Optimal Release Strategy for the Competing Software Vendors Based on Word-of-Mouth Effect

ABSTRACT

In the software industry, early release or delayed release are important decisions for the competing software vendors in the presence of the word-of-mouth (WOM) effect. This paper explores the optimal release decisions of two competing software vendors in two periods according to their estimations about WOM generated by their products. We find that in most scenarios, the firm expecting to have advantaged WOM adopts what is called an early-release strategy, whereas the firm expecting to have disadvantaged WOM adopts a delayed-release strategy. Of interest, when one firm’s expected WOM is sufficiently disadvantaged, it will also adopt an early-release strategy as long as the competitor’s expected WOM is sufficiently advantaged, which will never occur in a monopoly setting. When both firms’ expected WOM are negative and not too small, a counterintuitive equilibrium may appear in which the firm expecting to have disadvantaged WOM adopts the early-release strategy and the other expecting to have advantaged WOM adopts the delayed-release strategy. Two firms will be caught in a prisoner’s dilemma in the duopoly market when both adopt the delayed-release strategy and the difference between the two firms’ expected WOM is sufficiently small. As the market growth rate increases (decreases), competitive firms are more likely to adopt the delayed-release strategy (early-release strategy). Further, compared with the optimal strategy in a monopoly setting, a software vendor is more likely to adopt an early-release strategy in the competition setting, and the probability of adopting an early-release strategy will be higher as the intensity of competition becomes stronger. Firms’ interests and the social planner’s interest can be aligned under certain conditions, but this is not always the case. Finally, we examine the firms’ optimal release strategies and quality design strategies when we assume a specific functional relationship between WOM and consumer requirement uncertainty and product quality.

KEY WORDS AND PHRASES:

• Duopoly setting
• release strategy
• software markets
• software products
• software release
• word-of-mouth effect

Acknowledgments

This work was supported by the Key Program of the National Science Foundation of China (71631003), the General Program of the National Science Foundation of China (71371135, 71871155, and 71971153), and the National Science Fund for Distinguished Young Scholars of China (Grant 70925005). It is also supported by the Program for Changjiang Scholars and Innovative Research Teams in Universities of China. We are very grateful to the editor-in-chief, Professor Vladimir Zwass, and anonymous reviewers whose invaluable comments and suggestions substantially helped improve the quality of the paper.

Notes

1. See a description of this philosophy on Wikipedia (https://en.wikipedia.org/wiki/Release_early,_release_often).

2. Although Firefox and the subsequent Chrome are free for users, firms could obtain profits in other ways, such as advertising.

3. See a description on Wikipedia (https://en.wikipedia.org/wiki/SAP_ERP).

4. See an article from a senior iOS developer (https://www.pixelstech.net/article/1322490137-Software-philosophy%3A-Release-early-release-often-vs-polished-releases).

5. Early-release strategy is also called replacement strategy in other literature [4], and delayed-release strategy is also called skipping strategy [4] or leapfrog strategy [32].

6. If there exists no direct competition between two products, then it forms the market with two local monopolies, which is not different from the scenario with single monopoly vendor; therefore we do not consider this scenario.

7. When software vendors make pricing decisions, they usually attempt to maximize their total profit in the two periods by simultaneously choosing prices for the limited-quality version and full-quality version according to their estimations about future market condition (i.e. WOM effect). Thus we use the global-optimization method rather than backward induction used in traditional literature when deciding the optimal prices.

8. When $q_L^A\ne q_L^B$, the only influence is that the firm with a lower quality level will adopt a delayed-release strategy with a higher probability. However, the overall insights given from Figure 3 still hold; thus, we just present the results when $q_L^A=q_L^B$. The same reason can be applied to the following corresponding illustrations.

9. We do these experiments to confirm the correspondence of implications between WOM space and quality space. In the presented results, there is no DD equilibrium, because the presented market equilibrium results are only the subsets of the whole equilibrium results. When market growth rate is set higher or α is set lower, DD will appear. Because the present experiment results can derive the main insights, we omit the results including DD equilibrium.

10. We also examine the conclusions obtained in this subsection when the quality level of Firm B’s limited-quality version is low $q_L^B=0.2$ and high $q_L^B=0.8$, and we find that the implications (Proposition 6) that we obtained are held. Thus, we present the results obtained when the quality level of Firm B’s limited-quality version is moderate.

Additional information

Notes on contributors

Yu Wang

Yu Wang ([email protected]) is a Ph.D. candidate at the College of Management and Economics, Tianjin University, China. His research interests include economics of information systems, information systems and marketing, pricing, and competitive strategies. His work has appeared in Information Technology & Management and in conference proceedings, including International Conference on Information Systems, Workshop on Information Technologies and Systems, and China Summer Workshop on Information Management.

Minqiang Li

Minqiang Li ([email protected]; corresponding author) is a professor in the Department of Information Management and Management Science at the College of Management and Economics, Tianjin University, Tianjin, China. He received his Ph.D. in Systems Engineering and Management Science from Tianjin University. His research interests cover management science and decision support, electronic commerce, data mining and business intelligence, and evolutionary computation. His papers have appeared in MIS Quarterly, Journal of Management Information Systems, European Journal of Operational Research, Journal of Evolutionary Economics, International Journal of Production Economics, IEEE Transactions on Neural Networks and Learning Systems, Information Sciences, and other journals.

Haiyang Feng

Haiyang Feng ([email protected]) is an associate professor of information management and management science at the College of Management and Economics, Tianjin University, China. He received his Ph.D. in Management Science from Tianjin University. His research interests include economics of information systems, platform strategy, and business analytics. His papers have been published in such journals as MIS Quarterly, International Journal of Production Economics, Computers & Industrial Engineering, and Computers & Operations Research.

Nan Feng

Nan Feng ([email protected]) is a professor of information management at the College of Management and Economics, Tianjin University, China. He received his Ph.D. in management science and engineering from Tianjin University. His research interests include information security economics and risk management. His work has appeared in MIS Quarterly, Journal of Management Information Systems, Information & Management, Enterprise Information Systems, Information Sciences, and other journals.