, as it is a "disruptive technology" (Christensen, 1997; Christensen and Tedlow, 2000). "Transformed" organisations are seen as the top o f an ascending ladder of impact (Figure 1.14), as if changes o f this sort are progressive and cumulative. A n alternative conception is a more messy (or "muddled", Leyshon and Pollard, 2000) overlapping o f change (Figure 1.15). Here, rather than a linear progression toward greater levels o f sophisticated technology use across all functions, firms instead are seen as potentially utilising technology in greater depth in a variety of activity areas, (including logistics, purchasing and marketing), at different rates i n different areas, along with new synergies that might emerge in truly transformed organisations.  Figure 1.14: Transformative Power of the Internet: the ladder metaphor  5, transformed organisations 4. e-bbsiness 3, e-commerce  - open systems inf6rmatiph^\ customers^]] [suppliers dnd\y partners •' :  - integrate "Mppfychain •  •^newibiisiness¥i •mam^facture OH:mbdels!bdsed - order and yihdi^lrveiyfi-:'-, pay qrihne^::, become interworking,; i, jr^ducing:c0tsi \sedmless^:f between 1  2, website  1. e-mail  '•^p^dce)i^:'ff i-;maximise0^ accessibility . worldwide\rnarketfi{"}": ^nd'sp0d ; .' . :  internal and ^exiernal^ff^^ 'colnmumcahoh^  :|  [organisations:;':] ^-fnm^se'^f ^-j (:;:;Fpii:'I:'' y::• waste at every ' individuals • stage of the :  :  1 1!  ;: :  [St^fycMi^iMiti  Mwi0ow;pri% 'Woridrnde' isuppliers  Source: Taylor and Murphy (2003: 15), after Martin and Maday (2001)  A s The Economist (2001, as quoted i n Thrift, 2001: 419) reported, there was (and is) lasting bitterness about the self-serving nature of analyst recommendations: Only yesterday, so it seems, Wall Street equity analysts almost unanimously proclaimed a new economic paradigm. Out with these old equity valuation models, out with fusty concerns about earnings (actual or predicted), out with the business cycle, in with network effects, bum rates, and global scale. Forget, ugh, prudence: caution is the new  18  recklessness. Nowadays, as one reputable member of the breed then put it, the only danger is to be out of the market. Well, for the shrewd advice (as NASDAQ tottered at around 5,000) many thanks. For all those 'busy', 'hold' and 'accumulate' recommendations on stocks that cost $100 last year, and now cost $1.50, thanks a lot.  Figure 1.15: ICT Adoption in Businesses  Logistics and delivery  Automated b i l l rig  Finance  Purchasing and procurement  Automated stock replenishment  Operations, processing & assembly  r  <1  Mass custcmisation  Marketing and sales >  After sales service  Publish product warranty, senietng and statutory information  Capture and analyse customerfeedbackto rron itor product/service performance  C u s t o m e r driven relationship marketing  Customer driver design and relationship marketing  Source: Taylor and Murphy (2003: 16), after Foley and Ram (2002)  One reaction to these failures and the attending bitterness was a somewhat gleeful coverage of the mounting  losses  and  closures  among  the  "dot-com"  firms,  as  typified  by  the  site  "dotcomfailures.com" (Figure 1.16). Ironically the dotcomfailures.com website itself is another failure, and going to its defunct website in May 2003 brings up a directory for the website host (WebQuest). WebQuest's site had a 'banner ad' (Figure 1.17) that aims to deceive surfers into downloading browsercleanser.com's "privacy" software.  12  In case this seems like its going off-topic, my aim here is to replicate the "surfing" patterns of customers, whereby weblinks are followed, sometimes to detrimental ends. Banner ads posing as genuine security alerts are a hazard to unwitting consumers, as they may contain links to pornographic sites, viruses, or in this case an unabashedly unreflexive privacy site that has the tag line "Because it's nobody's business but yours". The website has this warning: Did you know your computer tracks EVERYWHERE you go on the net, and you CAN'T GET RID OF the records by yourself? What if you accidentally went to an illegal or adult site? The record is already there! It doesn't matter to you're your [sic] boss, or your wife! You are already guilty! Stay out of trouble! More importantly, get peace of mind! Workers are being FIRED from their jobs! People are losing MARRIAGES! It's in the headlines every day! (http://www.browsercleanser.com, accessed May 29, 2003)  19  Figure 1.16: Dotcomfailures.com: November 2000  D0IT • M FAILURES •  tc o  1 Si  I r s A L W A Y S D A R K E S T B E F O R E IT G O E S P I T C H B L A C K .  Source: www.dotcomfaaures.com, accessed November 23, 2000.  Figure 1.17: Dotcomfailures.com: May 2003  dotcomfailures.com Source: www.dotcomfailures.com, accessed May 29, 2003  A further reaction i n some quarters was to lampoon the whole idea o f electronically-mediated shopping. Woolgar (2002: 7-8), for example, highlighted an apocryphal and hypothetical story:  The advent of Internet shopping, it is said, is set to revolutionise the practices and experience of shopping (cf Knights et al., a degree never imagined by home shopping via television, or through mail order catalogues, the possibility of shopping onlin more and more people can opt for the convenience of ordering their groceries via the Internet. Instead of enduring the we slogging to the supermarket, filling large trolleys, queuing and processing it all through the check-out, transferring bags t driving home and unloading, it is now possible to order viayour PC and have the weekly shop delivered to your door.  However, this attractive scenario has a drawback. As this means of shopping becomes more and more popular, so more delivery vans are deployed in transporting the goods. As numbers grow, the traffic becomes increasingly congested. Road up. To prevent total seizure of the road system someone suggests a way of rationalising the delivery procedure. Instead of d individual households it is decided to combine deliveries to specific localities by depositing composite orders to local depo shoppers can then collect their orders from the depots. This not only reduces the delivery traffic, it also solves the emergin many Internet shoppers are not actually home to receive the deliveries when they arrive. Now they can collect themfrom t own convenience. The further idea arises that these depots could try diversifying their businesses by offering various sideli Internet shoppers might not have thought to order when online, newspapers, a lottery ticket, shoe laces, a box of matches. S up with a further brilliant idea: these depots could be called 'shops'! The appeal of this kind of story ... lies in the inversion of the claimed effects - specifically, the manner and degree to which the envisaged displacement of an existing system by new technology turns out to be misjudged. Something very like the old existing system re-emerges once the hype has died down. ... The inversion would be very much less marked if the original claims had been more moderate and cautious.  Woolgar  (2002) invents a neologism for Thrift's  "rhetorical flourish":  cyberbole, being the  application o f hyperbole for cyberspace, and something analysts and the media were particularly prone to during the dot-com bubble. A s Knights et al. (2002: 103) note, Consider, for instance, the anno mirabillis of 1998-9, which in the U S A was characterized by the seemingly irresistible rise of the so-called internet stocks. ... Thus Yahoo came to have a stock-market valuation higher than that of Boeing, while, at $30 billion, loss-making bookseller Amazon.com was briefly worth more than all other US booksellers put together [a highly dubious and unsubstantiated claim]. The experience of Active Apparel, a medium-sized sportswear company, is instructive. The firm's value had fallen to an all-time low of less than $2 million and bankruptcy loomed. Then, on 28 December 1998, it 'went online' - that is, commenced retailing via a web site. Within days, and on the strength of this (rather ad hoc) link with the internet, Active Apparel's stockmarket valuation increased by a factor of fifty. ... in February 2000 a strange conjunction arose in which Dixons,  20  one of the UK's most successful retailers, was valued at less than the company's own stake in loss-making IPO Freeserve, which implied that Dixon's 'traditional' businesses were worth less than nothing.  These types o f technology impact narrative are divided into two groups by Knights et al. (2002: 100): the "future shock" story type, following Toffler (1971); here organisations are (or stand to be) "overwhelmed" by new innovations (or "disruptive technologies", as Christensen (1997) terms them), and must follow the guidance of what Knights et al. (2002) term the "cyberprophets". The alternative narrative grouping is the "emperor's new clothes" story type, whereby Various instances of awe-mspiring technological or organizational prowess are 'exposed' as little more than cyberhype' (p. 100). Clearly most of the effects o f information and communication technologies lie somewhere inbetween these "book selling points", and this thesis attempts to avoid veering into either unabashed boosterism or excessive cynicism, by combining detailed empirical work with theory. The discussion above situates and contextualises the argument that follows in the remainder of this thesis. In the long tradition of geography the internet clearly presents a new space to explore, map, conquer, sell and trade in. A s the above discussion has shown, this virtual space is highly uneven, with great variation in technological development, access and interest, both witiiin countries (as demonstrated in the various "Digital Divide" reports from the United States, European Union and others (Taylor and Murphy, 2003)) and between them. Behind these reports, and the corrective policies that inevitably follow, lies the assumption that greater use of the internet is a sign of progression and sophistication, and is therefore to be encouraged. However, as Wyatt et al. (2002: 25) noted, access to the internet is seen as necessarily good. ... It is assumed that, once ... barriers to use are overcome, people will embrace the technology wholeheartedly. From the perspective of politicians, the hope is that people will use this knowledge to create wealth and employment, but maybe they will use it to look at pornography, play games, or trace long-lost friends and relatives. ... Maybe some people will not use it at all and — hard though it might be to accept — maybe its lack does not have to be a source of inequality and disadvantage.  1.3 Situating the Thesis Situated in the nexus between retailing and information technologies, and between supply and demand, this thesis necessarily draws on a range of literature. This section briefly summarises three of the 'core' literatures which form the context for the research that follows. It acts as a prelude to the more specific literature reviews that accompany and precede each empirical chapter.  21  1.3.1  Retailing  The retailing literature can be broadly categorised into three camps: "orthodox" retail studies of market areas, customer travel and expenditure patterns, and supply chain management; the "new retail geography"; and marketing studies of store and product brand management. Studies in the first tradition continue in the guise of Clark (1998) and Birkin, Clarke and Clarke (2002), and in operations research (Beamon, 1998; Lam, Vandenbosch and Pearce, 1998; Ganeshan, 1999; Hadjiconstantinou, 1999; Waller, Johnson and Davis, 1999; Abernathy, Dunlop, Hammond and Weil, 2000; Myers, Daugherty and Autry, 2000), but the more traditional lines of this research is dismissed by some as 'a largely descriptive and all too often simplistic mapping o f store location, location, location' (Crewe, 2000: 275) that is one o f the most undertheori2ed and 'boring of fields' (Blomley, 1996: 256). The 'new retail geography', on the other hand, following a call by Ducatel and Blomley (1990: 207) for a more 'systematic account of retail capital', is concerned with the geography of retail restxucftiring, particularly retail consolidation (Wrigley, 1987, 1997) and globalisation (Wrigley, 1993, 2000a, 2000b; Wrigley and Currah 2003a, 2003b, 2003c); the interface between retail and financial capital (Wood, 2001, 2002; Marston and Modarres, 2002); the "complex and contradictory" relations o f retailing with the regulatory state (Marsden and Wrigley, 1995, 1996; Wrigley, 1997, 2001, 2002a), and the spatial switching o f retail capital (Wrigley, 1991, 1993, 1994; Christopherson, 1993). M u c h work has also been conducted on the "backwards" connections between retailers and their suppliers, mcluding the interrelations of corporate strategy, corporate culture and market structures (Hughes, 1996a; Shackleton, 1998; Wrigley, 2000b); the social relations of 'production' in retailing (Marsden and Wrigley, 1996; Marsden, Harrison and Flynn, 1998); and the structure of and power relations within supply chains and retail channels (Doel, 1996; Foord, Bowlby and Tillsley, 1996; Hughes, 1996b, 1999a, 1999b). This body of work was at pains to note the "complex and contradictory" ways in which retail strategies are created and enacted by retailers and restrained or enforced by regulators, including, for example, the forced divestment of overlapping store networks in the United States as a result o f mergers and acquisitions; and the variously predatory and supportive relationships between retailers and suppliers. The most recent work within this body of research has considered "gaps on the map". The spatial switching o f retail capital (Lowe and Wrigley, 1996; Crewe, 2000; Wrigley and Lowe, 2002) away from socio-economicaily deprived areas has been shown to create "food deserts" (Whelan et al., 2002; Wrigley, 2002b; Wrigley, Warm and Margetts, 2003), with attendant effects on consumer  22  health. In a similar light, Marsden (1998, 2000) and Morgan and Murdoch (2000) have investigated food scares, food quality, and how retailers have stymied and could yet support  sustainable  agriculture and local rural development. The marketing strand of retail research has focussed on attempts to model branding and customer segmentation (Lockshin, Spawton and Macintosh, 1997; Burt, 2000; Uusitalo, 2001) and customer behaviour in searching for information and products (Smith and Carsky, 1996; Putrevu and Ratchford, 1997; Moorthy, Ratchford and Talukdar, 1997; Petrevu and Lord, 2001; Ratchford, 2001), among other themes. Marketers' interest in customer behaviour and satisfaction more generally has continued (Wirtz, 2000; Manila and Wirtz, 2001), combined with an emerging interest in how customers respond to various stimuli (sounds, smells as well as sights) in the store and to the "customer experience" as entertainment (Jones, 1999; Pine and Gilmore, 1999; Schmitt, 1999; K i m , 2001). More recently there has been an increase in consideration o f how the internet might alter existing marketing strategies and channels, although the focus is again on conceptual model-building (Strader and Hendrickson, 1999; Szymanski and Hise, 2000; Sang and Young, 2001). Clearly there is no particular uniformity in the research being conducted under the name of "retailing", whether "retail geography" or "retail marketing". In fact the three strands identified above can be seen as diverging: the first into a G I S - and model-oriented interpretation o f "geography", the second into regulation theory and culturally-sympathetic interrogations o f the economic; and the third primarily into model conceptualisation and empirical testing o f consumer behaviour. Although some o f the above work is directed towards the grocery sector (in particular the voluminous work by Neil Wrigley), there remains a significant gap on the implications of virtual technologies for retailing, and vice versa (though see Section 1.3.4 below for more on this).  1.3.2 Consumption Interest in the customer and acts of consumption has burgeoned in the past decade, promoted both by the desire of the 'new retail geography' to take both economic and cultural geographies of retail spaces seriously (Lowe and Wrigley, 1996; Crewe, 2000), and by the wider "cultural turn" in economic geography (Barnes, 1995, 2001a, 2001b, 2002; Crang, 1997; A m i n . and Thrift, 2000; Philo, 2000). Thus it was increasingly recognised that 'consumer spaces can be at once material sites for commodity exchange and symbolic and metaphoric territories' (Crewe, 2000: 275), and  23  there was greater awareness o f the 'diverse and often contested meanings of ... shopping places and their relationship to the identity of those who shop there' (Miller et al., 1998: ix). The meanings o f places o f consumption were given considerable attention by Gregson and Crewe (1997) and Crewe and Gregson (1998) i n the (rather British) car boot fair. In so doing they focus 13  on the importance of the gift-exchange (also studied by Belk, 1995) and the continuing use- and exchange-values of commodities well after initial 'consumption' at purchase — a point also raised by Jackson and Holbrook (1995), Pred (1996), and Crewe (2000). Their work, in common with others (Cook, Crang and Thorp, 1998, 2003), places heavy emphasis on discourses o f ' v o i c e ' and representation, and for consideration o f the customer's "dialectics of seeing" (Buck-Morss, 1989). Jackson and Thrift (1995: 218) call for textualising the 'voices of actual consumers, respecting their active role as creative agents before, during and after the moment o f purchase'. Crewe (2000) notes that this can be done through an exploration of the "rituals of consumption", something she shares in common with Miller (1998), by focussing on 'the cultural and symbolic dimensions o f exchange rather than simply quantifying the material worth o f a cornmodity transaction' (p. 280). Crewe (2000) draws on Appadurai (1986), who believes 'we should look at the commodity potential o f all things . . . [which] lies in the extent to which the social life of things can be defined by its exchangeability (past, present or future) for some other things' (p. 3). Cook and Crang (1996) and Cook, Crang and Thorp (1998, 2000, 2003) explore a "double commodity fetishism", whereby customers are largely ignorant of how, why and where products are produced, and yet assume certain knowledges (stereotypes) o f the places (if not means) of product origins. Hughes (2000, 2001b) fuses "geographies of consumption" approaches with the more materialistorientated analysis of commodity chains in investigating the 'production sites, labour processes, marketing chains and retail systems involved in the [Kenyan-UK] cut-flower trade' (Crewe, 2000: 282). The commodity chain approach to detenriining the 'finger marks of exploitation in the daily bread' (Harvey, 1989: 101) has been popular among cultural theorists (see, for example, the chapters in Hughes and Reimer, 2003). Commodity chain analyses are a 'way o f tying together material and signified realities, consumption and production, and activities separated by space and markets, providing a fuller interpretation of the material and representational worlds' (Hartwick, 2000: 1190) and so offer 'more culturally nuanced insights into the meanings of goods as they pass  The car boot fair is typically an informal gathering of individuals to sell (or more realistically, to swap) their surplus junk — a collective and-larger scale version of the yard sale or church fair (Gregson, Longstaff and Crewe, 1997; Crewe and Gregson, 1998).  24  through different places and phases of commodity circulation' (Crewe 2001: 631). Hughes and Reimer (2003) extend this work to include speculations on "ethical systems of provision", building on the "stakeholder approach" o f Whysall (2000a, 2000b; see also Chapter 3). Others with an interest in consumption geographies have focused on the body (Valentine, 1999; Crewe, 2001; Cutting Edge Women's Research Group, 2001) and fashion, looking for example at the globalisation and homogenisation o f fashion retailing and consumption typified by The Gap (Crewe and Lowe, 1996; Crewe and Beaverstock, 1998; Marston and Modarres, 2002). A t the consumption end of commodity chains, it has been suggested that the very act o f conspicuous consumption may induce behavioural changes in other customers, 'as with all fashion foods that were once exclusive can be taken up by the masses, and hence rendered less satisfactory as a means of distinction ... something once the privileged preserve o f the knowledgeable can quickly become identified by all' (May, 1996: 60). While this conception of Bourdieu's "cultural capital" might have some carryover to the interpersonal kudos gained by some customers of online grocery shopping — the envy o f paying someone to do all the shopping for them, and having it delivered sweat-free to the door — it is not practical to calculate commodity chains for the whole gamut of products supplied in a typical grocery order that is the focus of this thesis. Instead, as noted by Crewe (2001: 632), 'there remains considerable scope for work which critically interrogates the relations between consumption, individual identity, the power of agency and the potential formation of a collective politics of consumption', with which this thesis aims to engage explicitly in Chapter 5.  1.3.3  Information Technologies, Cyberspace and Culture  M u c h attention has been paid in the literature to computing, the internet and cyberspace over the years. Batty (1993, 1994), for example, outlined the development of the communications and computing technologies and institutional arrangements that gave rise to the internet, as also explained in Castells (2000, 2001). The development of industrial spaces and clusters around computing and software manufacture also have a well-explored path in economic geography (Saxenian, 1994; Coe, 1997; Sassen, 1997; Aoyama, 2000; Coe and Yeung, 2001; Romijn and Albu, 2002). Work on economic geographies of the internet has been primarily based on infrastructure (Zook, 2000, 2001, 2002a, 2002b; Malecki and Gorman, 2001; Gorman, 2002; Malecki, 2002) and mapping (Dodge and Kitchin, 2000; Brunn and Dodge, 2001), although there has also been work on labour (Chappie and Zook, 2002), urban implications (Graham and Marvin, 1996; 2001;  25  Graham, 1998, 2001, 2002), finance and discourse (Thrift, 2000, 2001), a comparative analysis of Japanese and American e-retail sectors (Aoyama, 2001a, 2001b), books (Dodge, 2001), and music (Leyshon, 2001, 2003). A m o n g non-economic geographers, researchers who have taken the internet seriously include Valentine and Holloway's series on children (Bingham, Valentine and Holloway, 1999; Holloway and Valentine, 2001a, 2001b; Valentine and Holloway, 2001a, 2001b, 2002; Valentine, Holloway and Bingham, 2002), and a whole raft of publications on community and identity (Jones, 1998; Kitchin, 1998a, 1998b; Crang, Crang and May, 1999). "Cyberculture", a genre o f writing on the interstices of cyberspace and cultural studies, likewise burgeoned in the late 1990s and early 2000s (Rheingold, 1993; Turkle, 1996; Hicks, 1998; Holeton, 1998; Levy, 2001; Woolgar, 2002a), with an emphasis on texmality, semiotics and metaphor (more on which in Chapter 5), and a sub-section dedicated to issues o f gender and identity (re) formation (Cutting Edge Women's Research Group, 2001; Green and Adam, 2001). Finally, there is a set of publications that attempt to historicize the internet, by comparing computing and communications technologies that comprise it to previous innovations, such as the telegraph (Standage, 1998). Graham (1999: 21), presents one such view: Here is an enormous incalculable force ... let loose suddenly upon mankind; exercising all sorts of influences, social, moral and political; precipitating upon us novel problems which demand immediate solution; banishing the old before the new is half matured to replace it;... Yet, with the curious hardness of a material age, we rarely regard this new power otherwise than as a money-getting andtime-savingmachine ... not many of those ... who believe they can control it ever stop to think of it as ... the most tremendous and far-reaching engine of social change which has ever blessed or cursed mankind.  in quoting Charles Francis Adams Jnr writing on the transcontinental railway in 1868 (as quoted in Boorstin, 1988: 581). Pred (1978: 366) likewise argues that the telegraph and train caused profound change and innovation in the mid-to-late 1800s, enabling the faster circulation of capital by allowing the placing of orders at short notice over long distance, and hence reducing the need for retailers to stock large inventories, de Sola Pool (1977, 1983, 1990) credits similar power to the telephone in the early twentieth century. Given these profound historical changes, it really begs the question: is the internet revolutionary or evolutionary? A s Graham (1999: 5) puts it, is it "radically new" or "merely novel", since: technology, while it obviously increases options in many instances, also frequendy Emits options. People who are very enthusiastic about technology are always telling us what it will do for us. They almost never address the question of what it will undo, (quoting Neil Postman — no reference given; interviewed by Pat Kane)  There is little question that the emergence of a global network of citizens and customers, retailers and producers — able to by-pass established means of communication in locating and transacting  26  with each other — will have considerable impact. There is also no question that this impact will be variable, just as there are differences in the infrastructure underlying the network, the speed and reach in the diffusion of key technologies, as outlined above, and the capabilities and interest of (potential) customers and suppliers. H o w the literature has reported on retailers grappling with these issues is the subject of Chapter 2.  1.3.4  E-Commerce and Online Grocery  I have presented my thesis as a work-in-progress at seven conferences: five Association of American Geographers Annual Meetings (1999-2003 inclusive), the 2000 Institute o f British Geographers conference, and the 1998 Western Canadian Association o f Geographers conference. For the first several presentations it was necessary to begin the discussion with definitions of what "electronic commerce" and the "internet" actually were, so absent were these terms from the academic lexicon, despite the best efforts o f Batty (1997) and Kitchin (1998a, 1998b). Ironically, only since the bursting of the dot-com bubble has there been a cohort o f academic work published on the internet, electronic commerce and cyberspace (Aoyama and Sheppard, 2003). The most hody debated issue around the internet and electronic commerce has been that of the imputed "death of distance", the news o f which has been gready exaggerated. O'Brien (1992) and Cairncross (1997) have alleged that information and communications technologies allow the seamless passing of information, capital and digitised products, and this hyperglobisation will sweep all spatial constraints before it. The rebuttal of this view by geographers was swift and brutal, as an attack was being made on the discipline itself (Malecki and Gorman, 2001). Clarke and O'Connor (1997: 90) noted that 'markets have a role not simply as locations of trading but as centres of information collection and the formal exchange places of specialist expertise'; Zook (2000, 2002a), Gorman (2002) and others have noted how spatially-specific and chscriminatory internet infrastructure and use is (Warf, 2001). There is a plethora of "self help" guides to web site and "e-business" creation, much of it rushed to the presses and instantly outdated. Accompanying this is a collection of more analytical but generally equally uncritical forecasts and reviews of nascent electronic commerce that inhabits the fuzzy border between "academic" and "popular" business research, much of it published in the early years.,Prolific and important contributors to this include Tapscott (1993, 1997; Tapscott, Ticoll and Lowy, 2000), Cronin (1994, 1996, 1997), Negroponte (1995), Whinston and his colleagues (Applegate et al., 1996; Kalakota and Whinston, 1996, 1997; Whinston et al., 1997;  27  Kalakota and Robinson, 2001), Evans and Wurster (2000), Shapiro and Varian (1999) and Kanter (2001). Many of these, and numerous other "how-to" guides, were temporary best-sellers during the dot-com boom, as investors sought to understand the language of the new economy and the apparent new rules. After the collapse of internet stocks from 2000, a whole new set of books were published on how greedy the dot-com entrepreneurs and venture capitalists were (and how gullible the investors) - prime among them Cassidy (2002), Liebowitz (2002), and Malmsten, Portanger and Drazin (2001, on the story of "Boo", the European online fashion chain). However, none o f this boosterish work was by geographers, or spoke particularly to the "geographical" (geographers' emerging contributions will be addressed in later chapters). Equally unsatisfying for the purposes of this research is much of the work published in the management and marketing journals. Earlier papers were devoted to cataloguing the numbers of retailers that had websites or speculation on what the effects of e-commerce on retailers might be (Cope, 1996; Spiller and Lohse, 1997; Griffith and Krampf, 1998; O'Keefe, O'Connor and Kung, 1998; Hart, Doherty and Ellis-Chadwick, 2000; Ellis-Chadwick, Doherty and Hart, 2002). More recent work has tended to construct and codify overly simplistic "models" of e-business and entrepreneurship (Grant, 2000; Palmer, 2000; O'Daniel, 2001; Cloete, 2003), or hypothetical models of logistics (Crowley, 1998; Holmstrom, Tanskanen and Kamarainen, 1999; Kamarainen, 2001; Punakivi and Tanskanen 2002).  1.3.5 Literature Summary These four broad literature groups have been summarised here for a simple reason: the nexus of retailing, consumption, technology and cyberspace lies at the heart o f the questions this thesis aims to address. The literatures introduced above underlie and inform much of the approach that follows. While not specifically covered in depth later, it is important to understand the lineage which this current work takes as a point of departure. Thus, existing bodies of knowledge have addressed questions of retail globalisation, retailer-supplier power relations, the nature of consumption and identity, and the construction and use of cyberspatial technologies, among other themes. This thesis aims to build on these bodies of work by exploring in-depth the example of online grocery shopping, and seeks to contribute to knowledge of institutional structures and processes of and responses to innovations, and how these affect and are in turn affected by consumption. I argue throughout this thesis that to really begin to understand the process of  28  innovation it is necessary to "get inside" (ethnographically) the physical spaces and mindsets of both customers and retailers. A s the review of the retail literature above suggests, "backward" linkages between retailers and their suppliers are well explored, particularly from a political economy perspective. Likewise, the tracking of individual commodities from production to consumption has been documented, notably for tropical fruit (Cook, 1994; May, 1996; Cook et a l , 1998, 2000, 2003) and fresh cut flowers (Hughes, 2000, 2001b). The forward and bi-directional linkages from retailers to customers are less well understood, particularly within recent economic geography. T o analyse these connections from an industry or regional perspective is difficult, especially i f the method of research undertaken is qualitative, because of the sheer numbers o f interviews that would be required. Quantitative studies are available in abundance from market research companies, but at a hefty price, and for new areas of analysis such as the internet their methods and results are of questionable usefulness. I contend that to deconstruct and analyse "the online market" for modern mass-produced commodities, the type stocked in a typical grocery store, requires privileging the (virtual) retail firm as a constrained analytical unit. It is from this vantage point that the connections between retailers and their customers can be explored in depth, through qualitative and ethnographic methods (as discussed in Chapter 2). By bviilding a body of detailed knowledge on the workings o f a few case study firms, we can have a more grounded and situated appreciation of, and provide an empirical basis for, more abstract models of technological and entrepreneurial innovation and diffusion. The next section develops in more depth the questions that are used to guide this process.  1.4 Research Objectives and Questions This thesis aims to provide theoretically-grounded empirically 'thick' descriptions and analyses of the whys and wherefores o f electronic commerce, told through the experiences and visions of three online grocers, the managers who lead them and the employees who work for them (and for companies that are contracted by them), and the customers who shop with them. These stories and interpreted analyses are constructed in two different places, Vancouver and Auckland: pardy to provide a spatial/locale comparison, partly because the numbers o f online grocers in any one place was limited at the time of research initiation (and even more limited now). As Barnes (1996: 250) wrote, in the context of economic geography as a discipline, 'The best we can hope for are shards and fragments ... not one complete story but a set of fragmented stories.' 29  The same is very much true of the internet and its effects. It is much too early, even i f ever possible, to come up with a generalisable (meta) theory of why and how organisations and individuals engage with electronic commerce, and how this engagement in turn changes them. Through the case studies explored here, I hope to collect some fragments o f textualised and pixellated images,' stories, and ideas, and as with an incomplete jigsaw attempt to place these fragments within a broader context: to derive a glimpse of what a 'bigger picture' or 'broader conversation' might look/sound like. O f what the internet means in daily institutional and individual lives, of the processes of retailing and consumption mediated by computer, modem, wires, trucks, and above all people. A s Leyshon and Pollard (2000: 209) note, accounts of technological change from an organisational perspective often assume the 'outcomes somehow reflect the "survival o f the fittest", the achievement of an "optimal" solution. It is more accurate to see industrial outcomes as products of a kind o f "muddling through'". Taking this to heart, this thesis aims to represent in text the "mudciling through" of the studied organisations and their strategies, and of the responses o f individuals — employees, customers — to this. T o accomplish the immodest ambition of glimpsing the "bigger picture", I have set a variety of objectives: a checklist i f you will for the organisation of the work that follows. Quite simply, this could be expressed as: "Why internet, why groceries, why now, so what?", with the underlying geographical question o f "why here?". For two of the organisations studied, the answer to the question "why groceries" is quite simple: they are grocers, food is what they do. "Why internet" is a much more difficult question to analyse, and brings with it a raft o f institutional and personal history, sense of identity, career progression, curiosity, and blind faith. For the other case study firms groceries were not necessarily a "natural" choice for an online business, and the combination of "online" and "grocery" evolved through trial and error, strategy and reaction. For the employees that work within these organisations, and for their customers, there are numerous reasons for their interest in and support o f "online" and "grocery" (as units and union). The "why now" is bound up in organisational and personal stories that I attempt to portray through the use of their own voices; the "so what" is addressed in Chapter 7. T o answer these questions, I have structured the work around retail institutions, customers, and connections between and beyond the them.  30  1.5 Structure of the Thesis While this thesis and the research that informs it could be categorised as 'retail geography', as the above discussion showed it is informed and driven by broader debates and questions than would normally fall under this rubric. With its interest in questions of relations among stakeholders, for example, it could be seen as a working example of Gibson-Graham's (1996) call to 'get out o f this capitalist place', in describing alternative means of organising and allocating resources other than reductionist economic models based on (shareholder-oriented) profit maximisation. It is also informed by, but not strictly beholden to, Storper's (1997) ideas o f untraded interdependend.es: although the emphasis here is not on a particular region and its socio-economic (relational) assets per se, I am attracted to his call for a 'social science of conventions' (p. 305), in this case a study of the conventions that must be acquired to inculcate a habit o f online shopping. The thesis is also to a large extent about how organisations organise and innovate. While this line o f research has guided Shackleton's (1998) work on strategy wifiiin Sainsbury. as it expanded into the United States, retail geography (even in its 'new' form) is rather more concerned about strategies of location rather than operation (Wrigley, 1997, for example, focuses on the required divestment of stores in the U S following a takeover due to regional monopoly). Further, the thesis has a distinct interest in food and its geographies, in issues of regulation (food safety and organic certification), local development and moral imperatives to 'do good' by society and nature; and for logistics. Lasdy, the thesis takes great interest in comparing the perspectives o f a range o f actors involved in and impacted by online food retailing, comparing in particular the views of managers and customers, but also of drivers, pickers and programmers (in a similar effort to that o f Suryanata, 2002, who more explicidy used an actor-network approach). It should therefore be clear that this thesis is rather eclectic in its sources of inspiration and precedence, drawing on literatures from geography, marketing and management, psychology, sociology, philosophy and anthropology.  14  It would not be possible to synthesise all these  influences into one coherent literature review chapter that forms the basis for the empirical work that follows, although I have summarised above the core "mforming'' bodies of work that form the context for the research that follows. Instead, the relevant literature is introduced as it is  This is not to presume that other economic geography writings do not draw on similarly broad (or broader) literatures, merely that I do not wish to oversimplify these works through a forced synthesis.  31  needed, in a thematic structure, and is therefore spread over every chapter in the thesis. Attention now turns to how this is done. Chapter 2 introduces the case study firms that lie at the core of the relations and actions that are studied in this thesis. It places the firms in a rough evolutionary context, by briefly exploring the biographies of the firms of interest and their contemporaries (and competitors). The chapter also explains the methods used in collecting information about these firms. It explains why case studies were used, and why interviews were the method of choice in gathering "thick descriptions". Chapter 2 presents details on who were interviewed, why and how, and also reflexively explores how the interviewer (me) is implicated in the types and extent of knowledge that was gained in interview. Attempting to research and present the voices o f numerous actors involved in the process and result o f online food retailing requires an organisational framework to give the thesis coherency. Chapter 3 introduces the framework used here, that of stakeholder analysis, and compares this to other contemporary theories of firms largely falling under the rubric of "relational economic geography" (Barthelt and Gluckler, 2002; Yeung, 2002). The "range of actors" described above fall into indiscrete stakeholder groups, and it was my objective while undertaking the research to interview as many (overlapping) stakeholder representatives as I could, to gain a multi-perspectival 'holistic' view of the firm, its purpose, functioning, and effects. Chapter 3 therefore describes how the stakeholder model can be operationalised in research and in firm strategy and purpose, and this understanding of what the firm is and who it is comprised of, and constituted for, underlies the remainder of the thesis. Because this approach is uncommon in geography (though see Hughes, 2001a, 2001b for an exception), links are made in Chapter 3 between the stakeholder approach and other more familiar techniques in unpacking actor relationships, mcluding actornetwork theory and the commodity chain approach. This thesis does not try to "apply" a particular model or theory to the situation at hand; it is instead an attempt to drink through and unpack the problems of identity, action and relation that organisations, individuals within those organisations, and individuals affected by those organisations face. It uses the stakeholder model as an platform on which to base the empirical findings and theoretical abstractions that follow, rather than a theory to prove or disprove. Chapter 4 is the first of the three empirically-informed chapters, and grapples with issues of innovation, corporate culture and strategy. It asks a series of questions about why companies innovate, who controls and influences the process, and how strategy is formulated, communicated 32  and enacted. Building on the stakeholder framework explained in Chapter 3, and the firm biographies of Chapter 2, Chapter 4 centres discussion on the inside o f the firm, in an attempt to understand what the firm is, how it works, and who it is for. The chapter also considers stakeholder perceptions of other stakeholder groups: the perceptions o f those inside the firm on those who are also inside (managers on their staff, and the staff on their managers, for example) and on those who transgress or are outside the firm's borders (on customers and contractees, for example). Discussion in Chapter 5 turns to customer reactions to the internet and online grocery shopping, and their actions (explicit or otherwise) in shaping the firm's strategies. It explores how customers have been represented in the retail and cyberspatial literatures. Customers of the case study companies are profiled using available statistics, with these profiles compared to that o f general internet users. Discussion then turns to analysing how customers get online, navigate the virtual stores, and interact with the online grocers and their employees. If Chapter 4 can be conceptualised as taking an analytical and discursive position from within the firm, considering contact witiiin and beyond the firm, Chapter 5 re-sites this analysis and discussion beyond the boundary of the firm, and considers how customer relationships penetrate the firm and its agents. A firm will not survive i f their products, no matter how virtual (digitised) or real (atomised), cannot reach its customers at the right place, at the right time, in the right quantity and quality, and at the right price. Logistics lie at the heart of these connections between retailer and customer, and therefore the last empirical chapter aims to show how the retailer and customer are intertwined strategically, relationally, and in future prosperity. Chapter 6 explores how retail firms attempt to manage their internal and external spaces and times in order to get their products to the customer. The chapter argues that firms cannot attempt to solve problems in the management of time and space equally at the same time: solving problems i n space tends to create problems in time, and vice versa, whether internal to the firm's store or warehouse, or externally in reaching out to customers. The chapter investigates how online retailers have managed these problems, how these problems have changed over time (and differ in the spaces of Auckland and Vancouver), and how the drivers and customers o f these organisations have reacted to and helped guide these innovative processes. Finally, Chapter 7 brings together the various findings of this thesis, and searches for the generalised conclusions that can be drawn from the empirical specificity, without destroying or erasing the contextualised and nuanced richness and embeddedness of this knowledge (Amin and 33  Thrift, 2000; cf Rodriguez-Pose, 2001). It also outlines where I see this research continuing and being built upon, as this thesis is by no means the final word on the nexus of retailing, consumption, food and electronic commerce.  34  2 Online Food Retailing: Research Context Surf: The act of moving around the Web using links until you've forgotten what you were looking for in thefirstplace. (MacFarlane et al, 2001) "The problem is the gap, between us on the map, and there's no easy way to reconcile it." (The Mutton Birds, 1994) Acknowledgement: parts of this chapter were originally published in Murphy (2003a)  2.1  Introduction  Chapter 1 included a brief review of cyberspatial and e-commerce literatures. It indicated that a stream of work is emerging on the internet and society, communities, children, consumption, and technology sociologies and histories, as well as the spatialities of technologies and infrastructure. What was lacking in the review, and the literature more generally, is substantial work on electronically-mediated spaces of retailing and consumption. Typically when this topic is addressed it receives only limited discussion — Kenney and Curry (2001: 49), Liebowitz (2002) and Cassidy (2002) all devote a few lines or paragraphs to the case of Webvan, a US-based online grocer, but even here a wider survey of the 'industry', and the actors involved in it, is decidedly lacking; until very recently, however, discussion of the "e-spaces" o f retailing and consumption has been absent altogether. This chapter aims to provide the necessary background to the development of online grocery shopping by exploring the biographies of some of the institutions that were central to it. It also provides the context for the case study firms on which this research and thesis is based, and a justification of the case-study approach.  2.2 Historical and international  context  Shopping from the home has a venerable heritage, dating back to the catalogue shopping foundations of Sears, Roebuck' and Montgomery Ward in the late 1800s and the U K ' s Kays (from the 1880s) and Great Universal Stores and Littlewoods in the 1930s (Davies, 1976; Coopey, O'Connel and Porter, 1999). Home delivery continued in micro-scale throughout the 20  th  century  by local stores with "bicycle boys" for customers ordering via the telephone (Figure 2.1, Cope, 1996). Some stores continue to this day with the practice of delivering to less-abled customers' homes the products which the customer has picked from the shelf to save the effort of bringing  The Sears Roebuck catalogue had over 10,000 items available, including houses and gravestones (Wrigley and Lowe, 2002: 235)  35  the goods to and beyond the doorstep. Remote ordering, where it occurred, was often locally organised on an ad-hoc basis, and was often either inherendy unresponsive to changing customer needs (by supplying a standing order o f staples, for example) or expensive to operate (due to the time required by store staff to takle and process orders, largely via the telephone). Fax ordering did make an appearance in some stores, but using fax ordering is complicated both for customer and store, due to the sheer number o f goods that a modern supermarket carries - anywhere from 10,000 to 50,000 Stock Keeping Units (SKUs). Local home delivery operations for a variety of 2  products (particularly milk, newspapers, hot food) still exist in many cities around the world, and in some cases suppliers o f these products are attempting to expand into other product categories, and to online ordering. Nonetheless, what is being attempted in food retail is a much larger scale operation than has been the case until now.  Figure 2.1: Sainsbury Home Delivery, 1913  Source: J Sainsbury Virtual Museum <www.j-sainsbury.co.uk/museum>  Computer-mediated ordering of groceries, although necessarily more recent, also predates the internet. The increasing penetration into home spaces o f computers with modems led to a rapid increase in the number of companies offering home delivery of groceries, first via more expensive (for both retailer and consumer) direct dial-up connections, and by the mid-1990s via the world wide web. A  1985 special issue of Environment and Planning B on  computer-mediated  communication (Bennison, 1985; Davies, 1985; Guy, 1985a, 1985b) outlined trials of "armchair shopping" in the U K using British Telecom's Prestel service, a combination o f dumb terminals, modems and dedicated telephone lines for the delivery o f data from 'information providers', including food retailers. The services, although foreshadowing to a surprising extent both the  Stock Keeping Units are the basic identifying tag for products and their packaging variations.  36  objectives and terminology of the world wide web, were never revolutionary in impact, attracting far fewer than the expected number of domestic customers (Bennison, 1985). A concurrent trial was held for the accessibility-challenged and other disadvantaged consumers (mainly the elderly and disabled) in the form of the Gateshead Shopping and Information Service, launched in 1980 using a Tesco supermarket for the limited-line groceries component (Davies, 1985). The service used a variety of connection methods: computer terminals placed in libraries and community centres, largely operated by an assistant; social services personnel who contacted the housebound by telephone for their orders; and retirement homes. While also using the Prestel service as part o f the communications infrastructure, the Gateshead system had a community focus, encouraging residents to meet and interact with each other, and used all-local suppliers (mcluding a chemist, bakery, dry cleaner and clothing store), all in or around Newcastle city centre. The service was deemed much more successful than general Prestel trials, partly because of free delivery and infrastructure provision. After the conclusion o f the trials, despite the bmgeoning private and inhome ownership o f computers, attention switched away from computer-mediated retailing to . equally short-lived experiments in interactive television. Tesco was sufficiendy unimpressed at the time of these trials to comment that it did 'not expect that there will be a major shift over to home shopping on the Internet' (Reynolds, 1997: 35). The United States has similarly had a number o f remote shopping food retailers for some time. San Francisco's Grocery Express began in 1984, using direct home computer links (Howard, 1995). Online retailer Peapod has operated since 1989, starling with a phone and fax ordering system, and by 1998 had expanded from its original base in Chicago to San Francisco, Houston, Dallas, Boston, Adanta and Austin, Texas (Schwartz, 1999). Its original business model was to pick groceries from the shelves of a grocer of the customer's choosing, which limited its income generation to the customer's willingness to pay for delivery (via Peapod's own distribution system) and a mark-up on groceries to cover the cost of picking (Stone, 1999). It moved to a proprietary warehouse fulfilment system in 1998, but struggled to attract sufficient customer growth and healthy margins (despite boasting customer numbers of around 100,000), and in A p r i l 2000 was taken over by the global retailer Royal A h o l d (now the fourth-largest U S food retailer) for US$73 million (Sandoval, 2000b; Wrigley and Lowe, 2002). Similarly, Streamline began in 1993 with phone/fax ordering and picking from local wholesalers, and then moved to the internet and a warehouse pick. It offered a rather unique service of "unattended delivery", supplying a locked  37  box (partitioned into ambient, chilled and frozen sections) which could be filled late at night (Cope, 1996; Schwartz, 1999; Kamarainen et a l , 2001). With the commercialisation of the internet in 1995, and thus a more reliable and cost-efficient method for customers to communicate direcdy with retailers via computer-mediated channels, the number of online retailers grew substantially — albeit briefly for some. Netgrocer, starting in N e w Jersey in 1995, fulfilled orders from a large warehouse, and used national courier Federal Express to deliver the groceries (Schwartz, 1998; Yamada, 2000). PinkDot (later renamed PDQuick), a Los Angeles-based 11-store retailer, offered a phone order delivery service from 1987, and at the turn of the millenium was planning a nation-wide service (Sandoval, 2000a). Table 2.1 summarises some of the key experiments i n North America in online grocery provision (in the "business to consumer" (B2C) market, as it came to be known — see Reynolds (2000) for an overview) — which received much of the attention of the media during the late 1990s.  Table 2.1: Selected North American Online Grocers US  Grocer  Location  St.iitu|) Date St.itus  Model (cli.inyes)  Grocery Express - • SanFraneisco • W a reh ouse J i f l l l S l l l i Y i ; i Y ,111:98111 :;Oipseci:;i980sjklj 1 Tried merger with Kozmo; bought by PinkDot/PDQuick Los Angeles . ';. • • Own-store pjck l l l i . i i f S i11; IWHygunO^ Chicago + 7 markets Other sto res to wa rehouse to ownTaken over by Royal Ahold, April 2000 stores Iri^SiSi^-r " Streamline Boston Wholesalers to warehouse; -CJosediHoTO unattended box :  Mil?/:  :  ;  Netgrocer HorneGrocer KOZmO: . ;v  WobHouse Club  11"' ifi W K ^ y r i O u t h  .GroceryWorks  New Jersey- US-wide Warehouse via Fed E x .. :  Seattle + 7 markets : New York; lift + 2 markets East coast  'Warehouse  1-. 11,' :  v;  :;:  1995  • St|lljoperatingJ ?:;;  1998  ;Bought by W e b y a n U u n f 2M  Other stores and contract delivery  1999  Auctions (Priceline subsidiary)  -1.  1999  lYY-  1999  San Francisco + 10 markets  Warehouse pick  San Diego, Los Angeles  Picks from Stater Brothers stores in Southern California, PinkDot stoes in Los Angelesj/  1999  Warehouse (cross-docking) to iSafeway stores 1&U:M Zf$^S  2000  Dallas  "  p:  :  :  Closed April 2001 Closed October 2000 ' Bought Homegrocer; closed July 2001 " ' '  '  Bought PDQuick in May 2001; acquired Webvan's customer ..tdatabase November 2001 ia-f:^n |:  Bought by Safeway (US) in April 2000, in joint venture with Tesco (UK)  C.inntl.i Peachtree Network  T ^~il-V:|ii;:E  : . : : ) . ; ; : ' : J : i k j l l t y . ' "iii i l l ; 1 i i : : l = Y E.iii Si h l i i l i 1: i i Closed November 2001 Ml "i :  Montreal  .;• Small Potatooa Urban Vancouver Delivery (SPUD) Stong's Market  Vancouver  Grocery Gateway Quick  Toronto .fi; Y Vancouver  .  Infornedi aiy (provided we b s e ryi c e to small retailers in Canada & US) Warehouse pick (organic box, unattended) 1  ; • • 1;Single-store pick Warehouse pick Warehouse pick  -,7:  :  1997  :  Took over 9 other organic home deliverers; still operating  1997  1998 : Part of Peachtree Network until Nov. : 2001; still ope rating f, |1 1999 . Still operating 'iYfy-; i . . 2000 i C id's ed Februan/20 0311;;',i. :  :  Source: Murphy (2003a: 1179)  38  :  These experiments can be separated into two broad types (though see Reynolds (2000) for a more nuanced view): the "store pick", where an existing (self-owned or other) store infrastructure is used to pick orders received from customers via telecommunications or internet networks; and the "warehouse pick", where an existing regional distribution centre or purpose-built distribution centre (DC) is used. Most North American experiments until 2001 followed the warehouse-pick model, while retailers in other nations focussed more on the store.  Table 2.2: Venture Capital Funding, Jan 2000  The Plavers. The Capital, and The Proaress Grocer  Webvan Peapod Streamline Homegrocer Shoplink Netgrocer Homeruns* HomeXpress** Grocery Gatewav**  Private  Capital Public  $396.0 $32.0 $40.0 $162.0 NA $12.0 $15.0 $0.1 $37.0  $375.0 $64.0 $45.0 $250.0 NA $0.0 $100.0 $1.3 $0.0  Total  $771.0 $96.0 $85.0 $412.0 $12.0 $115.0 $37.0  "Value based on recent sale  1999E Sales  Market Cao  $10.0 $100.0 $15.0 $20.0 $15.0 NA  $4,181.0 $175.9 $167.7 NA NA $60.0 $130.0 $7.3 NA  - -y:NpM^:  All figures in USSmln unless noted  "$Cdn Caicco (2000: 5)  One of the more unusual offerings was that of WebHouse Club, a spin-off of "name your price" virtual retailer Priceline, which promised to pass on individual shopper requests for goods at the customer's stated price and quantity to retailers, who would then issue vouchers to those whom it agreed to supply (Kane, 1999; see Priluck, 2001 for a detailed review). It went bankrupt in 2000 (Table 2.1). Many o f the online grocers in the United States went public, and/or raised substantial sums o f private capital from venture funds (Table 2.2), and this resulted in significant media 3  coverage and post -dot-com bubble analysis. Going public in November 1999, Webvan was one o f the largest Initial Public Offerings of that banner year for technology stocks, second only to Barnes and Noble.com, raising nearly $800 million in public and private finance. Webvan used its  Exact estimates of funds raised and market capitalisation (which is time-specific as it relates to current stock price) vary considerably among sources: Webvan has been quoted as raising between $700m and $lb, and its peak market capitalisation at anywhere from $4b to $8b, depending on whether the end-of-day price or maximum price is used.  39  significant Initial Public Offering (IPO) capital to purchase (via stock transfer) a prime competitor with equal nation-wide yearnings: Homegrocer of Seattle (Sandoval, 2000a). Webvan's market capitalisation then slumped from more than US$5 billion following I P O in November 1999 to a penny-stock valuation of under $300 million by the end o f 2000, less than a quarter of the initial cost of acquiring Homegrocer (Figure 2.2). By early 2001, with a price consistendy under US$1, it was threatened with deusting from the N A S D A Q exchange. It was in the process of a reverse stock split in order to boost its stock price when the company declared bankruptcy and closed in early July 2001 (more on this in Chapter 4).  Figure 2.2: Webvan Stock Value, $US Webvan daily close  Data.source: http://chartvahoo.com  Peapod, one of the longest-running computer-based home-delivery organisations, was acquired by the global retailer Royal A h o l d (Wrigley, 2002a), after struggling with slow growth and a low market capitalisation (Table 2.2). Many of the large retailers (particularly Safeway and Albertsons of the US, and Safeway and Sainsbury of the U K ) were slow to expand beyond tentative trials, but have increased their web presence as the "pure-play" startups folded. Safeway, the third-largest 4  "Pure-play" is generally taken to mean "unencumbered by non-intemet activities" and sunk retail capital. This includes internet-only retailers who have no store network, such as Amazon.com (although they may have extensive physical distribution networks). At the turn of the millennium pure-plays were seen as epitomising the  40  US food retailer, acquired Dallas-based online grocer GroceryWorks in 2000 and gradually rolled out a national service via subsidiary company Safeway.com, in which the U K supermarket chain Tesco acquired a 35 per cent stake (Olsen, 2000; Wrigley, Lowe and Currah, 2002). GroceryWorks was initially unique in outsourcing the provision of fresh produce, seafood and meat to local suppliers via a linked information technology system, collecting and distributing (through "crossdocking") these parts o f an order twice a day. The cost o f this system relative to the volume o f orders received caught up with the experiment, however, and in 2001 Tesco/Safeway.com closed the cross-docking warehouse and redirected fulfilment to Safeway stores using Tesco's U K derived computer technology and in-store picking methods (Table 2.3; Wrigley, Lowe and Currah, 2002; Murphy, 2003a; see Chapter 6 for more details).  Table 2.3: Selected online grocer sales in 2000 Grocer  Model  Sales, 2000, $US  store  $305m  warehouse and store  $168m  warehouse  $123m  store and warehouse  $61m  was warehouse, now store  |59m  was warehouse, now store  $51m  warehouse  $49m  Albertson's, US  store  $35m  Iceland, UK  store  |32m  Tesco, U K  a  Royal Ahold, Netherlands and US Webvan, US  c  J Sainsbury, UK GroceryWorks, US  d  Asda Wal-Mart, UK Homeruns, US  e  f  b  a Tesco's online sales for the financial year Mar 2001-Feb 2002 were £356m (US$51 lm), and £447m (US$681 m) for the year ended Feb. 2003 (Tesco 2002, 2003). Tesco is the only "bricks and clicks" grocery retailer to consistently and publicly reveal its e-commerce sales b Royal Ahold purchased Peapod during 1999 and added in-store picking from Royal Ahold's US chains to Peapod's warehouse model. c Webvan closed in July 2001; it acquired Homegrocer tn 2000. Caicco (2001: 3) estimates total sales (including Homegrocer) at $300m in the same year d GroceryWorks was acquired by Safeway US in 2000 to form Safeway.com, in which Tesco invested as a joint venture in 2001 e Asda closed its London warehouse operation in early 2002, and picks from stores around the UK f Homeruns closed down in July 2001  Source: Murphy (2003a: 1181); adapted from Finch (2001b)  modern, progressive and nimble, compared to the staid and conservative "bricks and mortar" organisations (Economist, 1999). "  41  In Canada the leading online grocer is Ontario- and warehouse-based Grocery Gateway. It began operations in 1999 in Toronto (Caicco, 2000), spreading to Hamilton and Kitchener in 2001 (Drexhage, 2001). Other major players include the I G A franchise chain in Quebec (but not its affiliates elsewhere in Canada), and independent grocers through the Peachtree Network, which spread across Canada to include grocers in most provinces, including Stong's Markets in Vancouver, as well as several grocers in the United States (Figure 2.3).  Figure 2.3 : Stongs Home Page: Peachtree Markets (November 2001) www.stongs com  Welcome Member's Entrance: Username:  to  Stong's  Shop Online New user?(click here) Registered user? (Click here)  LINKS  BECOME AN AFFILIATE ABOUT US  New to online shopping? It's easy.Simply:  1.  SHOP ONLINE THIS WEEK'S SPECIALS STONGS ONLINE SURVEY CUSTOMER SERVICE  Register for free, by completing the  HOME RECIPES/COOKING CONTACT US  form at: http://www.peachtreerietwork.com/  2.  Start shopping now!  Do it once and we know you'll love it because it's so convenient! Be sure to save your items into our handy quicklists This way, when it's time to shop again, most of your items are ready to go. give it a try! Click here: mew user link)  Home | Shop online | This week's specials | Survey C u s t o m e r service | Links | Become an affiliate About us | Recipes Cooking  The Peachtree Network was supplied and supported by Stria Communications, which developed web hosting software to communicate the product range, pricing and product image information for the thousands o f products provided by the various retailers through a common platform. The  42  Network handled all customer orders via the web, emailing back to the retailer an order list to be picked in-store and despatched. This concentration of software development into the hands o f an "infomediary" (Currah, 2002; Murphy, 2002; Wrigley, Lowe and Currah, 2002), in return for a percentage of turnover, made getting online for the small retailer a relatively quick and painless process. The software was not particularly sophisticated or cutting edge, however, and when Peachtree went into receivership in 2001, all small retailers lost their web interface. Stong's subsequendy re-launched after contracting a local web developer to construct a new interface. The experiments in North America were chasing a tantalising dream: a fortune to be obtained by becoming the "category killer" o f online grocery, with forecast revenues in the billions. Datamonitor, a research agency, predicted rapid growth i n the online grocery market to reach US$27b by 2005, equivalent to 4.8% of the "traditional" U S grocery market (Datamonitor, 2001); Palmer (2000) expected sales o f US$34b by 2002. Even more reticent forecasters (Caicco, 2001) were expecting 100-fold growth over at most a six-year period (Table 2.4), although these forecasts were downgraded as the dot-com market slumped.  Table 2.4: North American Online Grocery Market Predictions Internet Sales  Original Prediction: April 2000 2000 2001E 2002E  2003E  2004E  4700 420  7000 650  12000 800  Revised Prediction: Jan 2001 2000 2001E 2002E  2003E  2004E  3500 300  5000 500  2002E  2003E  2004E  0.43 0.25  0.67 0.49  0.93 0.79  1998  1999  120 5  300 15  1998  1999  120 5  300 15  800 25  1500 90  2200 150  % all Grocery sales  1998  1999  2000  2001E  US Canada  0.03 0.01  0.06 0.03  0.17 0.04  0.30 0.15  US(US$m) Canada (Cdn$m)  Internet Sales U S (US$m). Canada (Cdn$m)  1350 65  2500 180  Source: Murphy (2003a: 1180), adapted from Caicco (2001: 14)  Despite these rapid growth rates, however, Caicco did not expect online grocery to amount to more than one percent of the total grocery market in either Canada or the United States. The profitability and sustainability of this select niche remained elusive with the spectacular failures o f WebHouse Club and Webvan. Webvan failed due to excessive investment infrastructure costs of more than $30 million per fulfilment warehouse (see Murphy, 2003a, 2003b for an overview), while the Priceline subsidiary WebHouse Club had a bizarre business model of negotiating deals on single grocery items with manufacturers based on aggregated individual buy bids from potential customers (see Priluck, 2001; Wrigley, Lowe and Currah, 2002). The focus o f media attention on 43  the startup and subsequent failure of these operations has not been helpful to remaining businesses: Webvan hurt us in two ways: When they were first on the scene, everybody was sure they were going to win. Then when they shut down, everybody said we couldn't make it either. (Fritz, 2001)  Efforts in the U K continued to be led by Tesco. Despite its earlier scepticism Tesco launched a 5  C D - R O M based service in 1996, but this experienced slow growth due to the relatively low penetration of computers into U K homes at the time (particularly with modems and C D - R O M s ) , and the call cost of direct-dialling in orders. Using C D - R O M s did have the advantage o f a rich graphical interface, but customers had to download an update of stock availability and pricing from Tesco, which added to the cost and time. In 1999 it switched to a web-based interface, and growth thereafter was rapid, with more than a million registered customers by 2002, sales of over £350m, and the ability to serve over 95% o f the U K population. During 2001-02 Tesco expanded its online operations to Ireland, South Korea and the United States (with Safeway), making it the first multinational online grocer. Tesco's competing supermarkets were slow to respond to the challenge, despite (or perhaps because of) Tesco experiencing a rapid increase in general sales that saw it overtake J Sainsbury to become the largest chain in the U K . Sainsbury and Wal*Mart-owned A S D A both opened warehouse-based operations in London, but also experienced initial slow growth, and expanded through the rest o f the country via store-based operations ( A S D A closed its London warehouse in 2001, while Sainsbury struggled on). Waitrose, the supermarket arm of department store chain John Lewis Partnership, also entered the fray, originally through Waitrose-at-Work (deliveries to workplaces rather than the home), but by 2001 had expanded to home deliveries and created a second brand, Ocado, with which to do this. The U K chain Safeway (unrelated to the North American chain following a sale to Argyll in the 1980s) also began store-based operations at the turn o f the millennium, having earlier expressed disdain about the sustainability o f online operations and experimented with in-store customer self-scanning equipment.  A more thorough contextualising of Tesco's place in the food retailing sector and world follows in the next section.  44  2.3 Case Studies 2.3.1 Justification for Case Studies Approach A s mentioned in Chapter 1, statistics and generaliseable information regarding the "new economy" and electronic commerce are notable by their absence. The statistical agencies o f the United States, United Kingdom, Canada and N e w Zealand are all attempting to measure the size and importance of wholesale (B2B) and retail (B2C) sales conducted through internet channels, but traditional survey and census instruments have proved poor at doing so. This is pardy because o f the inherent difficulties in defining as well as measuring broad innovations in technology, business processes and scope of operations that span traditional industries and regions that have been the container for statistical information collection and analysis. The agencies have been more successful at measuring the penetration of computers and internet connections into homes and businesses, largely by adding questions on these to already existing industrial, labour force or household surveys and census questionnaires. These have shown a rapid increase in the proportion of households and businesses with internet-connections (Figures 1.6 and 1.7 from the Introduction), and indications from research on the so-called "digital divide" that usage tends to increase with household income (Figure 2.4; Taylor and Murphy, 2003).  Figure 2.4: US Internet Use with Income Intern* Use 1997,1990, 2000,2091  100 90 80 70 |  I  60 50 40 30 20 10 0 U M H - $111,000- $28,000- $36,000- $60,000- O w S.b.OOO $24,993 SM.S&i $4B,M0 $74,Mtt $76,000  Source: ESA- NTIA (2002: 10)  45  It could be inferred from these trends that richer households are likely to spend more money on online shopping, and that the proportion of retail revenue earned through online channels might increase with a rise in general economic wealth (such as during an economic boom), and as the diffusion of broadband internet access technologies increases. Quantifying these hypotheses has proved fraught with difficulties, however: there are many possible definitions o f what constitutes "electronic commerce", and national (and particularly international) statistics are prone to a definitional "muddle". For example, consider the following scenarios for a potential buyer, let's call him Andrew, of a transcribing machine: 1. Andrew goes to the University bookstore and buys the first machine that falls within his supervisor's expenses budget. H e pays by debit card through an E F T P O S terminal. 2. Andrew consults numerous websites containing reviews of various machines, posted by alleged users (since, unlike computers, cars and game machines, there are no "fan-zines" for transcribers), by the manufacturers, and by consumer testing bodies. H e then buys a machine that looks promising from said bookstore. 3. Andrew consults various online stores for pricing of said machine. He discovers the cheapest are in the United States but either won't ship to Vancouver, or shipping and likely customs charges are excessive, or the need for the machine is urgent and he can't wait for the 14 days for "standard" delivery. He buys the machine from said bookstore. 4. Andrew bids for a promising machine on E-Bay. The website automatically increases his initial low bid by small increments each time another bidder out-bids, up to a maximum. Andrew consults the auction website regularly to ensure his maximum isn't too low, especially in the vital last 10 minutes of the 3-day auction. H e wins the bid, since there are ten other identical machines on offer by others. The seller won't ship to Vancouver, so he persuades a Blainebased American friend to accept the item and bring it up to Vancouver with him. Andrew pays via E-Bay's "Pay Pal" online billing service subsidiary, which debits his credit card and credits the seller's account, which she then uses to purchase other items on E-Bay. 5. Andrew purchases a transcriber through Future Shop's website, but due to a website glitch ends up paying for the item by calling Future Shop's online customer support team and giving them his credit card details. The item is despatched from Toronto via Purolator within two  46  days, and arrives in Vancouver after five days. Since he wasn't home at the time o f delivery, Andrew picks the item up from a local Canada Post agency two days later. A l l o f these scenarios could be defined as involving electronic commerce: electronic funds transfer at point of sale (EFT-POS) for scenarios 1, 2 and 3; information gathering from online sources for 2 to 5; online transactions for 4 and 5. From a customer's perspective the delimiting of "electronic commerce" as a way of determining the extent of the "internet economy" is semantics, since they may use any or all of these configurations on any given day for all sorts o f different goods and services; it would be akin to trying to determine the extent of the "telephone economy" or the "postal economy". From a statistician's perspective it makes a significant difference: what value should be placed on the provision of information for searching online, but where transactions ultimately end up offline in-store? Should all banking transactions be included, since none can be made now without recourse to an internet-related infrastructure of fibre optics, routers and servers? Limiting the size of electronic commerce to purely digital transfers eliminates all of the above scenarios, since ail of them require a grounded infrastructure o f some sort for a physical product (couriers, telephone sales support, Interac E F T - P O S machines...), and may involve several organisations for which internet transactions are a minor part of their business, such as courier companies, P O S equipment suppliers, banks and consumer testing agencies. Thus electronic commerce is very difficult to measure either by tracking transactions, by occupations, or by organisations. Business research consultants such as Datamonitor or Jupiter have used a variety of "metrics" over the past seven years, largely focussing on counting users. First up was counting "hits", the number of HTTP-tagged links that users would activate. This was replaced by "page impressions" when it became obvious that complex pages would self-activate a number o f "hits" to bring in various components such as graphics, advertisements and icons. It was then realised that users o f complex sites with numerous pages were inflating scores, so "unique visitors" became the next big metric. This required the tracking of individual users, usually through their IP address or by placing a "cookie" (identifying tag) on the user's computer, both o f which were subject to problems with dynamic IP addressing (where the user's I D changes each time they log-in) and fears of privacy invasion, which led some users to mask or expunge their cookies. The most precise measures of internet use came from the use o f panel data, whereby a few thousand users would install special software on their computer to track their usage over a given period of time for that computer — but this method could not easily distinguish between multiple users of the same computer, or track the same user with multiple computers. This is still not the  47  most accurate, however, since it excludes the usage patterns o f those not involved in the panel programme (by choice or necessity). In any case, counting customers is not a great measure o f the value o f electronic commerce, merely its ubiquity. For all these reasons and more, national and sector statistics o f the size and importance o f electronic commerce remain deeply flawed, although their availability is improving (as is their accuracy with greater diffusion). For the purposes of this thesis, it was neither feasible nor advisable to investigate the total online sales o f an industry, even in one geographical area, since the number and constitution of participants is quite large and in flux (see Table 2.1 for North America alone). For this reason (among others) I decided to concentrate on a case study investigation of online grocers, with snapshot breadth replaced by longitudinal depth.  2.3.2 Case Study Options In the early days o f investigation (1999) the sample frame of firms available to me was limited numerically and spatially. Being based in Vancouver meant local operators were a natural target, but none of the major grocery chains (Safeway, I G A , Save-on and Costco) were accepting orders online. Only one grocery store (the single-store Stong's based in Dunbar) was operating in this manner, and became the natural first candidate; other candidates were the numerous small organic home delivery companies. Further afield I had access to N e w Zealand and the United Kingdom through family ties. In N e w Zealand one large national supermarket chain, Woolworths Supermarkets, had begun web-based operations in late 1998 and was a definite candidate; in the U K most national chain stores were contemplating the idea, with Tesco a leading experimenter, and I decided to target these i f at all possible (logistically through visits and methodologically through access to personnel and customers). The United States, while a natural option given its history (summarised above), was ruled out on logistics and financial grounds. I first approached Woolworths N e w Zealand (based in Auckland) in December 1998 and again in March 1999, along with Stong's Markets, based in Vancouver in March 1999, while preparing my thesis proposal. Both were keen and approachable, with the chief executive of Stongs and the manager of the Home Shopping (online) division of Woolworths giving me an hour o f their time to discuss research possibilities. Exploratory research began in the United Kingdom in January  48  2000, where I visited the Home Shopping section o f a Tesco store in Winchester, and interviewed 6  the section manager, two "pickers" (the employees who complete the orders) and a driver. This information was very useful, but Tesco themselves were not forthcoming for any wider study, and I had no plans to be back in the U K in order to explore this further. Woolworths remained very welcoming through email contact and telephone calls, and were willing to organise interviews with staff involved in the Home Shopping operation, plus contracted delivery staff and customers. This however was only organised upon my arrival in N e w Zealand in late January 2000 (see following section for details). Interviewing was done during February and March 2000, and further interviews (and re-interviews) were conducted in December 2000, January 2001, and December 2002. Through much of this time I remained in email contact with the managers of the service. U p o n return to Vancouver Stong's Market were initially welcoming. Initial interviews were conducted over the phone in December 1999 (to confirm earlier contact during the proposal process) and in person in February 2001, but after this the enthusiasm o f the store's management for the project distincdy waned and all further contribution by Stong's staff, and access to customers, was refused. In any case in November 2001 Stong's online host and infomediary, Peachtree Networks of Ontario, which until that point had been rapidly expanding across Canada and the United States to encompass more than 13 independent stores and chains (Figure 2.3), went into bankruptcy, bringing Stong's online provision to a sudden halt (Figure 2.5). Stong's replaced its web service with a more limited fax-based service, and by February 2002 was back online with the help of a local developer. Stong's had a policy of not releasing customer information to anyone outside the company (despite the fact that it had to do so in order for the infomediary Peachtree to operate). It would therefore have been difficult to locate online customers to interview without spending inordinate amounts of time in-store, and the number of staff involved in online order fulfilment was limited. This combination of barriers led me to cease pursuing the company as a research subject. Stong's is therefore not counted as one o f the case study firms in this research, although I make mention o f details where pertinent.  Woolworths and Tesco use the same terminology for their e-commerce division; they share many common traits and strategies, as will be explored throughout this thesis, but Woolworths has had virtually no contact with Tesco, formally or informally.  49  In the meantime two replacements were found: Quick.com, and Small Potatoes Urban Delivery, both also of Vancouver. I had been aware o f the efforts o f small, local organic food suppliers for some time, but because of their small size and scope, and limited operational experience, had preferred the once much larger and certainly more established Stong's Markets. Stong's once had a six-store chain with stores in N o r t h Vancouver and West Vancouver, as well as Dunbar and the East E n d , and had been operating since 1931 - a "heritage" it proudly promotes in its logo and advertising (Figure 2.5). It had slowly contracted to a single site in Dunbar as other retail chains increased in strength, but nonetheless had a significant presence in Vancouver and in particular in the local Dunbar community, which is relatively wealthy and well-educated. Figure 2.5: Stong's Tradition and Webstore Closure  Stongs  your home grocer since 1931  Our new site will be up and running very soon, we would like to thank our valued cutomers for their patience and look forward to you enjoying online shopping again. If you have any questions call us at 604.266.5191 ext.1 or E-mail us express@stongs.com or info@stongs.com Source: www.stongs.com, November 2001  O n the other hand, Organics to Y o u , the largest local organics company in Vancouver in 1999, had a mere few years' experience, no large store or warehouse, relatively few customers, and a "rustic" image of "Farmer Fred" (Figure 2.6). During 2000, however, Small Potatoes Urban Delivery (SPUD) had progressively acquired enough of its competitors to become a direct competitor (see Chapter 4 for details). In January 2001 S P U D merged with Organics T o Y o u , and  50  suddenly became a much more influential player in the local organic market. In proving their financial and business acumen, along with entrepreneurial flair, S P U D became an interesting research subject, and I approached the management team in early June 2001. Their response was as enthusiastic as Woolworths', and a full research project was quickly underway, with full access given to their customer database and to staff (see the next section for details).  Figure 2.6: Organics To You Image  Better. Naturally. Source: www.organicstoyou.com, December 30, 1999  Quick.com had begun operations in late 1999, but grew only slowly to the point where they became of interest to the local media. I had become aware of them in late 2000, and contacted the management in February 2001, once it became obvious that Stong's would not cooperate. The management of Quick were willing to discuss certain aspects o f their operation with me - on the condition that I signed a confidentiality agreement — but were more circumspect about allowing me access to their customer list. It was initially felt that Quick's Canadian customers are inured to incessant market research approaches and would not respond favourably to my research or to Quick's sponsorship o f it. Participants were instead provided (with the retailer's permission) by "Associates" of the company, independent sales agents who attract and train potential customers in using the web for grocery shopping, and in turn earn multi-level marketing commissions (on which more in Chapter 4). In actual fact I had no difficulties in convincing those approached to participate (they were already supposed to have given their permission to the associate, but several I spoke to had not even been contacted in advance) and most were very interested in the aims of the research and very happy to be involved. The limit on customer participation was not their enthusiasm, but their availability (it was summer and the Associates were slow in providing names) and time.  51  A t the close of the intensive research period, in August 2001, I" had detailed insider information about five online grocers: Woolworths (Auckland, N Z ) , Quick (Vancouver), S P U D (Vancouver), Stongs (Vancouver) and Tesco (Winchester, U K ) , and secondary information about several others in N e w Zealand, Canada, the United Kingdom and the United States. This thesis draws on all this information, but in particular on the wealth o f material provided by the in-depth interviews with the staff, suppliers and customers o f Woolworths, Quick and S P U D . A l l these interviews (ranging from 20 minutes to 3 hours, with most lasting between 1.25 to 1.75 hours) were recorded, with detailed notes taken (except where impractical, for example while being given a guided tour of a warehouse/store or when in a truck) and most were transcribed by myself and three contracted transcription assistants. Because the thesis draws so heavily on this material, Section 2.4 explores in more detail how and where this information was collected and from whom, while Section 2.3.3 gives more detail on the histories of the case study companies outlined above.  2.3.3 Case Study Company Histories As outlined above there are three companies for which intensive interview-based material is available: Woolworths N e w Zealand, based in Auckland; and Quick and Small Potatoes Urban 7  Delivery of Vancouver. In addition considerable material is available for Tesco o f the United Kingdom (including three interviews), Stong's Markets (two interviews) of Vancouver, and Webvan, formerly of the United States. These companies represent quite different scales and strategies, and a brief summary of their "positionality" in the retail scene is first required.  Tesco, United Kingdom Tesco is a globalising retailer, the largest national food retailer in the United Kingdom. It has store operations in nine nations (largely central Europe and southeast Asia), more than 200,000 employees and in 2001 made more than U K £ l b (US$1.45b) profit, only the second U K retailer in history to reach this mark. The Tesco.com Home Shopping service, begun in earnest in 1999 after C D R O M experiments since 1996, is available in more than 300 stores serving more than 95 percent of the country, and is being exported to Ireland and South Korea (Jones, 2001; Tesco, 2002, 2003; Wrigley, Lowe and Currah, 2002). It provides almost all its more than 30,000 in-store  Quick originally branded itself as Quick.com, when a domain name suffix was commercially advantageous (following the lead ofAmazon.com). As the market cachet of "dot-com" turned into "dot-bomb" or even "dotcon" during 2001 (Cassidy, 2002), Quick dropped the suffix. Both names are used through this document, largely depicting the period to which the name applies.  52  lines through the online store, plus separate dedicated online stores for a wine club, flowers and baby products which it sources from other suppliers. It also claims to stock (virtually, i n its database) most electrical goods and appliances, over a million books, and every C D , D V D and game on current release in the U K , which by 2003 had grown to 25,000 videos, 7,500 D V D s , and 1,800 electrical goods, along with over 800 wines, from its various specialty 'clubs' (Tesco, 2002, 2003; these items are also sourced from other suppliers). Tesco has also invested in Safeway.com, a joint venture with Safeway US resulting from the purchase o f GroceryWorks, and will introduce its Home Shopping technology through U S Safeway stores (at the time of writing, this was mainly in the western states o f Washington and Oregon). Tesco's U K online sales exceeded £350m (US$510m) during 2001 from more than 750,000 registered customers (85,000 deliveries per week), making it the world's largest online-grocer, and grew to £ 4 4 7 m (US$681m) during 2002, from more than one million registered customers (110,000 deliveries per week) (Figure 2.7; see Table 2.3 for comparable data for 2000).  Figure 2.7: Tesco sales, 1999-2003  500 450  '"mi  400  if  350 V  300  "lit  )  250  1  !|  t  200  " fi X  150 100  I j<''i">iil<  50  , i4':  i  0 1999-2000  2000-1  2001-2  HiijV  i  4  2002-3  Source: Tesco (2001, 2002, 2003), in G B P  W e b v a n , U n i t e d States The star o f the e-grocery and I P O show in the United States is undeniably Webvan. Formed i n 1996 by the founder o f Borders book chain, Louis Borders, and taking online orders since 1998, Webvan first rose to prominence with the appointment o f George Shaheen (previously head o f  53  Andersen Consulting), in late 1999, just prior to its Initial Public Offering, which was one o f the largest in that banner year. Shaheen's interest might also have been prompted by a favourable Andersen Consulting report on Electronic Grocery Shopping (Kutz, 1999). Webvan's mission was to automate the grocery business, by picking from large warehouses in edge-of-town locations. A t 300,000 square feet these "dedicated picking centres" are three times the size o f the largest supermarkets, and would hold "miles" of conveyor belts and computer-controlled rotating racks that would literally bring the goods to the picker (see for example Perman, 2000; Cassidy, 2002). Webvan's premise was simple: reduce "touches" in the supply chain. By using a warehouse rather than an existing "bricks and mortar" store, products could be organised so that stocking and picking could be done efficiendy, and with computer assistance. Its warehouses, at 330,000 square feet , cost between $25-40 million to build (Murphy, 2003b; Wrigley, Lowe and Currah, 2002). 9  Webvan claimed its 900 workers ought to be able to process 'nearly 10 times the productivity of a traditional "shopper" wheeling a cart through a store or warehouse' (Anders, 1999). It boasted a product range of up to 50,000 items, from the spectacular (live lobster) to the mundane (stamps), and in later months branched into photo developing deals. Webvan figured each of its 70 drivers could handle about 20 deliveries a day, which at an average order size o f around US$100 made a total daily turnover of $140,000, or $50m per year per distribution centre. The theoretical capacity of each warehouse was reputed to be 8,000 orders, which would mean revenues of $300m (Figure 2.8) (Cassidy, 2002).  Woolworths N e w Zealand Woolworths Supermarkets N e w Zealand was established by an Australian grocer in the 1920s as part o f an Australasian grocery chain, and was never related to the eponymous F W Woolworth of the United States or Woolworth's o f the U K . After several changes in ownership over the years (it parted company with Woolworths Australia decades ago), Woolworths became a subsidiary of  George Shaheen is credited with dramatically expanding the size of Andersen Consulting, and as CEO would have inherited substantial shares in the renamed demutualised firm Accenture. He is held responsible for insisting on the permanent separation of the consulting and auditing/consulting arms, for which the consultants are no doubt relieved given the subsequent implosion of Arthur Andersen. His decision to join Webvan in 1999 was seen as a vindication of pure-play investment, and a desire on Shaheen's part to "walk" the consultant "talk": facing 'a question all Andersen lifers eventually do: How good are you, really?' (Perman, 2000) By (indirect) comparison, large supermarkets are in the order of 40,000 square feet; hypermarkets are anywhere from 50,000 to 160,000 square feet and Wal*Mart's largest supercentres are just over 200,000 square feet (Tesco, 2001; Wrigley and Lowe, 2002). Of UK grocer J Salisbury's eight regional warehouses, each supporting hundreds of superstores, only five were significantly larger than Webvan's.  54  H o n g Kong-based Dairy Farm International (DFI) in the 1990s. D F I has more than 800 supermarkets and eight supermarket brands in seven Asian nations (Hong K o n g , Taiwan, China, India, Malaysia, Singapore and Indonesia) under various fascias, and Australia, under the Franklins brand.  Figure 2.8: Webvan sales and losses, 1999-2001  $m  Webvan Revenue and Loss per Quarter  100 50  Revenue Loss  0 -50 \  -100  \ \ \  -150 -200 -250  Mar 99 Jun99 Sep 99 Dec 99 Mar 00 JunOO Sep 00 Dec 00 Mar 01 Source: Webvan quarterly financial statements (10Q) to N A S D A Q (ww.w.nasdaq.com), in $USm  By the time of intensive interviewing in 2000 the N e w Zealand group contributed more than 10 percent to DFI's global trade. Woolworths was a nation-wide supermarket chain with 83 stores split among three retail brands or fascias (Woolworths, B i g Fresh and Price Chopper), the diirdlargest firm in a highly concentrated low-margin market. It first offered an online service in May 1996, initially via C D R O M (which held the catalogue of products and the software to run it, and required a modem to transmit orders and download price updates), switching to a web browserbased interface in November 1998 (Murphy, 2002; Davis & Buchanan-Oliver, 1999, 2000; see Chapter 4 for details on this change). It then expanded this pilot to same-day delivery witiiin the three largest cities, Auckland and Wellington in the North Island, and Christchurch in the South Island, and progressive expansion thereafter (Figure 2.9).  55  Figure 2.9: Woolworths Delivery Areas, 2000 and 2002 a: End-2000  b :  End-2002  North Island Auckland Map Key:  ^ Great Barrier Is. North Island Map Key:  Auckland  11! North Island Wide  Auckland Rural  No delivery  V I) •  Great Barrier Auckland Hamilton  • • H  New Plymouth Wellington Kaprti Coast  a  Auckland  H  East Auckland Rural  •  North Shore Rural  •  South Auckland Rural  •  Wast Auckland Rural  North Island Wide  New Plymouth  Wellington  South Island  °°o  ° South Island Map Key: III South Island Wide  Christchurch South Island Rural  O  •  Christchurch  •  Christchurch North  W  Christchurch South  South Island Provincial Town  Source: Murphy, 2002: 52; from website December 2000  Source: adapted from Woolworths' website, November 2002  In June 2002 Woolworths New Zealand was acquired by the Australian company Foodland, also owner of the second-largest supermarket group in New Zealand, Progressive Enterprises. The takeover culminated an acrimonious and protracted year-long batde between Foodstuffs, the largest New Zealand supermarket chain, and Foodland/Progressive Enterprises, involving all levels o f the New Zealand court system (including the Privy Council o f the U K ) . The deal was ultimately given Commerce Commission approval, requiring only minor divestment. Woolworths' three supermarket fascias were added to Progressive's (Foodtown, Three Guys and Countdown), presumably with some rationalisation to occur in the future. Online operations continue under the Woolworths brand, as Progressive was yet to commence, and in fact had been publicly disdainful of the project, particularly given its minor and unprofitable foray through TheSupermarket.co.nz (on which, more in Chapter 4). Woolworths offers more than 15,000 product lines (SKUs) through its online operations. The entire supermarket operations of Woolworths amounted to N Z $ 6 0 0 m (US$260m) in 2001, with the online operations a fraction o f that figure. Woolworths became increasingly careful about  56  disclosing sales information and customer numbers of its online division, particularly during and after the merger negotiation. Unlike with Tesco this information was not made publicly available through financial reports or statements as the company is not publicly listed. When I first interviewed the manager of the HomeShopping unit in March 1999, Woolworths had just re-launched into Auckland after settling a public dispute with its previous technology provider, The Great N e w Zealand Shopping Mall. A t the time it was operating in Wellington (relaunched November 1998), Auckland and Christchurch (February 1999). This had not changed by the time of my intensive interviewing period there, from January to March 2000, but plans were definitely afoot. In July 2000 a competitive threat arose from TheSupermarket.co.nz, a small one-store competitor in the northern South Island, supplied and funded by Progressive Enterprises. TheSupermarket.co.nz promised next-day delivery anywhere in the country, and in response (but allegedly also in line with existing expansion plans) Woolworths began limited-range next-day delivery via a link between its couriers and the national postal service to the South Island, as shown i n Figure 2.9a (it originally had a different delivery timetable to provincial towns and to the rest, as shown in the figure, by using different deliverers). By October it had added full service in the small North Island town of N e w Plymouth, and to the peri-urban area to the south o f Auckland. In November it also began a once-weekly service to Great Barrier Island, an island community two hours from Auckland without a supermarket, by putting a truck on a ferry. Woolworths expanded to restricted rural N o r t h Island delivery during 2001, and to the fifth largest city, Hamilton, during 2002 (Figure 2.9b) Woolworths is valuable as a case study because of the (relatively speaking) long involvement of key staff in online grocery (see Chapter 4), because of its operations are shop-based, and because it is a "bricks and clicks" .mixture of "virtual" and "real" retail spaces. Woolworths is an accessible facsimile for Tesco, albeit on a smaller scale.  ."  Quick, Vancouver Quick is a "pure-play" start-up based in Vancouver, Canada. It began operations in late 1999, funded by Leading Brands, a large Canadian product distributor, and assisted by a consortium of "partners" (Figure 2.10): Sentai, for "backroom" fulfilment software; Tompkins Associates, for logistics engineering; Descartes Systems for GIS routing software; Tribal D D B , for the initial website design (later embroiled in a lawsuit with Quick); and ZeroHype Technologies for data and network management (these will be discussed in more detail in Chapter 4). Quick appointed as  57  President a previous executive of Overwaitea stores, and appointed other staff with previous retail experience, in stark contrast to Webvan's initial technology and entrepreneurial appointments of Louis Borders and George Shaheen. Quick serves the Lower Fraser Valley, and to date has not expanded despite highly ambitious plans on inception (Chapter 4). It is a limited line retailer with 1(1  6000 S K U s , and on-sells specialty items and wine sourced from various other retailers through its "Best of Vancouver" market.  Figure 2.10: Quick Partners  ir )•( S C  TOMPKINS  NT A1  Tribal D O B  Vwcoumr I  LEADINGBRANOS"  Source: www.quick.com, "About Quick.com" webpage  Like Webvan, Quick is an e-grocery ("pure-play") "startup": it has no history or background in grocery stores and has no shop floor for customers to physically browse. A s a way of attracting customer attention that is lacking without a shop presence, Quick instituted a multi-level marketing programme o f "Quick Associates". In Quick's terminology this is a "Cooperative Marketing" arrangement where the company uses customer representatives to attract and train prospective customers, in return for a percentage of the markup on sales, in a financially complex programme (see Chapter 4 for more). The company used this marketing technique instead of the expensive television and newspaper advertising campaigns run by Webvan, Woolworths and Tesco: 'Quick will not be dependent on conventional advertising to promote itself. Instead its popularity will be spread by word of mouth and word of mouse through the networking power of Cooperative Marketing.' (www.quickreps.com, accessed February 22, 2001). In the website dedicated to its representatives, Quick promoted the benefits of online grocery shopping, and its  According to a prior staff member of Quick, orders in 2001 were so far below the expected growth rate that trucks were driving around the city, empty, just to advertise the business and make it look like they were busy. Quick had acted on the advice of its partners regarding expected growth rates and had built capacity far beyond its then needs, in keeping with the pace of development of Webvan (more on this in Chapters 4 and 6).  58  alleged strengths i n this business — a list that is worth extracting to illustrate the revolutionary potential for online grocery that was espoused by proponents: 1- The average family spends more than four hours shopping for the same grocery items every week. You now no longer need to drive to and from the store, fight for a parking space, struggle up and down the aisles, stand in line to check out or lug the grocery bags into your home. 2- We want to give you those four hours back; four hours each week to spend as you want, with your family, with your friends, exercising, reading a book or taking up a new hobby or endeavor. 3- We have brought together a wonderful team of grocery, logistics, technology and product experts to provide you with a shopping experience that is like no other. 4- We have designed every aspect of our offering and service to be quick, fun, simple, reliable and easy to use. Quick.com is to grocery shopping, what e-mail is to letter writing. We offer thousands of name brand grocery and related items at prices competitive with your local store.  The company has researched, watched and learned from the mistakes of others in the home grocery delivery market. Rather then just jump into the market place Quick allocated considerable resources developing the logistical infrastructure to support its delivery methods. In fact, most of our methods are proprietary to Quick. Quick has obtained exclusive rights for home delivery of some popular name brands, blocking out the competition.  One of the misconceptions about on-line grocery shopping is "but some people just like to shop, and won't want to buy their steaks from a computer". Well, an explanation of the whole shopping experience needs to be broken down. Typically, 80% of the manual shopping experience is buying dull, ordinary, non-perishables that require very little decision-making by the consumer. We purchase bathroom tissue because we have none left, and we either select our favorite brand, or we buy what's on sale. Most of what we struggle with to our car in plastic bags is filled with this sort of mundane, dull and time consuming shopping. While the other 20%, the picking and choosing of perishables, like steaks, chops, fruits and vegetables, require a necessary choosing process by the individual, and it's fun. When it comes to how much fat we want in our steak, lean or tasty, or how firm our peaches should be (are we going to eat them tonight, or let them ripen on the window sill?), we'd rather make this decision ourselves. Many people enjoy this process, and at Quick, we are about making life better, not about changing people. Our plan is to take away the dull 80% part of the shopping experience from the consumer, save families 4 hours per week on average, to allow them to enjoy the interesting part of shopping, and then to spend time with their families, to do something other then wait in a check-out line!  Once consumers realize that the checkout line is dead; that they can have hassle-free, better-priced groceries, and delivered to their doors, that's when a paradigm shift will occur. Those with the foresight to lay their plans and position themselves in the shift BEFORE it happens will reap the rewards. By 2005, most business commentators will be saying, "looking back, it was easy to predict the explosion in on-line grocery shopping - who would want to go back to that dinosaur age of standing in line behind a huge cart!" Quick will be the company to lead that paradigm shift into the future, taking thousands of Quick's Associates with them in making history in the New Economy.  Many of these sentiments would be shared by other online grocers, and indeed were explicidy hyped by analysts and the media. H o w well Quick and its Associates were able to "reap the  59  rewards" and "make history" will be explored in Chapters 4 and 5. Quick is a useful case study because o f its warehouse-based  operations, its "pure-play" origins, its Associates marketing  programme, and its objective o f revolutionising the "shopping experience". Quick is also a smaller-scale facsimile of Webvan's methods and aims.  Small Potatoes Urban Delivery (SPUD), Vancouver S P U D is an organic home delivery company based in Vancouver, Canada. Operating since 1998, it has taken over nine other Vancouver-based organic home delivery companies (Table 2.5) and is now the largest organic retailer in Canada, and is planning expansion to other cities. It supplies organic produce and a limited range (600 SKUs) o f health food within the Lower Fraser Valley and Whistler. While it initially took orders for its products by telephone and by faxed or posted order forms, options which exist and are still used by a minority, S P U D began accepting orders through its website witiiin a few months of inception — a method it encourages due to reduced processing costs. S P U D has gradually increased its market area