There is a mountain of literature about increasing the parallelism of projects to improve performance. While many have found this approach beneficial, others have seen it create havoc in their projects. It is important to have some insight into the dynamics of dependencies to understand how increased degree of overlap in activities can have such a high potential for both positive and negative consequences on a project.

"Accelerate the work available"
The first intended consequence of increased degree of overlap in activities is to increase the amount of work available at once. Assuming resources are available this can greatly accelerate the speed at which a project is done. In our modeling work on projects, we have found that a increase in the overlap of work in the early stages of the project can make a substantial difference in the overall time for the project. Preston Smith in his book Creating Products in Half the Time calls this early phase the "fuzzy front end" and he indicates that most organizations have substantial opportunities to increase cycle time by improving in this area. This loop is self-reinforcing. As you raise the degree of overlap in activities, the number of activities available for work goes up, and assuming that resources are available, the number of activities completed also goes up which make even more work available. It has been clearly demonstrated in the literature and several case studies that most assumptions about what is possible in terms of overlap is far too conservative. People tend to blend what is possible with what they perceive to be affordable. Is building the second story of a house dependent on the first floor being complete or even started? Not if there is a crane available. In this case, it may be possible, but necessarily not smart financially. In some cases time may be so valuable, that creative ideas about breaking apparent dependencies may be welcome even at extremely high cost.

"Resolve Interface Issues Early"
The second intended consequence of increased degree of overlap in activities is the reduction in work caused by concurrent engineering. If done properly, the increase in overlap will help force communication on interface issues to be dealt with early. This decreases the rework between functions and phases. Less rework means less unplanned activities and more resources allocated to the planned activities of the project. This helps to accelerate the project creating more time and earlier surfacing of interface issues. This is a self-reinforcing loop that rewards the team that increases overlap by eliminating non-value-added rework.

"Build on Poor Dependencies"
The unintended consequence of the increased overlap of activities is the increase in the error rate caused by starting work before the dependencies are completely set. Dependencies will change at times. If much of the work is done in parallel, the dependent activities will have to be reworked or scrapped when changes are made. This increase in work tends to pull time away from concurrent engineering activities and can set up a vicious cycle where error generation can create even more errors in future work.