This "postponement" strategy may be applied to almost any kind of new product development.
As an example, suppose we are developing a new consumer electronic product such as a printer. As we decompose the project, we see that there are some long lead-time, higher risk elements (say the ink application element or the digital control chips and software) and some low-risk and short-term components (power supply, page feeder, etc.). The temptation of many project managers is to treat all of these subprojects with equal priority and do them as much as possible in parallel, thinking that strategy will minimize the development time. Or they may even start the quick, low-risk parts first, rationalizing that some quick successes and completions will build morale for the team and look good on project reviews.
For these strategies, the problem is that we are not reducing the risk or time to develop the complete printer, possibly even depriving the high-risk elements of personnel, resources and mindshare. For example, perhaps the printed-circuit board layout department will defer laying-out the riskier controller board because they are busy laying out the power supply board (I have seen this very thing happen!)
A new product development process can be viewed as a process of amassing an inventory of knowledge about how to build, deliver, market and support the product. As this knowledge inventory is increased (gaining in value), the risk of failure in the marketplace is reduced. The value of bits of knowledge inventory, like cans of colored paint, can degrade with time if sitting still. Why? Components that may have been selected for the power supply, for example, may become obsolete or supplanted with better and cheaper components later. Thus, starting that part of the development early may have been a waste of time.
To reduce development time, we should first do a careful lead-time and risk-assessment of each component of the project. The goal of each subproject is to build our knowledge and reduce the risk of not meeting the goals of that component to an acceptable level. All of the resources and mindshare we can afford should be applied to the highest-risk and longest lead-time subprojects, postponing the low-risk and quicker subprojects as long as possible. We can thereby take maximum advantage of any new parts or techniques available and finish the development in time with finishing the bottleneck, higher-risk components.
In the printer example, our team may assess that it will take 4 months to develop and program the controller chips and software, and the risk is high that the development time may extend to 5 months - the longest lead-time and riskiest subproject. Whereas we may be confident that we can find a supplier to customize a power supply in two weeks. The best approach is postpone the power supply subproject (and like items) and spend all available resources on the controller development - and possibly other high-risk elements - until we conclude that those elements can be completed successfully in two weeks. Then we launch the power supply development to be completed at the same time.
This "postponement" approach builds our overall knowledge inventory about the product - and lowers overall risk - at the fastest possible rate. It also minimizes "rottage" of that inventory, significantly reducing development cost and minimizing time to market.