Finding and solving problems in software new product development

New product development is notoriously difficult, and software new product development particularly so. Although a great deal of research has investigated new product development, projects developing new software products continue to have problems meeting their goals. In fact, one line of research proposes new product development is difficult because it must solve an ongoing stream of complex problems. I integrate this line of research with two others to develop a conceptual framework of new product development as a process of finding and solving problems. From this framework, I develop four hypotheses that predict the probability projects developing new products will attain their development schedule and product quality goals. More specifically, I hypothesize that projects that generate access to, and integrate, large quantities of creative ideas, in-depth knowledge, and accurate information, should increase their probability of attaining schedule and product quality goals. Projects developing new products should both generate and integrate this “knowledge” to solve the problems that stand between them and their goals. However, how projects find problems also matters. Projects that search to identify problems earlier, rather than later, should also increase their probability of meeting schedule and product quality goals. To test these hypotheses I gathered data on 33 projects that tried to develop new software products from 23 firms, through interviews and questionnaires. Results from regression analyses support three out of four hypotheses. The projects that had high levels of both knowledge generation and integration had a significantly higher probability of attaining their product quality goals, but not their schedule goals. In contrast, projects that merely searched to find problems had a higher probability of attaining both goals. Moreover, projects that not only generated and integrated knowledge to solve problems, but also searched to find them, had the highest probability of attaining their product quality goals. This study illustrates the usefulness of modeling new product development as a bundle of problems to be found and solved. These results suggest that projects that combine practices to implement high levels of both knowledge generation and integration—not just one or the other—increase their chances of meeting product quality goals. This in turn suggests that focus on any single process or practice may be misplaced. Moreover, proactive search for problems may increase projects’ chances of meeting both schedule and product quality goals. In fact, search for problems was highly significant in this study, which suggests the way projects identify problems deserves further study. Although these prescriptions are preliminary, this study suggests they can help projects—and their managers—embody their visions in products and deliver those products to market.