The cost of speed: Work policies for crashing and overlapping in product development projects

Project management theory provides insufficient insight on the effects of crashing and overlapping in product development (PD) projects. The scholarly literature largely assumes that networks of project activities are acyclical. PD projects are iterative, however, and the time-cost implications of managerial actions, such as activity crashing and overlapping, differ in this context. We build a rich model that accounts for activity overlapping, crashing, and iteration in the project's activity network. Using a bank of thousands of artificial but realistic test problems, we test the performance of several alternative work policies - managerial rules about when to start work and rework, and when to allow crashing and overlapping - and compare their time-cost implications. We find that the time-cost benefits of crashing and overlapping increase with increasing iteration. However, whereas crashing works well in combination with an aggressive work policy that does not seek to minimize iteration and rework, overlapping ironically works better with a work policy that defers rework. No single work policy dominates in all situations, although an aggressive policy that precipitates rework may often be worth its added cost. We distill our findings into a set of propositions that pave the way for further theory development in this area. Our results also illuminate useful heuristics for managers of PD projects.