Developing products on "Internet time": The anatomy of a flexible development process
Citation: Alan MacCormack, Roberto Verganti, Marco Iansiti (2001) Developing products on "Internet time": The anatomy of a flexible development process. Management Science (RSS)
Internet Archive Scholar (search for fulltext): Developing products on "Internet time": The anatomy of a flexible development process
Tagged: Business (RSS) Product Development (RSS), Internet (RSS)
In many ways, MacCormack, Verganti, and Iansiti pick up where Eisenhardt and Tabrizi leave of. Their goal is to help unpack a set of mechanisms associated with development products on the Internet. The paper is empirical and present data on 29 completely development projects from the nascent Internet software industry and they demonstrate that increased flexibility (Eisenhardt and Tabrizi talked about experiential processes) are associated with higher levels of performance.
The authors describe the "stage gate" model as the traditional product development model where projects go through and passed through a set of phases that include concept development, design, integration, and testing and refining. MacCormack et al. argue that in uncertain and rapidly changing environments, this is ill-suited to good product development and that models that overlap these stages heavily, move prototyping up earlier, and build more flexibility in the process will be more successful.
The authors offer three hypotheses -- all focused on uncertain and dynamic environments:
- Greater investments in architectural design will allow more flexibility and changes later on and, as a result, will be associated with better performance.
- Earlier feedback on a projects system-level performance will be associated with better performance.
- Development teams with great amount of "generational experience" (i.e., people that have been through full product development processes before) will have the experience to support dynamic processes.
The authors present empirical evidence from 29 projects in 17 firms collected through a survey. The authors dependent variable was a measure of how good the product was as determined by a two-stage Delphi process using a panel of tech industry journalists and experts.
The results supported all three hypotheses and the second hypothesis with a very large effect size. The large take away was that in a flexible process, development teams should focus heavily on getting an early, unfinished, prototypic of the product to customers at the first opportunities so that the design can "co-evolve" with feedback. The three constructs together explained 50% of the variation in the success of projects.
Theoretical and Practical Relevance
MacCormack et al.'s paper has been cited 200 times since it was published 10 years ago in the innovation and product development literature and in the literature on management of Internet projects in particular.