Reinertsen & Associates, Tip of the Month: January 2011

Many product developers have a tendency to blend together the ideas of iteration and batch size. After all, the methods that make working in small batches feasible also make it easier to adopt an iterative approach. Nevertheless, I think it is useful to recognize the different mechanisms of actions between these two methods because it helps us recognize when they are useful.

Let us take a very simple example, painting the interior walls of a house. Assume that such painting involved a series of operations: patching defects in the wall surface, masking off the areas to be painted, applying a primer, applying one or more finish coats, and clean-up. A large batch size approach would complete all patching before beginning any of the masking. It would prime the entire house before putting the finish coat on any room. A smaller batch approach would work on a room at a time. The attractiveness of smaller batches depends on transaction costs and holding costs. You might prefer to have a single room out of service at a time. In this case you are concluding that the holding cost of having many rooms in process at the same time is too high. On the other, your brother may be a professional painter with paint-spraying equipment and the skill to use it. He is willing to come and put the final on your entire house in a few hours, but can only do this if all rooms are ready to go at the same time. In this case, you are using a single large batch to get scale economies by paying the transaction cost once.

What is the meaning of iteration is such a context? A simple example is whether I should apply one coat of paint or two. If I quickly apply one coat of paint there a likely to be minor areas where there are defects in my coverage. It is frequently more cost effective to apply two coats of paint recognizing that defects are unlikely to be in exactly the same place twice. Your alternative might be to slowly and meticulously paint every square inch trying to prevent all defects on the first pass.

The mechanism of action behind small batch size involves minimizing total cost which is the combined effect of the cost of doing a batch and the cost of not doing a batch.

Don Reinertsen