Mimetic behavior and success:
Follow or Leave the herd?
Bachelor Thesis by Jos Kiekens
Date:
26-06-18
Student number:
10791477
Program:
Economics and Business
Track:
Business Administration
Topic:
Can Elephants Dance?
Statement of Originality
This document is written by Student Jos Kiekens, who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document are
original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Abstract
Herd behavior is a common practice in business and has attracted scholarly attention. However, the long-term effects of herd behavior and leaving the herd on the performance of a firm have not gained much attention. This article examines the effects on performance of herd behavior on the long term, by comparing early and late adapters to in a sample of 120 public retail companies from North America and Europe. The right moment to adapt to the herd and the performance of the companies that have left the generic strategy of the herd have been examined. The strategy that has been examined is the e-commerce adoption and activities. Results indicate no significant difference in performance between early movers and late adopters. The moment of entry does not influence the performance significantly. However, leaving the herd has a significant
negative effect on the performance of a firm. This research gives a possible model for the advantages and disadvantages of early adopters and late followers. There is evidence that companies that leave the herd are worse off in
performance than companies that stick to the herd. Once a strategy is chosen, it is better to stay with it.
1. Introduction
Companies have been copying strategies and products from each other since the beginning of business. Especially in the 21th century, where technology seems to develop faster and faster every year and companies struggle to keep up with the competition (Adams, Nelson & Todd, 1992). Uncertainty is one of the forces that encourage players in the market to imitate each other’s strategies (DiMaggio & Powell, 1983). The current trend of online presence of companies is a good example. Many companies struggle with the profitability of their online operations, but due to forces from stakeholders and consumers they feel that they have to go with this trend. They are afraid to miss out on potential profits in the future (Goolsbee, 2018). Some strategies have been adopted for over 15 years now, so the long-term effects are measureable. The bandwagon effect and mimicking in strategic choices has been researched in many settings. There are
many theories about the adoption and cause of the bandwagon effect and herd behavior, but there has been lack of research to the actual success of adoption such a strategy. Research shows (Heugens & Lander, 2009) that companies adapt to these new strategies because they think that others have information they do not posses. So that there choices are based on this unique info, and they should adapt to this too. In reality do all companies look at each other’s strategies and copy each other, whilst sometimes underestimating their own judgment.
Adding to that, there are companies that tried to adopt their operations to the competitors, but in the end choose their own strategies (Porter, 2011; Ewusi-Mensah & Przasnyski,1991). Unsuccessful copying of strategies and abandoning strategic choices are all part of the current landscape of strategic management. The practice of taking the strategy of pioneers in the market and adopting a late-movers own strategy to it is not always successful and can end up in a weak performance. This research aims to provide answers on successful and unsuccessful herd behavior in strategic choices.
Current literature has focused on single country analysis for strategic renewal and herd behavior (Volberda et al, 2001). Suggested is a cross-country analysis and comparison of different continents to see if it is due to national culture to adapt to the herd. There has not been as much longitudinal research to mimicry in strategy and the bandwagon effect. The longitudinal effect to
successful or unsuccessful adaption is especially interesting, because to my knowledge there has been no research to this particular topic. The moment of entry in the market and the advantages of different marketing strategies for late adopters and early adopters is widely debated (Bowman & Gatignon, 1996). There are advantages and disadvantages for both groups. The moment of entry impacts the profits and revenues greatly, which makes finding the exact moment of hopping on the bandwagon an interesting subject.
In addition, the relative moment when you are a first mover or late adapter has been researched (Gal-or, 1985). Although this has been lively debated by other researchers. Suarez & Lanzola (2005) argue that there are many factors that influence the success of an early mover or late mover. Most of the success is derived from industry and firm specifics. Their research is based on existing literature. The relative profitability and performance of the
companies adapting first or last have yet to be researched. This can be a valuable result for managers to see if it is more interesting to be one of the early adopters or once more widespread adoption of a certain strategy occurs in terms of performance. A timeline for successful adoption might convert to a framework for strategic choices when choosing to adopt for a bandwagon move.
Finally, the abandoning of a bandwagon decision or the unsuccessful
adaptation of such a strategic decision has yet to be researched. Current research to abandoning of strategies has only been done to identify why companies
abandon certain strategies (Greve, 1995). Abandoning projects that are
financially and strategically unsuccessful can lead to more firms abandoning the strategic choice (Ewushi-Mensah & Przasnyski, 1991). Testing if the leaving strategy is a profitable one can lead to other firms following this strategy. This can be interesting, because some firms make the decision to follow a path because of the competition, which in time turns out to be not profitable. How they further adopt and perform after this move, compared to others, has yet to be researched.
Research to the difference between first mover strategy and second mover is through mixed findings a heavily debated subject. A longitudinal research over different industries and regions can bring an answer to this subject. In addition, the firm that decides to abandon the strategy and choosing a different path than the competition is yet to be researched.
The research question that is pursued in this paper is: What is the long-term
effect of herd behavior (following and leaving) on performance?
2. Literature review
2.1 Institutional theories
The basis of the behavioral theories within institutions derives from institutional theory. Institutional theory has been stated in different forms and arguments (Scott, 1987). The institutional theory provides a theoretical foundation for critical issues and theorizing multiple levels of analysis, which are essential for the research of companies (Kostova, Roth & Dacin, 2008). This theory is
transformation within a firm, it explains comparative business systems, it explains similarities of practices resulting from isomorphic pressures in the industry, to study diffusion within businesses and to explain the relationship between the organizations main country and the culture outside the main country (Kostova, Roth & Dacin, 2008). The focus in this research lays with the isomorphic pressures from the industry to adapt to industry norms and trends. Most often are these actions that are established as standard in the industry. Past research has been done to isomorphic pressures in local environments, comparing big national players to smaller local companies in the same industry. Isomorphism is important in many industries. For many companies it is used to improve the organizational legitimacy for stakeholders and new customers. They demand certain company structures or features, to serve their legitimacy in the industry and to make them a more normative player in the market (Deephouse, 1996). DiMaggio and Powell (1983) describe this as the iron cage. Generating more homogenous organizations over time due to better established industries. Structures like the multidivisional structure have been adopted by a great deal of firms to conform to industry standards. This even increases the likelihood of receiving a loan from the bank (Deephouse, 1996). The organizational field consists of a range of ‘normal’ strategies that a firm can adopt in a certain industry. These strategies are created through the isomorphic process (Van de Ven et al, 1995).
The isomorphic process that shapes the strategies of companies in an industry can be related back to herd behavior of humans. Investors act the same way to each other as companies do. They reflect the decisions of others in order to shape their own actions (Banerjee, 1992). The ‘herd externality’ is a model that defines that peoples actions depend on their own research in combination with the information they get from actions of others (Banerjee, 1992). The first few decision makers determine where the crowd will form. This also implies on strategy making. Governments and banks have certain prejudices of strategies that they find legitimate. If a company decides to alter their strategy, they come across barriers that make their financial position weaker (Deephouse, 1998). This pushes companies to adapt to the herd and take on a similar strategy. A research done under customers of different retailers pointed out that an online
presence and store add to the total trust towards a retailer (Everard & Galletta, 2005). A possibility for this behavior is the cost to obtain new information or different information than competitors already have. Also, the firm suspects that other firms have information that they themselves don’t have and they try to free ride this information by adopting their strategy (Kennedy, 2002).
2.2 early and late adopters
The herd behavior of subsequent users influences the adoption decision of the agent who moves first (Choi, 1997). This is risk-averse behavior, to follow the first-movers or the success stories in a particular industry. Technological adoptions are adopted by first-movers, whom influence followers in the market and set the standard in the industry. Meanwhile this decision does not have to be a good one (Choi, 1997).
Another factor that adds to copying of strategic decisions is the
bandwagon effect. This tends to derive from dominant logic (Prahalad & Bettis, 1986) or strategic focus as reflected in shared believes about certain
organizational structures and methods (Glynn, Barr & Dacin, 2000). The
bandwagon effect is often used to explain mass adoption of certain strategies in a particular sector in rapid succession. This massive adoption often comes without a firm’s individual research and judgment (Abrahamson & Rosenkopf, 1993). This has an impact on the profits of all firms in a market. The early adaptors lose their market share to new entrants and the new entrants easily copy costly investment in the new technologies. The late adopters have the advantage of using the technology that is available and combining it with their own marketing techniques. This allows them to Create their own differentiated brand and strategy within the industry. On the other hand, first movers have the advantage of being the first entrants to the market. They gain market share more quickly and in most cases set the industry standards (Gal-Or, 1985). First movers have in most cases the largest market share (Gal-Or, 1985), but according to Liebermann and Montgomery (1988) these firms spent many resources on maintaining this advantage in market share. This results in lower profits for these companies.
The most often occurring research on strategy suggests that one should find balance between conforming to industry norms and being different as others (Baden-Fuller, Porac & Thomas, 1989). Also, past research on the balance
between high conforming and high differentiation suggest that a firm is best of in profitability, once it finds balance in moderate differentiation (Hambrick et al, 2004).
The bandwagon effect on new technologies has only been researched in a cross-sectional research design. The long-term effects on revenue and profits are still to be researched (Roa et al. 2001; Strang & Macy, 2001). The insight of long-term strategy copying is necessary to see the development in revenue. New players set up their marketing strategy to gain new market share fairly quickly from the pioneers.
The first research hypothesis is the relative profitability of a certain choice. As Lieberman & Montgomery (1988) described as first mover
advantages, they state that first movers are pioneering in their industry by using new techniques. This happens also in mature markets. Late adopters might encounter higher cost to derive more market share in the new technology. The first mover also has more information to its disposal. Disadvantages that they state for first movers are the ability of the competition to free ride the new technology developed by the first mover. The last disadvantage is the incumbent inertia that makes it difficult to adapt to the new technology. These phenomena do relate to the e-commerce adoption of firms. The early adopters were indeed pioneers in their industry, but encounter various problems such as the amount of Internet penetration at the time they started their web stores. The net impact of first mover advantages is hard to determine and differs per industry and
situation. However, the advantages for an Internet based company are set to be different than the advantages brick and mortar companies have (Mellahi & Johnson, 2000).
Xia, Tan & Tan (2008) adopt a longitudinal approach to the bandwagon effect in firms moving to China after the open-border act. They observe a small group of early adopters, followed by a big herd of companies whom want to do business after the success stories of other western companies in China. These companies set the standard for all other companies expanding to China in that
era. Xia, Tan & Tan (2008) suggest more longitudinal research in specific
strategic choices, to strengthen the claims of the bandwagon effect and check the effects on performance.
Due to these mixed findings we select to different groups to determine if in this database we can find evidence for first mover advantages in the long run or late adapter advantages once the technology matures. Both groups have their advantages and disadvantages. Two hypotheses are set up to test these theories:
H1a: Early adopters are more profitable than late adaptors. H1b: Early adopters are less profitable than late adapters.
2.3 Moment of entry
Early adapters and late followers are subject of discussion in the first hypothesis, but the middle group, the early followers, is still left out. The early followers see the new development in the market and often bring a new innovation with their own entry into the market. Most of the time due to the infancy of the market, where there is a lot of technological innovation possible (Min, Kalwani & Robinson, 2006).
Isomorphic pressures can result in companies adapting to industry standards. They choose the strategy that all other companies are taking, just to not miss out on potential revenues (Kostova, Roth & Dacin, 2008). If this theory holds, the later the moment of entry and less considered the move would be for potential profits. Companies falling in this trap of isomorphic pressures often see their profits and stock prices drop (Kostova, Roth & Dacin, 2008).
On the other hand, late followers have the advantage of copying the successful parts of the strategy of the early adopter. They can fine-tune their strategy and enter the market later with a service or product that can enjoy higher growth rate and market potential than the pioneers (Shankar, Carpenter, & Krishnamurthi, 1998). The late mover is also able to overtake the pioneer in innovation and sharpen the competition between the pioneer and the late movers. This can enable the late mover to overtake the pioneer (Brendt et al. 1995).
Order of entry to the market has been demonstrated to have a significant effect on a firms market share. The early adopters and pioneers have a
significant greater market share than followers. But often these pioneers lack in marketing skills, where the late followers take their opportunities (Bowman & Gatignon, 1996). Taking this into account, with the transparancy of todays markets, it can be more profitable for late followers than for pioneers.
The moment of entry makes an important determination in the strategy of the firm. The relative moment of following the herd can influence the
profitability on the long run (Ruiz-Ortega & Garcia-Villaverde, 2008). The
previous research in this field has focused on the performance of three different categories: pioneers, early followers and late followers. The new aspect of this research is the relative timing and company in the industry that adopts a certain strategy. The timing between these categories is measured and can be brought down to a certain year.
H2: The later the moment of adopting the strategy, the worse for the performance of the company.
2.4 Leaving the herd
In addition to the other two hypotheses, leaving the herd after choosing a strategy has not been researched yet. Current literature only covers strategy changes and change management (Kavanagh & Ashkanasy, 2006), but not the abandonment of a certain type of strategy and its consequences on performance. In this research the relative performance of companies that choose to leave the herd is measured against the average of all the companies that stay with the herd. Leaving the herd is observed as a differentiation strategy. Choosing a different strategic path or proposition for a product or service can influence the performance and revenue of a firm (Porter, 2011).
The research of Choi (1997) suggests that an extra test is needed to confirm their theory about technological adaptation and herd behavior. One part that is relatively unexplained is the rejection and abandoning of certain strategies after adoption. Which makes the likelihood for other players to also reject the strategy larger (Roa, Greve & Davis, 2001). Once players exit the market successfully, the
other players are bound to follow on the exiting strategy as well. The question is if this exit out of the market is due to financial reasons or due to the decline phase of the market cycle.
Current literature suggests that projects and strategies are abandoned due to pressures of disappointing returns and high costs of development (Ewushi-Mensah & Przasnyski, 1991). They fail to acquire the extra market share to cover the costs of these developments. They choose to differentiate from the market and make their product or service more exclusive and outstanding (Porter, 2011).
Taken together, due to the first starting a strategic move and then again abandoning this choice, makes this a expensive move. The general tendency on these moves is negative for the overall performance of a company. This results in the following hypothesis:
H3: Firms that choose to stop following the herd perform worse than firms that stick to the herd.
3. Research design
The bandwagon or herd strategy can be applied to e-commerce adoption by retail firms. Companies struggle with profitability through these channels, but still seem to adopt this strategy due to pressure from stakeholders and industry conformity (Goolsby, 2018). Because most public traded e-commerce companies share their online revenues in their annual reports, this is an opportunity to compare these.
The analysis on performance of bandwagon or herd adoption to strategy needs to be in a longitudinal research design. The context for this research is the retail industry in the United States and Europe. The market is known for it’s strategic mimicking; many retailers follow other successful brands and try to differentiate just a little but from each other. There is a lot of data available from retailers in the US and Europe, which makes it more convenient and reliable to use this data as a basis for this analysis.
3.1 Type of research
The most convenient way to do research to herd behavior is to compare performance of different firms with each other to see if the companies compare in revenue and profit. This information is for most of the big corporations public information, which makes desk research the best way to gather enough
information. Secondary data is the basis of this research. There has been no primary data collected, only indexes and ratios are calculated based on this dataset. Quantitative research is used to measure the differences and draw a conclusion on these public figures. The advantages of this type of research are that an accurate and factual analysis can be made, without the interference of opinions or experiences of qualitative research. Quantitative information gives the most accurate description of performance and is thus the best measure to compare firms with each other. Another advantage is the amount of data that is gathered in a relative short period of time. The amount of data that is collected through the quantitative method is much greater than that can be collected through qualitative methods. A disadvantage of quantitative analysis is the choosing the wrong population or sample to represent the problem that is being addressed (Simon, 2011). Another disadvantage is that there are no exploratory new outcomes that are derived this type of research. There is a set amount of data and points that is collected and the data is significant or not. However, qualitative data can give new insights through interviews or conversations with the target population. Overall, the best way to analyze the performance and standardized measures of companies is through quantitative data.
This research is based on positivist method, relying on statistics. The quantitative method is used to look at correlations and relations between variables (Rolfe, 2006).
3.2 Population and sample
The total population consists of public trading retail companies in Europe and North America with experience in e-commerce. The companies are retailers that started their business in the North America and Europe. These markets fit the criteria, because these markets are generally the most advanced and score high on technological advancement. The total number of public trading
companies in North America and Europe is 5000. The biggest 1000 companies of both continents are taken into the population. A random sample calculation gives a sample of 120 companies taken from this population. The sample size is based on a confidence level of 95% and margin of error of 5%. The data is split in two main groups, USA and Europe, to avoid differences in the marketplaces defining part of the findings. To eliminate industry differences the industries are filtered for: Clothing industry, Supermarket industry and Electronics industry. These categories are compared separately from each other. The population that leaves the herd in e-commerce is difficult to define exactly, due to the information about quitting the e-commerce business is only found in news items and are reported separately for each case. The set of companies that first followed the herd and stepped out later is set to 10 companies due to the lack of information found.
3.3 Data collection
The database that is used consists of information from Orbis and
Annual reports from the companies selected. Orbis is a global office that gathers data from all around the world about private and public trading firms. They provide information about the structures, patents and financial information that is confirmed by third party due diligence. First, the Orbis database is filtered for public trading retail companies leaving only the relevant firms for the
population. The companies that have been selected are public companies. The financial data of these companies in this database is reported in a clearer and more complete manner than private firms. The Orbis database is updated
annually with the most recent financial information about six million companies in the retail industry. To complete the database that is used for the regression and t-tests, the total amount of years that a company operates online and the online revenues from 2013 to 2017 are added from annual reports of all the firms. This information is in most cases rounded off to the million or billions in sales, making it less accurate. In some cases, only percentages of sales are given by retailers to define their online sales. Secondly, the random number generator from Google is used to select random companies that enter the final sample. A total number of 120 companies are added to the final database. Finally, once all
firms are selected, the companies are individually researched for their online revenues and the year they started their online business.
3.4 Measurement
The measures used are compiled to percentages, making them suitable for comparison. Lambert and Larcker (1987) state that the most suitable variables to test performance of companies are Return on Sales and Return on Equity. In this comparison, due to the relevance of online sales as part of total sales, the measure of Return on Sales is used. The Return on Sales of the different groups are compared to each other to see if herd behavior positively or negatively influences performance.
Online adoption is measured as percentage online sales of the total sales of each firm. The years of online activity are derived from the company annual reports and indicate how many years ago the company started with online sales. For example, years of online activity with a value of 6 indicates that the firm started their online operations in 2011.
European and North American companies are separated with a 1 for
European firms and a 0 for North American firms. Industry type is indicated with a 1 for the clothing industry, 2 for the supermarket industry and 3 for the
electronics business.
Data indicators
Return on sales (Average 5 year profit / average 5 year total sales) *100
Years of online adoption (2017-year start online operations)
Average online adoption (Average 5 year online revenue / average 5 year total sales) *100
Region type 0= North America; 1= Europe
Early Adopters E-commerce adoption between 1998 and 2001
Late adopters E-commerce adoption from 2008 to 2013
Stop following All firms that started an online channel, but later ceased their
online sales
Moment of entry All companies ranked in their
moment of adoption to e-commerce
Profit growth The overall growth in profits from 2013 to 2017
Table 1: Data indicators
3.5 Reliability and validity
The measurement of the variables has been consistent and the same
standards have been used for every company to report their financial data. The data is all publicly available, making it convenient to replicate.
Return on sales is used to measure the performance of all the different variables. Creating a standardized method for comparison. A correlation table is used to examine the underlying connections between the variables.
The validity of the measurements is assumed due to the use of reports that are in compliance with accounting and financial reporting standards. Return on Sales is an accepted measure and performance indicator for firms (Lambert & Larker, 1987). The validity of the data is strengthened by the use of different industries. The findings can be compared to each other to validate the outcomes of the regressions. To strengthen the internal validity, control variables are added to the regression analysis. Variables like average online adoption and years of online activity are added. External validity of this research can be low, due to the limited sample size. The sample size is not representative for the total population.
3.6 Data analysis
The first hypothesis that is examined is the difference between early adaptors and the followers. The first movers are separated from the late adaptors. The first movers are determined as companies that started in e-commerce during the Internet hype of 1998 to 2001. The late adopters are the firms that adapted e-commerce after 2010. The difference is checked in a one-sample t-test, where the mean of the Return on Sales of the late adopters are tested against the mean of the Return of Sales of the early adopters. The early
region. American companies are compared to only American companies and European companies only to data of European companies. By comparing the percentage of profits on average from the periods, it is possible to see which strategy performs better. To standardize the measures the percentage of revenue from online relative to total sales is used.
The second statement that is derived from the literature is the moment of entry (Choi, 1997). The specific moment of entry is interesting to plot to see which data points are the most valuable in terms of performance. A comparison is made between the early and late adapters, but the middle group has yet to be included. By ranking all the companies in their order of moment of entry, it is possible to distinguish the best moment to adapt to the herd. For example, company X was the third in its industry to adapt to the e-commerce strategy and gets the number three assigned. The performance of the groups in different stages on the timeline can be examined to find perfect time to adapt to the herd. For this analysis an ordinary least squares linear regression analysis is used. The regression analysis shows the difference of adding the variable moment of entry to the model of control variables. Because these observations are all ranked, the top performers can be picked out of the sample and thus the best moment to enter the e-commerce market can be defined in a graph.
The last hypothesis focuses on the choice of a firm to stop following the herd if the strategy is not working for them. This relation is examined by performing a one-sample t-test. Companies whom started with e-commerce sales, but decided to stop are put into the test group. The performance of these firms is measured against companies that did continue with their e-commerce operations. The mean of the Return on Sales of the total population is tested against the mean of the Return on Sales of the test group.
By examining these hypotheses, it is possible to find answers for the research question: what are the long-term effects on performance for following the herd and to stop following the herd?
4. Results
The results of the different tests aim to answer the research question of what the long-term effects of herd behavior on the performance of a firm. The analysis is based on the random sample of 120 public trading retail companies, all
participating in the electronics, clothing or supermarket industry. These companies are derived from the Orbis database and filtered for their size and availability of data. First, the descriptive statistics and the interesting findings are discussed. The correlations table is discussed to identify correlating variables. Secondly, the results relating to the hypotheses are shown in tables and models to improve clarity of the results. Finally, some interesting remarks about the results are pointed out.
4.1 Descriptive statistics
The descriptive statistics indicate the normality in the variable return on sales in the sample, which is crucial to perform further analyses.
The descriptive test indicates no significance for Kurtosis and Skewness. The data indicates normality. The graph also illustrates the normality of the Return on Sales.
The data is split in two main groups, companies that originate from Europe and companies that originate from North America.
Descriptive statistics
Variable
N
Minimum Maximum Mean
Std.
Deviation
Average return on
sales
120
-5,5
21,7
4,068
4,45175
Average Online
Adoption
120
0
100
14,796
25,213
Years online
adoption
120
0
22
8,454
5,1036
Average profit
growth
120
-5,47
21,68
4,2543
4,53628
Table 2: Descriptive statistics
The descriptive statistics indicate a similar pattern for each industry and region. There are firms with relative high Return on Sales and firms with
negative returns in all the groups.
The two regions are tested separately. By splitting these groups, the cultural differences and adoption rates are not a factor. Both regions are approximately the same size in sample. North American companies are not expected to follow trends in Europe and vice-versa. The data of the North American group is more spread than the data of the European sample. It is
assumed that different industries do not copy each other in behavior, only within industries. Thus the industries are tested separately and differences between industries do not influence the results. The sample sizes of the industries are approximately equal.
4.2 Correlation between variables
The correlation table (table 3) shows the correlation between all the variables that are tested in the analysis for the different hypothesis. The
variables that are not normally distributed will be compared with the Spearman correlation. The variables that are normally distributed are compared with a Pearson correlation.
20
* Correlation significant at the 0,05 level ** Correlation significant at the 0,01 level
Table 3: Correlations
Correlation of Variables
Variable Correlation Average return on sales
Average online
adoption North America Supermarket Clothing Electronics
Years of online
adoption Moment of entry Europe Profit growth Average online
adoption Pearson correlation -0,214*
North America Spearman's Rho -0,071 -0,255**
Supermarket Spearman's Rho -0,182* -0,245** 0,21*
Clothing Spearman's Rho 0,14 0,254** -0,233** -0,471**
Electronics Spearman's Rho 0,21 -0,054 0,056 -0,431** -0,575** Years of online
adoption Pearson correlation -0,069 0,45** -0,172* 0,35 -0,075 0,18
Moment of Entry Spearman's Rho 0,087 -0,544** 0,151 -0,178* 0,098 0,081 -0,957**
Europe Spearman’s Rho 0,093 0,233** 0,689 -0,12* 0,21** -0,15 0,126 -0,093
The correlation table above shows significant correlations between the variables. The average online adoption correlates with all except the electronics industry. An interesting negative correlation is between Return on Sales and years of online adoption. These negative correlations indicates that the longer a firm is active in the e-commerce, the more it has a negative effect on their performance and the higher its online adoption, the lower the return on sales. Another interesting correlation derived from this analysis is the significant negative correlation between Return on Sales and the supermarket industry, indicating that this group has a much lower average Return on Sales than the rest of the industries. The clothing industry shows a significant correlation for
average online adoption, indicating that this industry has higher percentage online sales relative to total sales within this sample. Also the average years online are significantly higher than the other industries.
4.3 Early adopters performance versus late adopters performance Performance Early Adopters vs Performance Late Adopters
t-test for Equality of Means
t df Sig. (2-tailed) Difference Mean Difference Std. Error
Average ROS Europe -,550 28 ,586 -,8724 1,5850
North
America -,662 25,917 ,514 -,7564 1,1417
Table 4: T-test performance Early adopters vs. Late adopters
The t-test shows the comparison between the early adaptors (started between 1998-2001) and the late adaptors (started after 2008). The result indicates that there is no significant difference in performance between the early and late adapters in the Europe group with a significance of 0.591 . Equal
variances are not assumed, due to no significance of Levene’s test. The differences between the performance of the two groups are too small to be significant. For this test, only 6 early adopters were left in the sample. Tested
The t-test indicates no significant difference between the performance of the groups of the early adaptors in commerce versus the late adopters in e-commerce in North America. Equal variances are assumed in Levene’s test. The significant difference in means between the two groups is .713, which indicates no significant difference. For the North America group there are 12 observations for early adaptors and 18 observations for late adaptors.
4.4 The moment of entry and the performance of companies Coefficients: Dependent variable Average Return on sales
Variables
Unstandardized
Coefficients Model 1 Coefficients Model 2 Unstandardized
B Std. Error B Std. Error
(Constant) 5,406* (-2,530) 1,939 (4,362)
Industry type ,265 (2,333) ,099 (2,340)
Aantal jaren online actief ,000 (,093) ,197 (,222)
% av Online adoption -,041* (,019) -,042* (,019)
Supermarket -2,400 (2,323) -1,924 (2,374)
Electronics -1,086 (4,553) -,809 (4,563)
Moment of Entry - - ,095 (,097)
R Rsquared Adj. R squared
Model 1 .294 .086 .044
Model 2 .307 .095 .044
* Significance for 0,05
Table 5: Regression analysis moment of entry
To test the impact of the moment of entry for a firm to adapt to the herd, a regression analysis is used. The regression is divided in two separate models. The first model is only with control variables. The second model adds the moment of entry in the explaining of the total average Return on Sales, the performance measure. The moment of entry explains a small part of the total variable Return on Sales. The model shows that the moment of entry does not make a significant impact on the Return of Sales, or performance of a firm. With a significance of 0,332 it is not enough when tested with an Alpha of 0,10.
The R squared and adjusted R squared change slightly once the variable moment of entry is added to the model. The variable changes the R squared from 0.086 to 0.095, with a significance of 0.332. This is not enough at the alpha level of 0.10 to indicate a significant difference in the explanation of the model. The
moment of entry does help explain a very small part of the total Return on Sales for a firm, but not enough to be significant. Adjusted R squared does not change at all in these models.
The graph indicates the best moment to adapt to the herd for each industry. This helps clarify potential moments to choose the strategy and indicate where the potential best moment is for adaptation.
Figure 4: relation between the moment of adoption and the mean average return on sales.
Figure 4 shows the performance outputs for each moment of adoption. The results are widespread.
The first industry, the clothing industry, indicates a spread result in the returns. The highest returns are at the 37th mover into the market, which
The second industry, the supermarket business, shows that being the first mover or the 32th result in higher return on sales.
The third industry, the electronics industry indicates a spike in Return on sales in the moment of 11th entry after the first mover, just as the 36th (late)
mover enters the market with higher returns. However, these findings are only derived from visual analysis of the graph. No significance was found for these moments.
4.5 Leaving the herd strategy
The sample of companies that left the herd was relatively small. Only 11 companies could be identified that had chosen to abandon the strategy of their online sales. These companies have a mean of 1.214 and standard deviation of 1.762. The groups of companies that stick to the strategy of e-commerce have a mean of 4.328 and standard deviation of 4.6052.
This t-test indicates the comparison of the average return on sales from the sample of companies that choose to first join the herd, but later decided to abandon this strategy. The other companies consist of the rest of the complete sample of 120 companies. The t-test indicates a significant difference in
performance for the leaving companies compared to the companies that have stayed with the herd. The Levene’s test for equality of variances is significant, so equal variances are not assumed. The t-test indicates a significant difference with a significance of 0,000 at compared to our sample. The performance of
companies that leave the herd is significantly worse than companies that stay with this strategy.
5. Discussion and conclusion
The main purpose of this research is to indicate what the long-term effects are of the strategic choice to follow the herd or leave the herd. To answer this question, the research question is brought down to three hypotheses.
5.1 Main findings
Following the herd
The current literature suggests different possibilities and chances for late movers and early adapters. The early adopters and first movers have the
advantage of being pioneers in their new methods and strategy. This gives them the advantage to expand their operations faster (Lieberman & Montgomery, 1988). The downside to pioneering is that there is no great demand for the new product or service. This has to be promoted and build up by the pioneers, which brings additional costs. Late movers can take advantage to wait out this process and enter a more mature market at lower costs (Gal-Or, 1985). Another reason why companies join the herd is due to pressures from competitors. They fear to miss out on the potential revenues that competitors derive (Deephouse, 1996). Isomorphic pressures cause companies to conform their strategy to that of their competitors (Banerjee, 1992).
The results of comparing the early adopters and late adapters do not confirm the arguments of advantages or disadvantages for early adopters. The results of the t-test of the European and the North American market were not significant. Indicating that there is no difference in performance between the two groups. This could indicate that both strategies have their benefits, which in the end balance each other out. This effect is shown in figure 5.
Figure 5: Trade-off early adapter advantages and late adapter advantages.
The late adopters don’t have the high cost of child diseases and can copy the work of the early movers. The early movers have the advantage of being one of the first in the market and expanding their market share rapidly (Gal-Or, 1985). Previous research states that the moment of entry is highly dependent of the average technological cost to adapt to the market. Early pioneers can keep their advantage over late followers by using their superior capabilities to stay
innovative and use the firm specific capabilities to sustain a competitive advantage (Ruiz-Ortega & García-Villaverde, 2008).
The result could also be specific for this sample of e-commerce adoption. However, the results of this sample do support the research of Suarez & Lanzolla (2005). They state that first mover advantages and late followers advantages are overstated. The firms themselves can make the difference in their performance on an individual level, despite the moment of their entry.
The results of the regression analysis show the impact of the moment of entry to the e-commerce market on the total performance of the firms. The moment of entry shows to have no significant impact on the performance of the companies. The performance is only for a very small part dependent on the moment of entry, as both models concluded. A possible explanation is the fact that most of the companies that operate in the online environment have brick and mortar stores as their main generator of revenue. They rely on average for 4% of their revenue on their e-commerce operations. Figure 4 shows the
different performances for all the industries with the different moments of entry. The graph clearly shows how spread the returns are with each moment of entry, there seems to be no clear moment to get an advantage over the competition.
Leaving the herd
The current literature suggests that abandoning a chosen strategy is in most of the cases due to the lack of profits from a project (Ewusi-Mensah &
Przasnyski, 1991). Once projects become too costly to develop further or when the costs to get goods to the customer become too high, companies can choose to stop their online activities and leave the herd. However, the reason for
abandoning the e-commerce strategy has not been taken into account in this research. Differentiating strategies to make a product or service more exclusive are used in all the industries (Porter, 2011). Abandoning the online strategy makes the products less widely available, which is negative for the revenues of a firm. Although the customers percieve the stores without online activities as less trustworthy than stores that do have online presence (Everard & Galletta, 2005). Companies fear the quality issues that arise with the delivery of products.
The results of the t-test indicate a significant difference between the companies that are still part of the herd and companies that left the herd. The performance of the group that left the herd was significantly lower than the companies in the herd. This indicates that leaving the herd or differentiating from it in e-commerce perspective is not good for the performance of a firm.
Taking this into account, the abandoning of projects is mostly due to a
disappointing development in sales (Ewusi-Mensah & Przasnyski, 1991). It
indicates what other companies state as well: that the current practise of online sales is not always profitable. Also these findings confirm the theory of Ewusi & Przansyski (1991), if the companies abandon the e-commerce project because of their lack of liquidity. If this theory holds,tThe lower performance of the leaving group is a self-fullfilling prophecy. It does however contradict the theory of Porter (2011) if the firms quit their online activities due to strategy. Which states that differentiation is a strategy to be different than the rest of the herd.
5.2 Scientific contribution and management implications
This research adds to the gap in current literature that exists for herd behavior. Most of the research is done cross-sectional or is based on
psychological factors derived from literature (Deephouse, 1998; Roa et al, 2001; Strand & Macy, 2001). This reseach is taking a longitudinal approach and is explaining the difference in performance, where previous literature only covered market share (Bowman & Gatignon, 1998). This research is the first that uses profitability measures to compare the performance between companies that are early adapters or late adapters. Adding to that, the performance consequences of leaving the herd have never been researched. Previous research only covers theoretical explainations for differentiation strategies (Porter, 2011) or through questionaires to confirm theories (Ewushi-Mensah & Przasnyski, 1991).
The results of this research can be of value in the work field of managers. Managers have to make strategic choices in order to stay relevant. This research indicates that there is no significant difference in performance between late adapters and early adapters. Which indicate that both these groups have their advantages and disadvantages, but in the end these weigh out against each other. This research attempted to indicate the performance difference between late followers and early adapters, to indicate the right moment to enter the market. The findings indicate that performance is more related to company specific performance than to moment of entry to the market. This creates the
opportunity that it is never to late to adapt to the competitors operations or enter a new market. If the value proposition of the company is good, the performance can be positive.
This research does indicate that once a company decides to follow the herd, it is not a good choice to leave it again. The performance of the firms that left the herd was significantly lower than those that stayed with the herd. The performance data that is gathered originates from 2013 to 2017, which indicates that in this time the strategic choice would be of negative influence on the
5.3 Limitations
Although the tests done in this research confirm what existing theories state, there are a few limitations to the test. First of all, the sample that has been taken only accounts for large public retailers. Most of them have been relying on their brick and mortar sales for more than 25 years. The e-commerce adoption is still in its infancy, with average adoption in the sample of four percent. Once e-commerce adoption starts to gets more general, this might impact the revenues and profits a lot more. Including other smaller companies and more internet focussed firms might help for a better view of the early adopters and late movers in the industry.
Another limiting factor is the availability of information about online operations. Not all public firms provide information about their online sales, which led to some firms that needed to be excluded from the sample.
The sample of the firms that abandoned their online strategy was hard to find. The sample is not random, due to hand picking of the companies that fitted this group. This influences the reliability of this group, although other methods of finding these companies were not available. Leaving the herd performance needs to be tested in future research to confirm the findings of this test.
Looking at the regression model, a relative small part of the variance (a difference in R squared of 0,009) is explained by the present model. This
indicates that there are mainly other factors that have a much greater impact on performance than moment of entry or online adoption has. Possible other variables could be for example firm size or macro economic factors.
The leaving strategy for other companies might be different than it is in this sample. Looking at the companies that were in the sample of ten, a clear pattern of companies that have difficulties with staying profitable. The online operations were simpely too costly to maintain and most of the companies are electronics companies that choose to stay out of the internet sales and
differentiate on services. If more companies were taken into account, the
playfield might be more leveled and the performance of this group might not be as bad.
possibly lead to more significant findings. Because this sample only consists of 120 firms, the effect of outliers is proportional. There are two companies that have a return on sales of twenty percent, which is five times the average return on sales. Putting these high performing companies in separate groups would be interesting for future research.
5.4 Suggestions for future research
The first proposal is to take a look at other ways to conclude herd behavior within companies. E-commerce is one strategy to look at, but other strategies like store locations or other strategic choices where herd herd behavior may play a role. Testing these against the long term performance of firms might increase the knowledge about the long term effects of herd behavior.
The correlation between performance and online adoption shows ground for new research in this direction. The negative correlation between online adoption and Return on Sales in this sample indicates that the higher the online adoption, the lower the performance. This could be investigated further to prove the significance of these findings. This might strenghten the theory of herd behavior, because this indicates that this strategic choice is influenced by herd behavior and is not a path choosen for extra profits.
Another proposal for future research is to look into the motivations of firms to adapt to a certain strategy. This could strengthen the argument for herd behavior if firm managers explain why they have taken certain strategic steps. Implementing this with the moment of entry could make an interesting research and insight in the late followers and early adaptors motivations.
A greater sample with more focus on late followers and early adapters and their specific performance linked to this strategic choice would make an interesting research to the best moment of entry.
If possible and with the correct data, it would be interesting for future research to see what the specific strategic choice contributes in extra profit. If these profits can be measured and compared to different groups, the conclusion can be drawn of which companies choose the strategy because other firms do so and which companies adapt the strategy because they see a revenue model.
5.5 Conclusion
This research contributes to the phenomenon of herd behavior in strategic decisions for firms. To effectively answer the research question on what the long-term effects on profitability are, early adapters and late followers are tested for performance, as well as the moment of entry. According to this sample there was no difference in performance between early and late adopters. Also, the moment of entry did not make a significant difference. This sample has limitations, which future research has to expand on. The second part of the research question however, can be explained. Leaving the herd has shown a significant negative effect on performance. The returns of firms that choose their own path and deviated from the herd are worse off in terms of performance. More research is required to accurately state the long-term effects of herd behavior on performance. Also leaving the herd needs to be investigated further to assure these findings.
Bibliography
Abrahamson, E., & Rosenkopf, L. (1993). Institutional and competitive bandwagons: Using mathematical modeling as a tool to explore
innovation diffusion. Academy of management review, 18(3), 487-517. Adams, D. A., Nelson, R. R., & Todd, P. A. (1992). Perceived usefulness, ease
of use, and usage of information technology: A replication. MIS quarterly, 227-247.
Banerjee, A. V. (1992). A simple model of herd behavior. The quarterly journal of economics, 107(3), 797-817.
Berndt, E. R., L. Bui, D. Reiley, and G. L. Urban (1995), "Information
Marketing and Pricing in the U.S. Anti-Ulcer Drug Market," American Economic Review, Papers and Proceedings, 85 (2), 100-105.
Bowman, D., & Gatignon, H. (1996). Order of Entry as a Moderator of the Effect of the Marketing Mix on Market Share. Marketing Science, 15(3), 222-242. Choi, J. P. (1997). Herd behavior, the" penguin effect," and the suppression of
informational diffusion: an analysis of informational externalities and payoff interdependency. The Rand Journal of Economics, 407-425. DiMaggio, P., & Powell, W. W. (1983). The iron cage revisited: Collective
rationality and institutional isomorphism in organizational fields. American sociological review, 48(2), 147-160.
Deephouse, D. L. (1996). Does isomorphism legitimate?. Academy of management journal, 39(4), 1024-1039.
Deephouse, D. L. (1998). Legitimacy. Encyclopaedia of Management, eds. CL Cooper and C. Argyris, Oxford: Blackwell.
Everard, A., & Galletta, D. F. (2005). How presentation flaws affect perceived site quality, trust, and intention to purchase from an online store. Journal of Management Information Systems, 22(3), 56-95.
Ewusi-Mensah, K., & Przasnyski, Z. H. (1991). On information systems project abandonment: an exploratory study of organizational practices. MIS quarterly, 67-86.
Gal-Or, E. (1985). First mover and second mover advantages. International Economic Review, 649-653.
Gefen, D., & Straub, D. W. (2000). The relative importance of perceived ease of
use in IS adoption: A study of e-commerce adoption. Journal of the
association for Information Systems, 1(1), 8.
Goolsbee, A. D., & Klenow, P. J. . (2018). Internet Rising, Prices Falling: Measuring Inflation in a World of E-Commerce. . Stanford University working paper, 1-6.
Glynn, M. A., Barr, P. S., & Dacin, M. T. (2000). Pluralism and the problem of variety. Academy of Management Review, 25(4), 726-734.
Greve, H. R. (1995). Jumping ship: The diffusion of strategy abandonment. Administrative Science Quarterly, 444-473.
Hambrick, D. C., Finkelstein, S., Cho, T. S., & Jackson, E. M. (2004). Isomorphism in reverse: Institutional theory as an explanation for recent increases in intraindustry heterogeneity and managerial discretion. Research in organizational behavior, 26, 307-350.
Heugens, P. P., & Lander, M. W. (2009). Structure! Agency!(and other quarrels): A meta-analysis of institutional theories of organization. Academy of
Management Journal, 52(1), 61-85.
Hitt, L. M., & Brynjolfsson, E. (1996). Productivity, business profitability, and consumer surplus: three different measures of information technology value. MIS quarterly, 121-142.
Kavanagh, M. H., & Ashkanasy, N. M. (2006). The impact of leadership and change management strategy on organizational culture and individual acceptance of
change during a merger. British Journal of Management, 17(S1).
Kennedy, R. E. (2002). Strategy Fads and Competitive Convergence: An Empirical Test for Herd Behavior in Prime‐Time Television Programming. The Journal of Industrial Economics, 50(1), 57-84.
Kostova, T., Roth, K., & Dacin, M. T. (2008). Institutional theory in the study of multinational corporations: A critique and new directions. Academy of management review, 33(4), 994-1006.
Lambert, R. A., & Larcker, D. F. (1987). An analysis of the use of accounting and market measures of performance in executive compensation contracts. Journal of Accounting research, 85-125.
Lieberman, M. B., & Montgomery, D. B. (1988). First‐mover advantages. Strategic management journal, 9(S1), 41-58.
Mellahi, K., & Johnson, M. (2000). Does it pay to be a first mover in e. commerce? The case of Amazon. com. Management Decision, 38(7), 445-452.
Min, S., Kalwani, M. U., & Robinson, W. T. (2006). Market pioneer and early follower survival risks: A contingency analysis of really new versus incrementally new product-markets. Journal of Marketing, 70(1), 15-33. Min, S., & Wolfinbarger, M. . (2005). Market share, profit margin, and marketing
efficiency of early movers, bricks and clicks, and specialists in e-commerce. Journal of Business Research,, 58(8), 1030-1039.
Porac, J. F., Thomas, H., & Baden‐Fuller, C. (1989). Competitive groups as cognitive communities: The case of Scottish knitwear manufacturers. Journal of Management studies, 26(4), 397-416.
Porter, M. E. (2011). Competitive advantage of nations: creating and sustaining superior performance (Vol. 2). Simon and Schuster.
Prahalad, C. K., & Bettis, R. A. (1996). Dominant logic. Managing the Multibusiness Company, 398-420.
Rao, H., Greve, H. R., & Davis, G. F. (2001). Fool's gold: Social proof in the initiation and abandonment of coverage by Wall Street analysts. Administrative Science Quarterly, 46(3), 502-526.
Rolfe, G. (2006). Validity, trustworthiness and rigour: quality and the idea of qualitative research. Journal of advanced nursing, 53(3), 304-310.
Ruiz-Ortega, M. J., & García-Villaverde, P. M. (2008). Capabilities and competitive tactics influences on performance: Implications of the moment of entry. Journal of Business Research, 61(4), 332-345.
Shankar, V., Carpenter, G. S., & Krishnamurthi, L. (1998). Late mover advantage: How innovative late entrants outsell pioneers. Journal of Marketing research, 54-70.
and adaptive emulation. American journal of sociology, 107(1), 147-182. Suarez, F., & Lanzolla, G. (2005). The half-truth of first-mover advantage.
Van de Ven, A. H., & Poole, M. S. (1995). Explaining development and change in organizations. Academy of management review, 20(3), 510-540.
Volberda, H. W., van den Bosch, F. A., Flier, B., & Gedajlovic, E. R. (2001). Following the herd or not?: Patterns of renewal in the Netherlands and the UK. Long Range Planning, 34(2), 209-229.
Xia, J., Tan, J., & Tan, D. (2008). Mimetic entry and bandwagon effect: the rise and decline of international equity joint venture in China. Strategic
