Essays on software systems design, release, and early promotion

In today's volatile software market, both quality and time to market are critical to a new software product's success. However, can a firm accelerate the delivery and adoption of its software products without trading off software quality, or improve the qualify of its software without delaying release and market adoption? To answer this question, my dissertation study focuses on the (1) design, (2) testing and release, as well as (3) early promotion phases of the software development and marketing process. I propose three strategies that a firm could consider: (1) usage-based optimal design, (2) patching, and (3) free software offer. The three strategies are addressed in three separate essays.; In the first essay, I study an important usage-based design issue, the attribute-value conflict problem that is often encountered when merging data from multiple sources. I propose a framework for making the cost-minimizing decision based on the distribution of possible values and costs of errors associated with decision problems of interest. Analyses show that the cost-optimal solution provides a substantial amount of cost saving as compared to a solution that always pick the most likely value.; The second essay studies the software release policy with patching considered. I show that when patching is available and not very costly, a firm can release its software early and continue to test even after release. This policy has several advantages over the policy without considering patching. First, the total expected cost is reduced by a significant amount. Second, the reliability of the software product is higher after testing stops. Third, the expected number of software failures in the field is lower.; The third essay addresses the free software offer strategy. In this essay, I answer the following questions: what kind of products can benefit from free offer, and when and how should the free software be distributed to maximize a firm's own benefit? I conclude that a software firm can benefit from free offer because the word-of-mouth effect of the free adopters can accelerate the whole diffusion process.