I'd rather play golf

In my previous post I began this book review by concentrating on the first two important themes Don Reinertsen describes in The Principles of Product Development Flow [1]: Economics and Queuing. In this part deux, we briefly summarize another important theme: Variability.

Variability

Completely eliminating variability in the product development process is often seen as the ultimate goal: e.g. if only our estimates were accurate, and we planned for every contingency, we could operate at much closer to 100% capacity, become optimally productive, and produce products on time and on budget. Sound too good to be true? You’re right. The issue for many product development shops is that at high capacity utilization the effect of queues is exacerbated by variability–here is Figure 3-6 from page 73, used with permission:

Reinertsen’s thesis is that since we know that estimates are rarely accurate, and surprises in product development occur constantly, it is foolhardy to adhere to fixed project plans. Rather, one must manage the problems of variability by exploiting the opportunities it creates (page 96):

Product development creates economic value by producing the recipes for products, information, not by creating physical products. If we create the same recipe twice, we’ve created no new information, and consequently, no economic value. If we wish to create value, it is necessary, but not sufficient, to change the design. But changing the design introduces uncertainty in outcomes. In fact, we must introduce uncertainty to add value in product development. Risk-taking is essential to value creation in product development.

Risk-taking plays no equivalent role in manufacturing processes. We can manufacture the same product one million times and we will create value every time. For a manufacturer, reducing variability always improves manufacturing economics. This is not true in product development. In product development we must carefully distinguish between variability that increases economic value and variability that decreases it. Rather than eliminating all variability, we must actively put it to work to create economic value.

Because development projects have unlimited downside (we can spend an infinite amount of resources chasing an ultimate product failure) and only limited upside (once customer preference is achieved, adding more features produces little incremental value), managing variability is important and can be achieved in two broad ways: reducing variability, and changing its economic impact.

Reinertsen looks at seven ways to reduce variability; pooling projects together is one way to reduce the variability in individual projects. Besides pooling (in several ways), other ways to reduce variability include counterbalancing–the example being the cross-training of development staff to handle peak workloads–and time buffers, in which one trades an increase in cycle time for a reduction in variability.

Changing the economic impact of variability includes shifting variability to the development process stage where its cost is lowest, or addressing peaks using a different performance variable, which is discussed in Principle V14: variability substitution. In product development, we constantly tradeoff expenses to stabilize cost, schedule, and product performance. Making these choices wisely can be done only if the economics behind these factors are well understood.

In part trois, we’ll finish this review by discussing two more themes from this volume: batch size and Work-in-process (WIP) constraints.

[1] Donald G. Reinertsen (May 2009). The Principles of Product Development Flow: Second Generation Lean Product Development. Celeritas Publishing, Redondo Beach, California. ISBN 1-935401-00-9.

[2] Donald G. Reinertsen (1997). Managing the Design Factory. Simon and Schuster, New York, New York. ISBN 0-684-83991-1.

[3] Preston G. Smith and Donald G. Reinertsen (1998). Developing Products in Half the Time: New Rules, New Tools (2nd edition). John Wiley and Sons, New York, New York. ISBN 0471-292524.