Better Never than Late

A well-known maxim advises “Better late than never,” but in the world of capacity management, the reverse can often be true. Boston Consulting Group vice presidents George Stalk and Thomas Hout addressed the problems of overcorrection based on a model originally discussed in a 1958 Harvard Business Review article by Jay Forrester, the inventor of system dynamics.5 When demand is cyclical, responding slowly can be worse than not responding. As demand grows, capacity is added, but by the time it is added, demand has fallen off, so there is even more excess capacity. As Stalk and Hout observe in analyzing a production response to a demand increase, if you attempt to “smooth the demand by advertising and promoting in the troughs and raising prices in the peaks, the oscillations would actually increase.”6

If we consider the “better safe than sorry” analysis, we realize why “better never than late” can in fact be a good strategy. If we try to chase random fluctuations, we will end up picking points other than F. But we already have seen that F is the best choice when we don’t know what’s coming. Since we don’t know what’s coming, any other choice is worse. In the tennis game, if we were to use a shot on the right sideline as an indicator that all coming shots will be on the right, we have just worsened our position in returning the uniformly distributed shots that may be as far off as the left baseline.

Although we are addressing capacity here, these insights ...

Get Cloudonomics: The Business Value of Cloud Computing, + Website now with the O’Reilly learning platform.

O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.