Analyzing Software as a Service with Per-Transaction Charges

Publication Type

Journal Article

Publication Date



Software as a Service (SaaS) delivers a bundle of applications and services through the Web. Its on-demand feature allows users to enjoy full scalability and to handle possible demand fluctuations at no risk. In recent years, SaaS has become an appealing alternative to purchasing, installing, and maintaining modifiable off-the-shelf (MOTS) software packages. We present a game-theoretical model to study the competitive dynamics between the SaaS provider, who charges a variable per-transaction fee, and the traditional MOTS provider. We characterize the equilibrium conditions under which the two coexist in a competitive market and those under which each provider will fail and exit the market. Decreasing the lack-of-fit (or the cross-application data integration) costs of SaaS results in four structural regimes in the market. These are MOTS Dominance → Segmented Market → Competitive Market → SaaS Dominance. Based on our findings, we recommend distinct competitive strategies for each provider. We suggest that the SaaS provider should invest in reducing both its lack-of-fit costs and its per-transaction price so that it can offer increasing economies of scale. The MOTS provider, by contrast, should not resort to a price-cutting strategy; rather, it should enhance software functionality and features to deliver superior value. We further examine this problem from the software life-cycle perspective, with multiple stages over which users can depreciate the fixed costs of installing and customizing their MOTS solutions on site. We then present an analysis that characterizes the competitive outcomes when future technological developments could change the relative levels of the lack-of-fit costs. Specifically, we explain why the SaaS provider will always use a forward-looking pricing strategy: When lack-of-fit costs are expected to decrease (increase) in the future, the SaaS provider should reduce (increase) its current price. This is in contrast with the MOTS provider, who will use the forward-looking pricing strategy only when lack-of-fit costs are expected to increase. Surprisingly, when such costs are expected to decrease, the MOTS provider should ignore this expectation and use the same pricing strategy as in the benchmark with invariant lack-of-fit costs.


software as a service, game theory model, pricing based on transactions, competitive strategies, lack-of-fit costs, economies of scale


Computer Sciences

Research Areas

Information Systems and Management


Information Systems Research





First Page


Last Page







INFORMS (Institute for Operations Research and Management Sciences)