The optimal software licensing policy under quality uncertainty
Proceedings of the 5th international conference on Electronic commerce - ICEC '03
The rapid rate of standard software development, and the need of users to stay current, has unleashed unprecedented levels of process and product innovations in the software industry. A new service model has emerged which delivers application software and services over the Web on a lease or subscription basis. Software vendors such as Sun, Oracle, and Microsoft have already adopted this innovative business model. They have expanded their sales offerings with lease contracts that augment their
... aditional one-time purchase transactions. Our paper studies the optimal licensing policy of a software vendor that uses that business model. We look at software vendors that are both selling (at a posted price) or leasing their products where as lessor they guarantee that the lessee will always have the latest version of the software on their desktop. We address some of the specific issues of implementing this policy at the packaged software market, including the impact of network externality, negligible marginal production costs, and upgrade compatibility. We show that by properly defining their pricing structure, software vendors can segment the market and realize effective seconddegree price discrimination and show how and when software vendors can maximize their profits through the use of this new licensing policy. Software vendors continuously invest in research and development to improve product quality in response to the threat from potential entrants. There are several important dimensions to software quality including speed, compatibility with available operating systems, functionalities, user interface, ease of learning, warranty, service and support, and other characteristics that affect the users' valuation of the product. It may be the perceived quality, which includes both real improvement and a successful marketing component. Potential buyers can assess the quality of existing software, yet the quality of the next version (upgrade) is uncertain. Users who choose early on the lease option with periodic upgrades are exposed to the risk of buying into future upgrades with unknown quality. On the other hand, users who just purchase the software can decide on buying an upgrade after the quality of the new version has been realized. Software can be used for a period of time without replacement, although its value may depreciate. In this sense, it is a kind of a durable good. Yet software as a commodity has some special characteristics that differentiate it from other durable goods. First, it is hard to resell or appropriate because of intellectual property rights. There is no other source of a license but the original vendor, the retailers, or service providers. A second-hand market like that for used cars, therefore, does not exist for software. Second, with the development of information technology, it is easy to improve the value of already installed software through upgrades without interfering with the original customization. In addition, the use of software has a strong network externality coming from two sources: (1) direct externality: the larger the population of consumers using the software, the easier it is to get information, training, and communication about the product, so it becomes attractive to more users; (2) indirect externality: supports and other complementary products also increase with the population of adopters. This network externality, therefore, is like the "chicken-and-egg" problem (Katz and Shapiro 1986). As a result, selling and leasing software have some common features with selling and leasing other durable goods and some special characteristics, too.