Value Creation in E-Business and Financial Performance: Researching Polish Online Companies with Amit and Zott’s Model.



The paper presents the outcomes of the survey of the managers of Polish internet companies with the aim of verifying to what extent the e-business value creation model developed by Amit and Zott could be used to explain different profitability levels among firms. The study included 150 businesses with the largest representation of internet retailers and service providers. The major outcome of the study was developing logistic regression model that allowed to establish which variables were statistically significant predictors of operational return on sales ratio. It implied that out of four elements comprising Amit and Zott’s model only Complementarities were closely linked with improved financial performance. The two other elements – Customer Lock-in and Novelty – were not viable predictors while the forth component (Efficiency for Customer) was negatively related to the performance metric.

Keywords: business models, value creation, e-business, e-commerce, Internet, CEE, Poland
TrackNew Technologies and E-marketing


The objective of the study was to establish to what extent the value creation model originally proposed by R. Amit and C. Zott could be linked to financial performance of internet companies (Amit, Zott 2001). The presence of various elements of Amit and Zott’s concept was measured with Likert-type 7-point scale comprising a list of statements pertaining to the four value drivers of the model incuding: novelty, lock-in, complementarities and efficiency. According to Amit and Zott these four categories represent groups of factors that can enhance the total value created by e-business. In particular they refer the following specific kinds of business solutions:

  • Novelty is defined by the level of uniqueness of goods or services offered by a firm with regard to how the customers’ needs are satisfied and what those needs are; thus novelty can be achieved by devising new ways of satisfying existing needs or finding and addressing entirely new needs.
  • Lock-in concerns the various kinds of costs that customers need to bear if they were to replace the firm’s offering with one of competitors; it is assumed that the higher are such switching costs the customer lock-in is tighter.
  • Complementarities are determined by the extent of completeness of customer needs fulfillment by the firm and its business partners.
  • Efficiency pertains to the benefits that customers can reap from savings in time, effort and financial costs that are brought about by the firm’s offer.

There were so far only a few attempts to test the Amit and Zott’s model empirically with quantitative approach, though the outcomes were rather inconclusive and context sensitive. Christensen and Methlie (2003) having researched implementations of e-business solutions in traditional companies noted no significant improvement in financial performance, which was attributed to the early stage of the implementation and the lack of time for the positive effects to manifest. According to the later research of Amit and Zott (2007) Novelty positively influenced financial performance while Efficiency did so only during a period of resource scarcity. In another paper, the same authors studied how the fit between Novelty- and Efficiency-centred business models with different product market strategies can enhance firm performance (Amit, Zott 2007).  Using a population of general companies (including both internet enabled and traditional businesses), Malone et al. found that some models perform financially better than others (2005).

In light of the above, if the concept of the value creation model in e-business were to have considerable practical merit it could be proposed that “the increased involvement of a company in any of the four value drivers results in its improved financial performance”. To test this hypothesis it was necessary to develop adequate measures of elements of value creation model and competitive performance, which will be discussed in the subsequent section.

Research methodology

The data on characteristics of business models employed by internet companies were collected through CATI survey in August 2012. The interviewed respondents were the managers of Polish firms which were utilizing Internet as a distribution channel for retailing and services with the exception of major Internet portals, advertising and web design agencies, media brokers, telecommunications companies, banks, insurers and operators of large popular news and lifestyle portals. The exclusions were made on the assumption that those companies employed so unique and diverse value creation mechanisms that making meaningful comparisons among them might not have been possible based on survey data. The net sample of 150 units was drawn at random from a database of major internet companies compiled by the authors of the study from several available rankings and listings of various types of internet businesses operating in Poland. The final set of studied companies was made up in 57% of retailers and 43% of service providers, with 63% of them having sales of tangible products as the main revenue stream, 17.3% generating most incomes from sales of virtual products and 16% relying above all on proceeds from advertising.  Around 25% of businesses generated more than half of their sales outside of the internet.

Since they are multidimensional and difficult for direct measuring, the four sources of value in the model could be thought of as constructs or latent variables and evaluated accordingly, with appropriate multi-item Likert scales. Building on existing works in management science, both conceptual and empirical, and drawing from their own experience the authors came up with 24 statements that were listed in the Table  1.

Table 1: The list of statements of the 7-point Likert scale used for measuring the four sources of value creation

Value creation source: Novelty

Only a few firms offer solutions similar to ours
In markets we operate in we are recognized as pioneers
By offering our solutions we shape needs or behaviors of customers
In large part customers choose our solutions for the sake of their innovativeness
Our solutions have been imitated by competitors
Our key business partners do not cooperate with our competitors

Value creation source: Lock-in

Resigning from our offer and changing to our competitors’ brings about high switching costs to our customers, such as extra time, effort or financial expenses
It happens that customers are not fully satisfied with our offer but they stay with us due to switching costs
We provide our customers with personalized solutions
Most of our customers make use of our personalized solutions
We consider it important to maintain for as long as possible even these customers who are less profitable
Regular customers are rewarded through loyalty programs and other measures
We have implemented specific mechanisms for maintaining customers
Our key partners have strong impact on uniqueness of our offer for customers
An important criterion of selecting our business partners is enhancing our capacity for maintaining customers

Value creation source: Complementarities

Our offer is among the most comprehensive in the industry
Our firm has been systematically widening the extent of customer needs that we are able to satisfy
Our key partners have strong influence on how comprehensive our offer is
Our customers are choosing our offer for the attractiveness of the available complementary products
Customers try to use our solutions together to benefit from synergy effects
Value creation source: Efficiency
Our solutions allow customers to take advantage of savings in time and effort
Because of our solutions customers can solve their problems more easily
Internet grants customers more efficient access and use of our products than traditional channels
Our partners have significant contribution to savings in time and effort afforded customers by our offer

The original set of 24 individual variables was reduced with principal component analysis to a more manageable set of 8 composite variables representing various dimensions of the four theoretical constructs. Factor analysis also solved the problem of multicollinearity  among variables which tends to inflate standard errors in modeling. The 8 components were then used as independent variables in regression modeling. Using factor loadings on individual variables for interpretation, the following components were isolated: (1) novelty: customer perspective, (2) novelty: competitors and business partners perspective, (3) lock-in: personalization, (4) lock-in: exit barriers, (5) lock-in: loyalty programs, (6) lock-in: cooperation with business partner, (7) complementarities, (8) efficiency.

As can be seen, the constructs Complementarities and Efficiency were represented by a single component variable each, while the construct Novelty had two independent dimensions. The most complex in structure was Lock-in which was made up of four apparently independent aspects. The components extracted with exploratory factor analysis allowed to retain about 60% of the original variance.

As an indicator of financial performance served the operational return on sales ratio obtained by dividing operational income from transactions on the internet by sales achieved on the internet. To increase item reply rate from survey participants, the variable did not have open-ended format but was categorized with four predetermined answer classes. To be usable in binary logistic regression the variable was dichotomized with the value of 0 denoting firms with operational returns on sales no greater than 20% (67.7% of the sample) and the value of 1 indicating the companies that had the ratio of more than 20% (33.3% of the sample). The split point of 20% was chosen to yield two groups with possibly comparable sizes – it transpired that any other merger of categories in original variable would result in more disproportionate subgroups.

To obtain the model with the best fit to the empirical data and possibly the best prognostic qualities it was decided to use binary logistic regression with backward elimination of predictors to exercise more control over the suppressor effect. The original set of independent variables included 8 components representing various aspects of value creation model supplemented with several terms pertaining to different characteristics of the firms encompassing number of employees, percentage of sales from transactions on the internet, percentage of loyal customers, year of funding and year of starting operations on the internet.

Research results

The regression procedure finished at the 9th step, after eliminating 8 statistically insignificant candidates for predictors, resulting in the model with the following goodness-of-fit and pseudo R-Square measures (Table 2.)

Table 2: Goodness-of-fit measures for the outcome logistic regression model

Step Chi-square Sig. -2 log-likelihood Nagelkerke R Square
9 45.735 <0.001 145.219 .365

The overall model provides statistically significant improvement in predicting group membership of the companies (i.e. low versus high sales profitability) with error reduction of around 36.5% in comparison to the baseline solution containing only fixed term and no independent variables.  The classification table contains a comparison of groupings of firms using regression equation and the baseline setting when all observations were assigned to the more numerous category of the dependent variable.

Table 3: Outcomes of classification of the participant firms into groups with low and
high profitability using regression model and baseline knowledge of the dependent variable distribution



ROS > 20%

Percentage Correct



Baseline classification

ROS> 20%








Overall Percentage


Classification with the final model

ROS> 20%








Overall Percentage


The classification table indicates that predicting group membership of companies with the logistic regression model results in 75.3% of correctly classified cases compared to 66.7% of rightly predicted memberships when all firms were considered to be low profitability (0). Thus, the model offers a considerable improvement in predicting capacity over baseline scenario and could be potentially useful for practical applications. The variables included in the model and their coefficients are shown in the Table 4.


Standard Error












Percentage of total revenue from internet transaction







Number of employees





















The final regression equation contains 4 statistically significant predictor variables. Interestingly, out of 8 different aspects of the value creation model only 2 display sufficiently strong association with financial performance to be used along other variables in the regression function. Quite unexpectedly, the Efficiency variable remains in negative relationship with financial performance with higher scores on Efficiency diminishing odds of a company falling into higher profitability category by the factor of 0.529. Other variables with higher values tend to increase the odds of a firm being classified as highly profitable. The strongest factor here is the number of employees which with each progressing category on the scale increases the odds by more than twofold (2.559). Involvement in Complementarities and Percentage of Total Revenues from Internet Transactions seem to be equally important stimulants of profitability with similar impact on odds ratio of 1.8 following one unit increase on their respective scales with other factors remaining equal.

The above findings seem to give only limited support to the hypothesis of the study, as only one of four drivers comprising the Amit and Zott’s value creation model in e-business (Complementarities) seems to be positively associated with financial performance of surveyed companies and another driver (Efficiency) exhibits what appears to be an inverse relationship.


The lack of the two variables (Novelty, Lock-in) representing e-business value creation model in the regression may suggest that this concept could only have limited potential as a conceptual lens through which to explain success and failure of business models. One relevant caveat though could be the idiosyncratic features of the studied internet companies, which even though were representative for Poland, included mostly e-retailers and providers of rather uncomplicated services. It may well turn out that explanatory power of the value creation framework could be more potent if applied to analyze more complex business models, like those represented by the companies excluded from the sample.

The fact that Efficiency remained in negative relationships with operational ROS doesn’t necessarily have to imply causal association. And even if cause-and-effect link did exist it is possible to propose at least two plausible explanations.

The companies which offerings allowed the consumers to solve their problems quickly, possibly also cheaply and reliably, may have had taller operational cost structures (for instance had to employ more staff or use more expensive network infrastructure) and consequently had to accept lower sales profitability. Such situation didn’t have to entail weaker financial results and worse market position as only one metric was used to evaluate financial consequences of involvement in e-business, which is an apparent limitation of the study. It can be easily imagined that lower ROS ratio could be compensated by higher volume of sales to achieve higher profitability of the company.

Another explanation could lie in very intense competitive rivalry within the industries in which some companies had to operate. In such conditions advanced and comprehensive solutions increasing efficiency for customers could be forced upon a firm by strong pressure from rivals and as such could be a necessary measure to remain in the market and not a source of competitive advantage translating into increased mark-ups and profits. Unfortunately the data collected in the survey was not sufficient to assess the intensity of competitive struggle in each firm’s market environment.

The substantial positive effect of Complementarities on financial performance could be linked to economies of scope which result in lower cost structure and higher unit profitability due to offering wide range of products addressing a group of related needs. The other partial explanation could be offered by considerable costs of customer acquisition in internet channels.

Among all significant predictors the strongest one was the company size as measured by the number of employees with larger firms tending to display higher levels of profitability. It may imply that certain economy of scale mechanisms could be responsible but the fact that the studied businesses operated in various markets, compounded by the lack of knowledge about their pertinent characteristics, makes such a conclusion debatable.


  1. Amit, R., Zott, Ch. (2001). Value Creation in E-Business, Strategic Management Journal, 22 ,493-520.
  2. Amit, R., Zott, Ch. (2007). Business Model Design and the Performance of Entrepreneurial Firms, Organization Science, 2 (18), 181–199.
  3. Amit, R., Zott, Ch. (2008). The Fit between Product Market Strategy and Business Model: Implications for Firm Performance, Strategic Management Journal, 29, 1–26.
  4. Amit, R., Zott, Ch. (2012) Creating Value through Business Model Innovation, MIT Sloan Management Review, 53, 41-49.
  5. Christensen, G.E., Methlie, L.B. (2003). Value Creation in eBusiness: Exploring the Impacts of Internet-Enabled Business Conduct, 16th Bled eCommerce Conference eTransformation, Bled, Slovenia, June 9-11, 2003, 27-43.
  6. Malone, T.W. Weill, P. Lai, R.K. D’Urso, V.T. Herman, G. Apel, Th.G. Woerner, S.L. (2006), Do Some Business Models Perform Better than Others?, MIT Sloan Working Paper 4615-06.

One thought on “Value Creation in E-Business and Financial Performance: Researching Polish Online Companies with Amit and Zott’s Model.

Comments are closed.