May 21, 2013

Adopt a gradual rollout of solutions, if possible

I’ve heard that the Swedes make fun of the Norwegians by attributing them silly stories. The French, on the other hand, make fun of the Belgians. Here is one of these fictitious histoires belges:

“To demonstrate their frustration with the European Union, the Belgian parliament have decided to follow the Brits: starting this summer, Belgian traffic will now drive on the left side of the road. To facilitate the transition, the Parliament have adpoted a gradual roll out: in June only 18-wheelers will switch, while the rest of traffic will join them in August.”

Favor a gradual rollout, if possible

In some situations rolling out a solution gradually isn’t an option. But in other cases, it is. If you are in such a case, you should consider it.

A gradual rollout reduces the impact of imperfections. If your problem is complex/CIDNI, chances are that the solution won’t be trivial either. A gradual roll out allows to learn from the experience of pilot groups and adjust accordingly. That was Atul Gawande’s experience in developing a checklist to reduce surgical mistakes (by the way his book, the Checklist Manifesto, is directly relevant to people solving complex problems, and it’s a fast and entertaining read). Pfeffer and Sutton also recommend it as part of evidence-based management through the use of pilot programs.

That is also our experience at Rice as we are rolling out a system to manage campus visitors. In fact, in this case, we get to solve two problems with a great deal of similarity; only one has more complexity than the other. Following this post’s idea, we have settled on solving the simpler problem first before tackling the second. And even for the first problem, we are rolling out the solution to pilot groups first to make sure that our mistakes impact as few people as possible.

A gradual rollout allows to create success stories. When rolling out big changes, it makes sense to “think big, start small, scale fast.” A gradual roll out enables the “start small” part. It allows you to create success stories that you can then share with the rest of the target population and, hopefully, make them envious.

Of course, the impact of a gradual rolled out solution takes longer to notice, so I’m not advocating for it in all situations. All I’m saying is that you should consider it.


Pfeffer, J. and R. I. Sutton (2006). “Evidence-based management.” Harvard business review 84(1): 62.

Pfeffer, J. and R. I. Sutton (2006). Hard facts, dangerous half-truths, and total nonsense: Profiting from evidence-based management, Harvard Business Press.

Gawande, A. (2009). The Checklist Manifesto. New York, Picador.