THE ACTION-BASED MODELOF ORGANISATIONAL EVOLUTIONIN Essay Samples

THE ACTION-BASED MODELOF ORGANISATIONAL EVOLUTIONIN THE e-COMMERCE ENVIRONMENT

...r price cut initiated by a firm to improve or defend its relative competitive position. They define a competitive response as a market move taken by a competitor to counteract the initial competitive action of a rival. They classify competitive actions into four categories: entrepreneurial, Ricardian, deterrent and cooptive. They define entrepreneurial actions as ones that implement new combinations of existing resources, which often involve innovation and first-mover advantages. Ricardian competitive actions are ones that exploit unique and strategically relevant resources owned by the firm. 18 Deterrent actions are ones derived from external market power and market leadership which are intended to exploit market power, including limit pricing, predatory pricing, extensive advertising and promotion, investing in excess capacity, product proliferation and preemptive patenting. These three types of competitive action are designed to out-compete rivals. A co-optive action, on the other hand, is one designed to limit or reduce rivalry, such as advance signalling on price changes, disclosing price information and other competitive information through regular announcements, and consistent use of base-point prices. These action types are summarised in Table 1. Basic Components of the Action-Based Model of Organisational Evolution Type of Action Key Resource Strategic Intent Entrepreneurial Entrepreneurial discovery Delay response by exploiting uncertainty Ricardian Factors of production Delay response by exploiting scarce resource ownership and scarcity Deterrent Market share and reputation Delay response by exploiting market power and intimidation Co-optive Parity Avoid rivalry through tacit coordination Table 1 (Source: Grimm & Smith, 1997: p. 68.) 19 Consider the S-curve depicted in figure 6, which shows the evolution of the firm segmented into three stages, start-up or turnaround, growth and maturity. In the first stage the firm has limited resources and a poor competitive position in relation to its rivals. It must rely on entrepreneurial discovery or the managers’ capability to exploit uncertainty and blind spots by taking entrepreneurial actions. As the firm progresses and evolves to a high-growth stage, its attention will shift away from entrepreneurial discovery and entrepreneurial action to the exploitation of specific resources through Ricardian actions. Successful Ricardian actions will be based on the firm’s specific resource advantages. As the firm matures, its orientation shifts to defending its market position with deterrent actions. Relative Resource Strength Time Key Task: Exploiting competitive uncertainty & blindspots Key Task: Exploiting relative resources Key Task: Exploiting market position Entrepreneurial Actions Ricardian Actions Deterrent Actions Figure 6 (Source: Grimm & Smith, 1997: p. 70.) Action-Based Model of Organisation Evolution 20 It should be noted that as a firm evolves it does not focus exclusively on the competitive actions further up the curve. While a firm in the first stage can only attempt entrepreneurial actions, a firm in the second stage can choose between entrepreneurial and Ricardian actions, and one in the third stage can chose between entrepreneurial, Ricardian and deterrent actions. Thus, as a firm evolves its strategic choices widen. The importance of this model is the insights it gives into (i) how firms should act at different stages in their development, (ii) what a firm should expect from a rival given that rival’s relative stage of development, and (iii) how to react to rivals’ actions, having regard to the respective stages of development of the focal firm and the rival. Firms with limited resources and a poor market position can engage in entrepreneurial actions to improve their competitive position and their competitive advantage. Examples of such actions are new product or service introductions or improvements, or copying effective actions from one geographic area or market segment to another. Such firms have no choice but to delay competitive reaction by exploiting uncertainty and competitive blindspots. Firms with relative resource advantages over rivals can use Ricardian actions to impede and delay competitive response from those rivals and the greater the resource imbalance between the firm and its rival the greater will be the delay in the rival’s response. Ricardian actions include exploiting economies of scale or scarce assets. However, the aggressor firm must be cognisant of how their Ricardian actions will affect and be reacted to by rivals. Firms with strong market positions can defend against rivals by taking deterrent actions to exploit their market power. Such actions could 21 include limit pricing, product proliferation or excessive advertising. It is important to respond quickly to challengers’ actions, and to react with a broad array of deterrent actions. However, dominant firms should be careful not to do so in a way which would breach competition rules – e.g. predatory pricing is an extreme form of limit pricing, but predatory pricing is illegal in many jurisdictions. In markets where both resources and competitive strengths are roughly equal among firms, co-optive action will often be in all their best interests. An example of such an action is advance signalling of price changes. Again careful attention to competition rules is required. Table 2 compares and contrasts entrepreneurial, Ricardian, deterrent and co-optive actions. 22 Key Distinctions among Entrepreneurial, Ricardian, Deterrent and Co-optive Actions Categories Entrepreneurial Ricardian Actions Deterrent Actions Co-optive Actions Goal: Avoid rivals Engage rivals Deter rivals Reduce rivalry Source: Opportunity based Resource based Market based Driven by lack of advantage; resource & market share parity Intent: Spontaneous, based on opportunities created from dis-equilibrium Deliberate outcome of plan to exploit resources Deliberate outcome of plan to defend resource position Deliberate outcome to exploit joint behaviour Criteria for evaluation: No guidance on conformance; newness Conform to criterion of economic efficiency or maximisation Conform to criterion of long-term profit maximisation, but not economic efficiency Conform to criterion of joint profit maximisation, but not economic efficiency Credit: Credit goes to the discoverer Credit goes to owner of resource Avoid credit Avoid credit Outcome: Unpredictable Predictable Predictable Unstable Table 2 Source: Grimm & Smith (1997), p. 160. 23 Literature Review: e-Commerce What is Meant by e-Commerce? There are many definitions of e-commerce: “… the buying and selling of goods and services on the Internet, especially the World Wide Web. In practice, this term and a new term, ‘e-business’, are often used interchangeably. For online retail selling, the term ‘etailing’ is sometimes used.” (www.whatis.com) “… the seamless application of information and communication technology from its point of origin to its end point along the entire value chain of business processes conducted electronically and designed to enable the accomplishment of a business goal. These processes may be partial or complete and may encompass business-to-business, as well as business-toconsumer and consumer-to-business transactions.” (Wigand, 1997) “In its simplest definition [e-commerce] is any electronic means used to transact business… As technology advances and Internet use becomes more widespread, ecommerce has come to refer to a wider scope of methods such as e-mail, informational websites, on-line catalogs, Internet ordering, direct product delivery (downloading), and product customization… The most common forms of electronic commerce are: ¨ Business to Business - Businesses selling to other businesses ¨ Business to Consumer - Businesses selling to consumers ¨ Business to Government - Businesses bid on government proposals. ¨ Consumer to Consumer - Consumers sell to other Consumers via auction sites etc.” www.wvebiz.com 24 “… the conduct of a financial transactions by electronic means. With the growth of commerce on the Internet and the Web, ecommerce often refers to purchases from online stores on the Web, otherwise known as ecommerce Web sites. They may also be referred to as ‘virtual-stores’ or Cyber stores... E-commerce can be business to business (B to B) or business to consumer (B to C).” www.straight-on.com “… the use of networked computer technology to facilitate the buying and selling of goods and services between trading partners.” Bathia (1999) While all of the above definitions are valid, the purpose of this dissertation is to test the applicability of an existing model in the new electronic environment. The most appropriate definition is therefore one which focuses on “pure” e-commerce, i.e. with as small an element of off-line activity as possible. Consequently the definition selected is an adaptation of Straight-on.com’s definition, as follows: The emergence of the Internet and e-commerce as a major component of global economic activity has occurred very quickly over the latter part of the last decade. The pace of development and change has been very fast, and many “new economy” business segments (e.g. web consultancy, free content sites) have been born and some so-called “old economy” segments have undergone significant structural change (e.g. the disintermediation and reintermediation of travel agents in the US airline industry (McCubbrey, 1999)). For the purposes of this dissertation e-commerce means consumer purchases from online stores on the Web, otherwise known as e-commerce web sites, virtual-stores or cyberstores, by vendors selling exclusively online. 25 There has been considerable debate in academic and business circles about these changes, and the impact of the Internet and e-commerce on business strategy. However, most of this writing has been focussed: (i) on the identification and classification of different business models as they emerge (e.g. Calkins, Farello & Shi, 2000; Timmers, 1998), (ii) at a tactical level within particular industries (e.g. Cavalli, 1998), (iii) on the modalities of exploiting particular technologies (e.g. Fried, 1998), or (iv) on the extraction of logistical benefits from technology to improve the supply chain and/or reduce costs (e.g. Rayport & Sviokla, 1995). Porter (1996) argues that issues such as item (iv) above are issues of operational effectiveness, and not of strategy. The Author suggests that items (ii) and (iii) are also issues of operational efficiency rather than strategy, and that while item (i) is the only one of the four resulting from a truly strategic decision, many of the relevant works answer the question “what” rather than “how” or “why”. So far as the Author has been able to ascertain, despite the thousands of square miles of Amazonian rain forest which appear to have been consumed, few writers or researchers have sought to extract generic strategic lessons from the admittedly brief history of e-commerce to date. There follows a brief exploration of the writings of those few whom the Author has been able to identify. 26 Wigand & Benjamin Wigand & Benjamin (1995) examined the effects of e-commerce on markets and identified four areas of opportunity and risk: (i) The consumer will benefit through maximum choice at lower cost derived from free market access to all suppliers willing to pay an interconnection cost. (ii) There will be lower co-ordination costs throughout an industry’s value chain. (iii) There will be lower physical distribution costs from cheaper electronic transmission of information and the elimination of intermediate distribution links and inventory carrying costs. (iv) There will be a redistribution of, and a potential reduction in, total profits. Wigand & Benjamin also identified freedom of market access and the potential for reconfiguration of an industry’s value chain as two issues which business strategists need to address. Bloch, Pigneur & Segev Bloch, Pigneur & Segev (1996) propose a framework of ten components to improve, transform or redefine products, processes or business models to increase the business value of e-commerce. Their ten components are summarised in table 3. They go on to map their ten business value components onto Porter’s analytical framework. They do this in a complex way, simultaneously considering the implications of their ten components for (a) a single company in an industry, (b) a new entrant or substitute product, and (c) 27 suppliers to or customers of the industry. Through each of these three perspectives they map each of their ten components onto Porter’s three generic competitive strategies, and explore the implications of each mapping. The significance of this work is that it seeks to show how Porter’s generic strategies can be applied in the e-commerce environment. Calkins, Farello & Shi Calkins, Farello & Shi (2000) have identified three fundamentals of online retailing. They are Content, Community & Commerce (the 3 Cs). Content is defined as whatever appears on the website itself and on hot-linked websites. Table 3 Source: Bloch, Pigneur & Segev (1996) Bloch, Pigneur & Segev’s Framework 28 Community is defined as the core of dedicated consumers who, as a result of site-to-user and user-to-user forms of interactivity (such as chat rooms) become avid marketers of the site. Commerce is defined as including offering goods and services directly, or marketing those of another company for a fee, thus helping to cover the fixed costs of site operations and to offset customer acquisition costs. By applying these fundamentals to the four main emerging e-tailing business models they identified, they further identify seven “levers” which they consider important to achieve success in e-tailing. The four business models are (i) channel supporter – using the internet to support existing channels, (ii) category killer – dominating a particular category of online merchandise, (iii) auctioneer – matching sellers with buyers, and (iv) vertical portal – specialising in a particular industry or product category. The seven levers for success in e-tailing are (i) exploiting two channels to close one transaction, (ii) maximising the value of the whole transaction through premium products, bundling products and services, (iii) leveraging low customer acquisition costs, (iv) exploiting alternative revenue streams, (v) purchasing scale at low volumes, 29 (vi) minimising customer churn, and (vii) maximising pricing potential. Timmers The most interesting examination of e-commerce business models is that of Timmers (1998), which has real generic value. In his rigorously analytical approach he defines what he means by business model, and he defines marketing model. He then develops a systematic approach to identify business model architectures based on value chain deconstruction and reconstruction. He goes on to list eleven business models, and then plots them with examples on a grid as set out in figure 7. One significant outcome of this analysis is that, according to Timmers, the business models in the top right hand quadrant cannot exist in a traditional “old economy” form. Figure 7 Source: Timmers (1998) Timmer’s e-Commerce Business Model Grid 30 Application of Traditional Strategic Perspectives to e-Commerce As can be seen from the above, other than the work of Bloch, Pigneur & Segev (op. cit.), the Author has been unable to locate any writings on e-Commerce strategy from the perspective of the positioning approach, or the resource-based view of the firm. Nor has he been able to find any writings on competitive rivalry in the e-commerce environment. 31 Summary The literature review began with an exploration of the positioning approach to business strategy and the resource-based view of the firm. On the positioning side, Porter’s approach to industry analysis was discussed, and there was a discussion of some of the issues from the perspective of the resource-based view of the firm. Developments from these perspectives were discussed, including Peteraf’s four cornerstones of competitive advantage, D’Aveni’s work on hypercompetition, and Chen’s work on competitor analysis and interfirm rivalry including his seven propositions. Finally Grimm & Smith’s work on strategy as action was described. D’Aveni and Grimm & Smith argue that sustainable competitive advantage, a fundamental concept underlying both the positioning approach and the resource-based view of the firm, is not possible. In recognition of this, and in support of D’Aveni’s identification of hypercompetition, which they document with specific measures across a range of industries, Grimm & Smith have postulated the Action- Based Model of Organisational Evolution. This is an empirically-derived model for markets in a constant state of flux and dis-equilibrium, where organisational outcomes are only partly determined by strategic choice. The Action-Based model categorises a firm’s competitive action into four generic categories, which are related to the relative resource strength of the firm in comparison to its rival. A firm’s ability to engage in different types of competitive action is a function of its relative resource strength. 32 The literature review then looked at e-commerce, and noted the dearth of literature relating to strategic management or competitive rivalry in an e-commerce context. There is a brief exploration of such generic strategic writing for e-commerce as was discovered. In summary therefore, while there has been considerable research and thought on generic strategy making by firms, there is very little in the literature focussed on generic strategy making in the e-commerce environment. 33 Chapter 2 – Research Methodology Problem Definition Broad Statement of Problem “In the space of just a few years, the Internet has become the most exciting medium for commerce and business communications… “No one really knows exactly how the Internet will develop, and there are currently more questions than answers… “The story of the 1990s so far has been about change, and the Internet poses one of the biggest ‘challenges of change’ to change. It also offers one of the biggest commercial opportunities. Given the current size, future growth prospects and attractive demographics of the new media, can any company afford not to investigate the possibilities? Using the Internet and the World Wide Web can provide a competitive edge…” (How the Internet can work for you, 1998, p. 22.) The quotation above highlights the importance of e-commerce to modern business, and explains why businesses can ill afford to ignore it. However, as the literature review discovered, there is almost no guidance available to managers on generic business strategy in the ecommerce environment. A broad objective of this dissertation is to begin to fill this gap. The literature review has identified many research streams dealing with generic business strategy. Of these, the work of Grimm & Smith (1997) is recent, and is focussed on industries characterised by globalisation, rapid technological change, deregulation, intense competition and increasing time pressures. Given these criteria, it seems likely to be 34 relevant in the e-commerce environment, and it is therefore considered that it would be of value to investigate the applicability of Grimm & Smith’s Action-Based Model of Organisational Evolution in the ecommerce environment. The Focussed Research Question The broad statement of the problem above (“to investigate the applicability of Grimm & Smith’s Action-Based Model of Organisational Evolution in the e-commerce environment”) is too general a research question to deal with in a dissertation of this nature, and it is therefore necessary to narrow the focus of the question by concentrating on a particular e-commerce industry or organisation. The criteria for selecting such an industry/organisation should be: · One which is likely to be broadly representative of e-commerce generally in terms of strategic issues (ideally one which exists only in an e-commerce market) · One where the business model is well-defined, so that the strategic issues are not clouded by business model confusion · One where there has been a (relatively) long trading history so as to increase the likelihood of firms having progressed through as many stages of the Action-Based model as possible · A highly competitive environment/business approach Cyberatlas (2000) describes computers, books and music/videos as “the ‘veterans’ of e-commerce”, recognising their comparatively long presence as e-tail products. Gross revenues for these sectors are quoted by Cyberatlas as set out in table 4. 35 e-Commerce Revenue by Category Quarter 1, 2000 Computer hardware/peripherals $852 m Books $461 m Music/video $340 m Table 4 Source: Harris Interactive e.commercePulse Given the average cost of a computer and the average cost of a book or compact disc (CD), it is likely that the figures above represent more online book or CD transactions than computer transactions. Therefore, taking account of their “veteran” status, and the scale of activity, etailing of books & music/videos was selected. Calkins, Farello & Shi (2000) and Timmers (1998), in their works on business models in e-commerce, all cite Amazon.com (Amazon) as examples of the “category killer” and the “virtual community” models respectively, and both articles identify the building of virtual community as an integral part of Amazon’s business approach. Virtual community as a concept is unique to e-commerce and Amazon is one of the oldest businesses on the Internet as well as one that operates exclusively on the Internet. Finally there is a media and public perception that Amazon and Barnes & Noble are engaged in a bitter Internet “dogfight” for market dominance. For these reasons Amazon fulfils all the criteria laid out above, and is therefore selected as the specific organisation to focus on in this investigation. 36 Research Approach Approach Underlying the Original Model Since Grimm & Smith’s model was empirically derived, it seems appropriate that in determining the research approach to be adopted in this dissertation, the research approach adopted by Grimm & Smith in their original work (1997) should be considered. They set out their research methodology in appendix 1.1 of their book. Grimm & Smith derived their model based on two field studies and four archival studies carried out over ten years, and covering various periods between 1975 and 1993. Industries studied were (i) high technology electronics, (ii) computer retailing, (iii) US airlines, (iv) brewing, telecommunications & personal computers, (v) software, and (vi) leader/challenger in 41 diverse industries. The field studies comprised (i) interviews of a total of 72 chief executives, and (ii) completion of questionnaires by those chief executives and some of their senior executives. The focus of the research was to identify competitive actions to which the subject firms had responded, and to gather data about the nature of the action and the nature and timing of their response. While each of the four archival studies had a different time period and research focus, all used the “Structured Content Analysis” methodology of Jauch, Osborn & Martin (1980). This involved the careful reading and coding of thousands of news headlines and abstracts under the headings pricing, marketing, innovation, legal, licensing, and product 37 announcements. A total of 8,546 actions and responses were identified across all the industries studied. In summary, Grimm & Smith derived their model from a combination of primary research among 74 firms, and secondary research on 8,546 competitive actions and reactions. As can be seen from these figures, the majority of empirical data underlying the model is secondary data from four archival studies. Sources of Data Primary data “are collected by [the researcher] … especially to answer [his] research question” (Cooper & Schindler, 1998, page 256), and “secondary data is [sic] data that was [sic] developed for some purpose other than helping to solve the problem in hand” (Fay, 1997). Primary and secondary data each have advantages and disadvantages for the researcher. Advantages of primary data usually include the collection of exactly the information wanted, the exact specification of operational definitions, and the elimination (or at least the monitoring and recording) of extraneous influences on the data (Cooper & Schindler, 1998). Advantages of secondary data include faster speed and lower costs of collection. On the other hand, secondary data have been collected for other purposes and may not meet the needs of the researcher. There may also be definitional difficulties, or other forms of inconsistency among the data. Assessing the accuracy of the data may also be a problem. According to Cooper & Schindler “There is nothing wrong with using primary data under many circumstances, or secondary data under different circumstances, or, rarely and prudently, substituting one for 38 the other when either might be suitable. But the basis for substitution has to be well understood and good judgement applied” (Cooper & Schindler, 1998, p. 256). Furthermore “… secondary data may be used as the sole basis for a research study, since in many research situations one cannot conduct primary research because of physical, legal, or cost influences… Retrospective research often requires the use of published data” (ibid., p. 257). The difficulties inherent in undertaking meaningful primary research to investigate this particular research problem are many. In order to illustrate the applicability or otherwise of Grimm & Smith’s model to ecommerce, a very large sample of data is required, covering as long a period of time as possible. To get data covering the later stages of the model a market segment which is relatively mature in e-commerce terms (e.g. books, CDs, software) must be included, and there are no Irish companies with market leadership positions in any of those industries, thus adding distance (and therefore delay and costs) to the research effort. Furthermore, primary research would involve questioning industry participants with regard to historical events. In the Author’s view it is doubtful if participants’ potentially partisan views carry any greater degree of credibility or veracity than contemporaneous independently documented records. Furthermore, any bias of reporting is likely to be self-compensating, since it is anticipated that major competitive actions and reactions would have been reported in more than one journal. Also, to the extent that media reports may be based on corporate press releases, any resulting inherent bias will not distort the research findings, since the bias underlying a firm’s reporting of its own actions is likely to be an inherent element of the competitive action itself (i.e. it is often a competitive action to “spin” a particular story). 39 Given the above considerations, this situation clearly fulfils almost all of Cooper & Schindler's criteria for preferring secondary to primary data (retrospective research, and physical, legal, or cost constraints), and therefore, in these circumstances, it is considered that the use of secondary data is preferable. Approach Used in This Dissertation As noted above, all four of Grimm & Smith’s archival studies were carried out using the “Structured Content Analysis” methodology of Jauch, Osborn & Martin (1980). It is intended to use this methodology in this study. Jauch, Osborn & Martin put forward their structured content analysis as an alternative to the questionnaire in developing and testing hypotheses in organisational research. This approach uses a carefully developed Content Analysis Schedule to draw specific relevant information from published cases. Content analysis of published data is widely used in history and anthropology. A number of examples of the use of this approach in organisational and strategic analysis are given, e.g. Miller & Freisen (1977) and Schendel & Hofer (1979). The advantages of this methodology versus those of the questionnaire are discussed, as set out below. A typical questionnaire attempts to gather the perceptions of a number of purportedly knowledgeable individuals, and while this is an appropriate approach in many situations there are the usual problems of less than perfect validity, reliability and representativeness. Problems particularly associated with questionnaires include single time collection, same source data and lack of depth regarding dynamic and comprehensive organisational conditions and processes. It is suggested that, in fact, a questionnaire may be completed on behalf of 40 a respondent by a (less knowledgeable) assistant, or alternatively that the respondent may not answer the questionnaire with appropriate care. There is also the possibility of deliberate bias, where the respondent wishes to put a particular (often favourable) spin on his or his organisation’s actions or stance. Case analyses can provide data from multiple sources over several time periods in considerable depth, and are therefore more suitable to explore dynamic changes over time, as is often the case in history and anthropology. However, it isn’t always clear who is the Author of the published data, or what may be the effect of any editorial process that has taken place. Missing data can be a source of bias under either approach. In questionnaire studies an information gap might be considered unimportant, and the researcher could claim his study to be restricted to a particular domain, perhaps even citing an unresolved question or issue as worthy of further research. Furthermore, in questionnaire studies partial responses are often accepted, whereas in content analysis it is suggested that unclear or incomplete data should be discarded. In summary both the questionnaire approach and the content analysis approach can give rise to problems of objectivity, accuracy, distortion and consistency. It is therefore suggested by Jauch, Osborn & Martin that one methodology should be used to verify the findings of the other. The Content Analysis Schedule, like a questionnaire, is designed to estimate specific variables specified in advance by the researcher. Information in the case that is unrelated to the researcher’s interests is ignored, and a case that provides insufficient information is ignored just as a blank questionnaire is considered a non-response. Analysis of research questions rests on the theory behind the schedule and the extent to which it taps the constructs enumerated by the researcher. 41 Just as a valid questionnaire study does not rely on just one response, a structured content analysis should rely on a large number of cases. 42 Implementation Design and Pilot Testing of Content Analysis Schedule A Content Analysis Schedule was designed for this investigation using Microsoft Excel, and is attached as appendix A. As a pilot test of the Content Analysis Schedule, a Harvard Business School case study (Ghemawat & Baird, 1998) which had been used in class in the “e- Commerce: Strategy and Operations Management” elective module, was analysed, and the design of the Content Analysis Schedule was found to be satisfactory. Source and Analysis of Data The structured content analysis data gathered during the pilot test was retained for use in the main study. In addition, further articles were obtained from two searches carried out over online databases of published materials. Details of the sources of data underlying the research are set out in table 5. 43 Data Sources Used in the Research Date of Search Source Number of Articles Generated Not Applicable Ghemawat & Baird (1998) 1 14th July 2000 Infotrac: Newspapers Database 366 23rd July 2000 Infotrac: General Business File International 2,247 Total number of articles 2,614 Table 5 All of the articles resulting from the searches were printed out and the entire articles were “speed read” (the Author has been trained in speed-reading) to identify competitive action by Amazon and its competitors. All such competitive actions were written up and analysed using the Content Analysis Schedule, and the data were then sorted by “Actor”, and within “Actor” by “Date”. The results of this work are described in Chapter 3 – Results. Validation of Technique Jauch, Osborn & Martin (op. cit.) assert the importance of checking the reliability of the technique in a manner similar to that used for questionnaires. In accordance with their recommendations, two Reader-Evaluators were trained in the use of the Structured Content Analysis Schedule, and provided with definitions of the key terms used therein. A copy of the training documentation given to each Reader- Evaluator is attached as Appendix B. Each Reader-Evaluator was asked to read a sample of 30 articles from the population of 2,614 articles that formed the basis of the research. Each Reader-Evaluator 44 was also given a blank Structured Content Analysis Schedule (Appendix A), and was asked to identify competitive actions in the sample of articles and to write them up and analyse them using the Structured Content Analysis Schedule. The sample articles were selected using the fixed interval sampling method based on an initial randomly selected item. The results of the Reader-Evaluators’ analyses were then compared with the results of the Author’s analysis of the same articles. The results of this evaluation are set out in table 6. Results of Validation Testing by Reader-Evaluators Sample Size Number of “Errors”1 in Sample Reader-Evaluator 1 30 3 Reader-Evaluator 2 30 2 Total 60 5 1 An “error” is considered to occur unless the Author and the reader identify the same competitive action(s) from the particular article, and classify it/them similarly on the Structured Content Analysis Schedule. Table 6 The extent of the agreement between the Reader-Evaluators and the Author can be taken as a measure of the reliability of the research methodology. The results of the validation testing can be statistically evaluated as follows: The fixed interval methodology used is a quasi-random method considered to give a sample that is representative of the population 45 from which it is drawn. The sample size is considered large enough to be statistically valid and the population size is considered large in relation to the sample size. Assuming the validity of the “central limit theorem” (provided that the sample size is large, the theoretical sampling distribution of the arithmetic mean of the sample is approximately normal), using a 95% confidence level, we can establish the degree of reliability of the sample as being representative of the population, as follows: (1) Where: p = proportion in sample possessing attribute, (2) q = (1 – p), (3) and n = sample size, (4) then the standard error of the sample proportion = Ö(pq/n) (5) and the confidence interval = p ± 1.96 Ö(pq/n) (6) = 55/60 ± 1.96 Ö(((55/60)(1-(55/60)))/60) (7) = 0.9167 ± 1.96 Ö(0.9167 x 0.0833)/60 (8) = 0.9167 ± 1.96 Ö0.0764/60 (9) = 0.9167 ± 1.96 Ö0.001273 (10) = 0.9167 ± (1.96 x 0.03568) (11) = 0.9167 ± 0.058 (12) 0.8587 < confidence interval < 0.9747 Put another way, we can be 95% confident that the error rate in the population is in the range 2.53% to 14.13%. 46 Chapter 3 – Results Introduction As noted in Chapter 2 – Research Methodology, all of the articles resulting from the database searches were printed out and “speed read” to identify competitive action by Amazon and its competitors. All such competitive actions were written up and analysed using the Content Analysis Schedule, and the data were then sorted by “Actor”, and within “Actor” by “Date”. A summary of the results of this work is set out in table 7. Summary of Results of Structured Content Analysis Number of articles read 2,614 Number of publications used as sources of competitive actions (listed in appendix C) 100 Total number of competitive actions identified 552 Number of competitive actions by Amazon 380 Table 7 A more detailed analysis of Amazon’s competitive actions over time is set out in table 8. 47 Analysis of Amazon’s Competitive Actions Over Time Date (Quarters) Entrepreneurial Ricardian Deterrent Co-optive Total 1994: Q2 5 0 0 0 5 1994: Q3 1 0 0 0 1 1994: Q4 0 0 0 0 0 1995: Q1 0 0 0 0 0 1995: Q2 0 0 0 0 0 1995: Q3 23 0 0 0 23 1995: Q4 0 0 0 0 0 1996: Q1 0 0 0 0 0 1996: Q2 0 0 0 0 0 1996: Q3 0 2 0 0 2 1996: Q4 0 6 0 0 6 1997: Q1 3 8 0 0 11 1997: Q2 1 12 0 0 13 1997: Q3 0 6 9 0 15 1997: Q4 0 4 2 0 6 1998: Q1 1 1 1 0 3 1998: Q2 0 9 1 0 10 1998: Q3 0 3 14 0 17 1998: Q4 1 4 13 0 18 1999: Q1 2 6 17 0 25 1999: Q2 1 11 29 0 41 1999: Q3 2 0 20 0 22 1999: Q4 2 14 50 2 68 2000: Q1 1 8 43 2 54 2000: Q2 2 3 20 8 33 2000: Q3 0 1 6 0 7 Totals 45 98 225 12 380 Table 8 48 Identification of Amazon’s S-Curve Having established a table of competitive actions by Amazon, it was then necessary to plot the results of this work against the Action-Based Model of Organisational Evolution. The first stage in this exercise was to establish Amazon’s S-curve. It will be recalled that the x- and y-axes of the model’s curve (figure 6) are “time” and “relative resource strength” respectively. 49 Plotting the X-Axis The x-axis, time, is easy to measure and plot. Amazon was founded on 16th April 1994, near the beginning of quarter 2 of that year, and writing in August 2000 it is feasible to plot against time from quarter 2, 1994 to quarter 2, 2000 (both inclusive), a total of 6¼ years. When it was subsequently found that Barnes & Noble data had also to be plotted (see below) and that this was not available after quarter 1, 2000, the time plotted on the x-axis was reduced to the 6 years from quarter 2, 1994 to quarter 1, 2000, inclusive. 50 Plotting the Y-Axis Meaning of “Relative Resource Strength” The y-axis, relative resource strength, was more difficult to plot, and called for considerably more judgement. Grimm & Smith subsequently redefine relative resource strength to comprise relative internal resources and relative external market positions. They state (op. cit., page 186): “the action model focuses on the relative internal resource and external market positions of firms. Although a manager may attempt to relate those positions of his or her firm to industry averages or to a set of specific competitors, the analyses should generally be conducted on a pairwise basis, between a focal firm and a single direct competitor”. Grimm & Smith quote from Chen (1996) in support of this view. This pairwise approach requires the evaluation of Amazon’s resources not in absolute terms but in relative terms, i.e. relative to a specific competitor. Grimm & Smith define competitors as “firms that are vying with the focal firm for the same customers in the same markets” (p. 187). As noted supra, Amazon is generally perceived to be in an Internet “dogfight” with its main rival, Barnes & Noble. Therefore, for the purposes of evaluating Amazon’s relative resource strength, it was decided to measure Amazon’s relative resource strength in relation to Barnes & Noble. However, Barnes & Noble Inc. is the world’s largest “old economy” book retailer, not just an Internet bookseller like Amazon. Therefore the 51 question had to be addressed whether to measure Amazon’s relative resource strength in relation to Barnes & Noble Inc., or only in relation to its Barnesandnoble.com division (in latter years, a joint venture with Bertelsmann AG, and since 1999 a publicly traded company (20%)). Since it is an objective of this dissertation to assess the applicability of the Action-Based Model of Organisational Evolution in the e- Commerce environment, it is considered more relevant to confine the comparison to Barnesandnoble.com only. References to Barnes & Noble in the remainder of this dissertation should be construed accordingly. It may be noted in passing that this issue highlights the difficulties with Grimm & Smith’s definition, supra, of competitor. It is arguable whether the “bricks & mortar” division of Barnes & Noble is competing with Amazon for the same customers in the same market. Throughout this dissertation care has been taken to refer to the e-commerce “environment” rather than the e-commerce “market”, since the Author is not convinced that it is a separate market. While Grimm & Smith recommend that “relative resource and competitor analyses are most appropriately conducted for the focal firm and every one of its primary competitors” (p. 187), for reasons of time, resources and scale, this investigation will focus on Amazon vs Barnes & Noble only – in any event the Author has not identified any other significant competitors with a long enough history of e-commerce to make a competitive analysis worthwhile. Relevant Resources According to Grimm & Smith once the principal competitors have been defined, the next step is to analyse each rival’s key resources in relation to those of the focal firm. Resources for this purpose means 52 “all information, knowledge, capabilities, organizational processes and tangible assets controlled by a firm that enable the firm to undertake competitive action”. They suggest that many models are available for assessing a firm’s resources, and then say “we take a relatively simple approach … by examining key functions of the organization, such as marketing, finance, innovation, management, and so on, focusing on dimensions where there are important differences between the focal firm and its rival”. However in the two industry examples that they explore (pp. 188 et seq.), Grimm & Smith arrive at what they term a “subjective” depiction of relative resource positions. Grimm & Smith’s approach to plotting the y-axis therefore requires a first subjective judgement as to what specific types of resource are relevant, and a second subjective judgement as to the quantification of each rival’s score. The Author tried, to the extent possible, to bring as must objectivity to these issues as possible. The principal source of data for this part of the exercise was the US Securities and Exchange Commission (SEC) Electronic Data Gathering, Analysis & Retrieval (EDGAR) database. Data in relation to both Amazon and Barnes & Noble were extracted from documents filed by each of those firms. Following extensive analysis, the measures selected for plotting on the y-axis were (a) revenues by quarter year, and (b) the sum of (i) marketing & sales expenditure and (ii) product development expenditure. This exercise was facilitated by the adoption by both entities of similar classifications for financial reporting purposes. The approach adopted had the effect of retaining a judgmental element in the selection of the specific types of resources considered to be relevant, but once they were selected, the scores of the rival firms in respect of each type of resource was objectively measured. 53 Furthermore, it is considered that the level of judgement involved in selecting revenues as a measure of external market share is negligible, since relative sales revenues must, by definition, be an accurate measure of relative market share. In selecting the sum of (i) marketing & sales expenditure and (ii) product development expenditure as the measure for internal resources regard was taken of the nature of e-commerce, and the strategy pursued consistently by Amazon of building market share and breadth of service/product range at the expense of profitability (e.g. Bannan, 2000). It is noted that in later years descriptions such as “technology and content” were substituted for “product development” in SEC EDGAR filings. 54 Plotting the Curve Figure 8 was created from the data extracted from SEC EDGAR. For the purpose of plotting the diagram it was necessary to calculate an index of relative resource strength. This was done as follows: For relative external market share, in early periods neither firm had sales revenues, so an index of zero was used. In subsequent periods when Amazon has sales revenues but Barnes & Noble had not yet commenced trading, an index value of 1 was used. In later periods where both firms were trading, Amazon’s revenue was divided by Barnes & Noble’s revenue to arrive at the “index” value. A similar methodology was used to index internal resource strength. It is apparent from table 8 that Amazon’s first detected Ricardian action occurred in quarter 3, 1996, its first detected deterrent action in quarter 3, 1997, and its first detected co-optive action in quarter 4, 1999. If one assumes, as appears reasonable, that Amazon commenced these types of actions as soon as it could, then the windows for each action category are as shown in Figure 9. Finally, the numbers of each type of competitive action were plotted against time and overlaid on the S-curve – see figures 10(a) to 10(e). Figures 10(a) to 10(e) illustrate the point that as Amazon’s resources strengthen, the firm is able to broaden the menu of competitive actions available to it in its battle with its rivals. This is as predicted by the theoretical model. 55 F ig u r e 8 : A m a z o n .c o m v s B a r n e s a n d n o b le .c o m R e la t iv e R e s o u r c e S t re n g th s 0 5 1 0 1 5 2 0 2 5 3 0 3 5 4 0 94/Q2 94/Q3 94/Q4 95/Q1 95/Q2 95/Q3 95/Q4 96/Q1 96/Q2 96/Q3 96/Q4 97/Q1 97/Q2 97/Q3 97/Q4 98/Q1 98/Q2 98/Q3 98/Q4 99/Q1 99/Q2 99/Q3 99/Q4 00/Q1 T im e Relative Resource Index E x t e r n a l m a r k e t s h a r e I n te r n a l R e s o u r c e s 56 Fig u r e 9 : Am a z o n .c o m v s B a r n e s a n d n o b le . c om A m a z o n ' s A c t ion-B a s e d M o d e l o f O rganis a t io n a l E v o lution 0 5 1 0 1 5 2 0 2 5 3 0 3 5 4 0 94/Q2 94/Q3 94/Q4 95/Q1 95/Q2 95/Q3 95/Q4 96/Q1 96/Q2 96/Q3 96/Q4 97/Q1 97/Q2 97/Q3 97/Q4 98/Q1 98/Q2 98/Q3 98/Q4 99/Q1 99/Q2 99/Q3 99/Q4 00/Q1 Tim e Relative Resource Index E x t e r n a l m a r k e t s h a r e In t e r n a l R e s o u r c e s Entrepreneurial Actions Ricardian Actions Deterrent Actions Cooptive Actions 57 F ig u r e 1 0 ( a ) : A m a z o n ' s R e l a t i v e R e s o u r c e S t r e n g th s a n d C o m p e t i t i v e A c t io n s 0 1 0 2 0 3 0 4 0 5 0 6 0 94/Q2 94/Q3 94/Q4 95/Q1 95/Q2 95/Q3 95/Q4 96/Q1 96/Q2 96/Q3 96/Q4 97/Q1 97/Q2 97/Q3 97/Q4 98/Q1 98/Q2 98/Q3 98/Q4 99/Q1 99/Q2 99/Q3 99/Q4 00/Q1 T im e Number of Competitive Actions 0 5 1 0 1 5 2 0 2 5 3 0 3 5 4 0 Relative Resource Strength Index E n t re p r e n e u r ia l R ic a r d ia n D e t e r r e n t C o - o p t i v e E x t e r n a l m a r k e t s h a r e I n te r n a l R e s o u r c e s 58 Figure 10(b): Amazon's Relative Resource Strengths and Entrepreneurial Competitive Actions 0 5 10 15 20 25 94/Q2 94/Q3 94/Q4 95/Q1 95/Q2 95/Q3 95/Q4 96/Q1 96/Q2 96/Q3 96/Q4 97/Q1 97/Q2 97/Q3 97/Q4 98/Q1 98/Q2 98/Q3 98/Q4 99/Q1 99/Q2 99/Q3 99/Q4 00/Q1 Time Number of Competitive Actions 0 5 10 15 20 25 30 35 40 Relative Resource Strength Index Entrepreneurial External market share Internal Resources Figure 10(c): Amazon's Relative Resource Strengths and Ricardian Competitive Actions 0 2 4 6 8 10 12 14 16 94/Q2 94/Q3 94/Q4 95/Q1 95/Q2 95/Q3 95/Q4 96/Q1 96/Q2 96/Q3 96/Q4 97/Q1 97/Q2 97/Q3 97/Q4 98/Q1 98/Q2 98/Q3 98/Q4 99/Q1 99/Q2 99/Q3 99/Q4 00/Q1 Time Number of Competitive Actions 0 5 10 15 20 25 30 35 40 Relative Resource Strength Index Ricardian External market share Internal Resources Figure 10(d): Amazon's Relative Resource Strengths and Deterrent Competitive Actions 0 10 20 30 40 50 60 94/Q2 94/Q3 94/Q4 95/Q1 95/Q2 95/Q3 95/Q4 96/Q1 96/Q2 96/Q3 96/Q4 97/Q1 97/Q2 97/Q3 97/Q4 98/Q1 98/Q2 98/Q3 98/Q4 99/Q1 99/Q2 99/Q3 99/Q4 00/Q1 Time Number of Competitive Actions 0 5 10 15 20 25 30 35 40 Relative Resource Strength Index Deterrent External market share Internal Resources Figure 10(e): Amazon's Relative Resource Strengths and Co-optive Competitive Actions 0 0.5 1 1.5 2 2.5 94/Q2 94/Q3 94/Q4 95/Q1 95/Q2 95/Q3 95/Q4 96/Q1 96/Q2 96/Q3 96/Q4 97/Q1 97/Q2 97/Q3 97/Q4 98/Q1 98/Q2 98/Q3 98/Q4 99/Q1 99/Q2 99/Q3 99/Q4 00/Q1 Time Number of Competitive Actions 0 5 10 15 20 25 30 35 40 Relative Resource Strength Index Co-optive External market share Internal Resources 59 Chapter 4 – Analysis & Discussion Applicability of the Model in the e-Commerce Environment The results of the structured content analysis, summarised in table 8, indicate a progression by Amazon from entrepreneurial competitive actions through Ricardian, deterrent and co-optive competitive actions, as predicted by the Action-Based model. It is also noted that Amazon continued to engage in entrepreneurial actions when it had relative resource advantage sufficient to allow it to engage in Ricardian, deterrent, or co-optive action types. Similarly, Amazon continued to engage in Ricardian actions when it had relative resource advantage sufficient to allow it to engage in deterrent and cooptive actions, and it continued to engage in deterrent actions even when it had relative resource advantage enough to engage in co-optive actions. This is entirely consistent with the Action-Based model, as outlined in Chapter 1 – Introduction and Literature Review. However the Author was initially surprised by the level of continuing entrepreneurial and other early stage activities occurring in later stages when Amazon had the capacity to compete on later stage activities. While Grimm & Smith do not deal with the relative proportions of different activity types over the life of the model one would intuitively expect that a competitor with the resources to compete in a way their rival cannot match would exploit that advantage. Indeed this logic is implicit in the model. 60 Summary Analysis of Amazon’s Competitive Actions Over Time Date (Quarters) Entrepreneurial Ricardian Deterrent Co-optive Total Entrepreneurial Phase 94/Q2 to 96/Q2 29 0 0 0 29 Ricardian Phase 96/Q3 to 97/Q2 4 28 0 0 32 Deterrent Phase 97/Q3 to 99/Q3 7 44 106 0 157 Co-optive Phase 99/Q4 to 00/Q1 5 26 119 12 162 Totals 45 98 225 12 380 Actions Types as Percentage of All Actions in Each Summary Period Date (Quarters) Entrepreneurial Ricardian Deterrent Co-optive Total Entrepreneurial Phase 94/Q2 to 96/Q2 100% 0% 0% 0% 100% Ricardian Phase 96/Q3 to 97/Q2 12% 88% 0% 0% 100% Deterrent Phase 97/Q3 to 99/Q3 4% 28% 68% 0% 100% Co-optive Phase 99/Q4 to 00/Q1 3% 16% 74% 7% 100.0% Totals 12% 26% 59% 3% 100.0% Table 9 61 Table 9 summarises the data from table 8 into the four time periods demonstrated in figure 9. Of the total 380 competitive actions studied, as many as 205 actions (54%) (those italicised in table 9) were action types occurring in periods when more advanced competitive actions were possible. However a closer examination revealed that 119 of those 205 “after time” actions were deterrent actions occurring in the final or “co-optive” phase. It seems likely that as time goes on the proportion of deterrent actions in the final time phase will give way to co-optive actions. The summary in table 9 also draws attention to another trend very noticeable from table 8. This is the general increase in the absolute number of competitive actions with time. This may be due to a number of reasons: 1. There may be correlation between passage of time and the intensity of competitive rivalry. As time progresses and a firm develops it may be able to increase the number of competitive actions it initiates. As discussed in the literature review, Grimm & Smith have observed increasing levels of competitive activity over time in the US airline industry, the software industry, and in other industries. 2. It is likely that as Amazon.com developed and built its market presence, more newspapers and magazines became aware of its existence and took an interest in its activities. Thus it became increasingly likely that more of its competitive actions would be reported. 3. The research data was drawn from two online databases of periodicals. One of those databases (Infotrac: General Business File International) covers articles from 1980 onwards, but the other database (Infotrac: Newspapers) covers articles only from 1997. It was initially suspected that the inclusion of articles from the Infotrac: Newspapers database could be distorting the result, but a close 62 review of the raw data indicated that of the 380 competitive actions studied, only 45 (12%) were drawn exclusively from that database. Furthermore, another seven (2%) were drawn exclusively from the HBS case study (Ghemawat & Baird, 1998) which discusses the competitive situation of Amazon and Barnes & Noble in 1997 and gave a history of competitive activity in the online book retailing market up to 1997, thus counterbalancing, at least partially, any skew from the inclusion of articles from the Infotrac: Newspapers database. In the author’s judgement, it is likely that the first and second reasons advanced above each contributed to the increasing number of competitive actions over time. Overall, therefore, the first conclusion from the results of the research is that the Action-Based Model of Organisational Evolution does apply to Amazon.com in its rivalry with Barnesandnoble.com. 63 Other Matters Resource Levels The identification of appropriate measures of relative resource strength was the most judgmental issue in the research work undertaken. If different measures of relative resource strength had been used the general shape of the Action-Based Model curve in figure 9 may have conformed more closely to the generally accepted S-curve shape as in Grimm & Smith’s depiction of the theoretical model (figure 6). However a closer examination of the data underlying the graph in figure 9 is instructive. This data is set out in table 10. The method of compiling the index was described under “Plotting the Curve” in Chapter 3 – Results, as follows: For relative external market share, in early periods neither firm had sales revenues, so an index of zero was used. In subsequent periods when Amazon has sales revenues but Barnes & Noble had not yet commenced trading, an index value of 1 was used. In later periods where both firms were trading, Amazon’s revenue was divided by Barnes & Noble’s revenue to arrive at the “index” value. A similar methodology was used to index internal resource strength. It can be seen from figure 9 and table 10 that a spike occurs in the internal resource curve in quarter 1, 1997, and in the external market share curve in quarter 2 of the same year. In fact, if quarters 1 and 2, 1997 were eliminated from the internal resource curve, and quarters 2 and 3, 1997 were eliminated from the external market share curve, the curves would align reasonably closely to a flat elongated version of the classical S-curve portrayed in the theoretical model in figure 6. Table 10 64 Data Underlying Amazon's Action-Based Model of Organisational Evolution 94/Q2 94/Q3 94/Q4 95/Q1 95/Q2 95/Q3 95/Q4 96/Q1 96/Q2 96/Q3 96/Q4 97/Q1 97/Q2 97/Q3 97/Q4 98/Q1 98/Q2 Amazon $000 $000 $000 $000 $000 $000 $000 $000 $000 $000 $000 $000 $000 $000 $000 $000 $000 External Market Share: Sales Revenue 0 0 0 0 0 170 341 875 2,230 4,173 8,468 16,005 27,855 37,887 66,011 87,375 115,977 Marketing & Sales Expenses 0 0 0 0 0 67 133 205 696 2,251 2,938 3,906 7,773 10,979 16,306 19,503 26,452 Product Development 8 15 15 43 43 43 42 263 394 755 901 1,575 2,808 3,582 4,520 6,729 8,060 Total Internal Resources 8 15 15 43 43 110 175 468 1,090 3,006 3,839 5,481 10,581 14,561 20,826 26,232 34,512 Barnes & Noble $000 $000 $000 $000 $000 $000 $000 $000 $000 $000 $000 $000 $000 $000 $000 $000 $000 External Market Share: Sales Revenue 0 1,593 3,252 7,060 9,013 11,380 Marketing & Sales Expenses 81 413 2,954 5,407 6,613 19,471 Product Development 74 458 1,098 1,626 1,804 2,129 Total Internal Resources 0 0 0 0 0 0 0 0 0 0 0 155 871 4,052 7,033 8,417 21,600 Amazon'sRelative Resource Strength Index Values Index Values Index Values Index Values Index Values Index Values Index Values Index Values Index Values Index Values Index Values Index Values Index Values Index Values Index Values Index Values Index Values External market share 0.000 0.000 0.000 0.000 0.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 17.486 11.650 9.350 9.694 10.191 Internal Resources 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 35.361 12.148 3.594 2.961 3.117 1.598 98/Q3 98/Q4 99/Q1 99/Q2 99/Q3 99/Q4 00/Q1 $000 $000 $000 $000 $000 $000 $000 153,648 340,371 293,643 314,377 355,777 676,042 573,889 37,517 49,551 60,717 86,165 86,555 179,713 140,111 13,374 18,644 23,477 34,213 44,608 57,424 61,244 50,891 68,195 84,194 120,378 131,163 237,137 201,355 $000 $000 $000 $000 $000 $000 $000 15,561 25,880 32,317 39,065 49,051 82,134 78,244 14,685 29,654 18,909 22,881 26,289 43,474 32,204 2,046 2,553 3,519 4,049 5,251 8,187 6,364 16,731 32,207 22,428 26,930 31,540 51,661 38,568 Index Values Index Values Index Values Index Values Index Values Index Values Index Values 9.874 13.152 9.086 8.048 7.253 8.231 7.335 3.042 2.117 3.754 4.470 4.159 4.590 5.221 65 Why then, did these spikes occur? They occurred because Barnes & Noble commenced developing their website in quarter 1, 1997 (or at least that is when it appears to have started based on publicly available information), and the website opened for business in quarter 2, 1997. That the relationship between Barnes & Noble’s development expenditure in its first quarter and Amazon’s corresponding expenditure in its 12th quarter should be a factor of 35, or that Amazon’s market share in its 8th quarter of trading should be 17 times Barnes and Noble’s market share in its first quarter is hardly surprising. Clearly the spikes and the distortion in the shape of the curve are the result of extreme short-term disproportionalities in the relative resources of the two firms, rather than evidence of a real divergence between Amazon’s Action-Based Model and the theoretical model as depicted in figure 6. Fine-Tuning the Model for the e-Commerce Environment Grimm & Smith’s theoretical model is depicted in figure 6, and it is based on the classical S-curve. The classical S-curve is a theoretical representation of the lifecycle growth of a firm or product. Usually time is plotted on the x-axis and progress (corporate growth, product market share, or similar) along the y-axis. In the early part of the curve, a lot of time and effort is expended for very little progress. However, the pace of progress gradually quickens and the rate of increase of progress accelerates until a very high rate of growth is achieved. Growth usually continues at a rapid rate for some time before a gradual levelling-out occurs. In its early stages, this model implies a young, growing organisation, with limited resources (financial, management, market share, etc.), struggling to survive. This picture is consistent with the conventional view of the “old 66 econo...

• View Full Text