Learning about the Learning Curve: Technology Vendor Strategies and Adoption Outcome

Abstract

Learning to use new technologies takes time and repeated practice. The value of such technologies is uncertain. So customers have to make prolonged learning effort without knowing their returns. Economic theories on vendor strategies have not modeled customers' post-purchase learning processes. We consider two aspects of learning: 1) Customers learn to use the software according to a learning curve. 2) The shape of the learning curve is unknown to customers, but can be learned through experience. We investigate a software vendor's optimal licensing strategy and strategies to reduce learning difficulties as well as the resulting customer adoption outcome.
Prior literature recommends subscription-based licensing for promoting the adoption of uncertain technologies because it limits customers' initial financial commitment. We show that while subscription licensing can create more initial adoption under some conditions, perpetual licensing leads to more continued usage and lower customer churn. Surprisingly, the long-term adoption rate can be higher under perpetual licensing. Extending the subscription term length can reduce customer churn. Moreover, learning curve uncertainty can make either perpetual or subscription-based licensing more profitable for the vendor but through different mechanisms. The vendor may not offer the licensing scheme that maximizes long-term adoption. Finally, the vendor benefits from investment that reduces the initial learning barrier or accelerates customer learning. Although both strategies reduce learning difficulties, they are not necessarily substitutes for each other. Investment in one can increase the return to investment in the other. Vendor's incentive to reduce learning difficulties can be higher under perpetual or subscription-based licensing.

Keywords: Learning curve, Stochastic learning, Bayesian analysis, Software licensing