As companies mature, it’s easy to believe that the core experience and most user needs have been resolved, and all that’s left to work toward are the marginal benefits, the cherries on top. Cherries on top might add delight and panache, but they rarely cause fundamental shifts in performance and success. And as a business, even a mature one, we’re looking for the innovations that tangibly impact the KPIs we care about. 

Because we’re testing things that have a lower chance of causing top-line impact, experimentation as a practice and method can become a questionable task. Why spend time and effort to prepare and run an experiment if the results are inconclusive? It’s an understandable and fair question. At the end of the day, a business needs to prioritize actions that contribute to its objectives.  However, it would be the wrong conclusion to think we have to write off experimentation altogether — we can, instead, change our approach to it. 

In order to make better use of the experimentation method and achieve more tangible impact for the business in a mature context, we ensure we are following three rather straightforward strategies:

1. Start with the decision that needs to be made.
2. Utilize localization to innovate for homogeneous populations.
3. Break the feature apart into its most critical pieces.

Starting with the decision that needs to be made

Our quest for information, as humans and as organizations, often stems from the need to make decisions. We search for information about travel destinations and flight options to plan a vacation; companies try to decide whether to acquire organizations by gathering yearly statements and other financial data. And in the case of product development, we need to consider what to build, how to build it, for whom to build it, when it will launch, and ultimately decide whether it might be worth it to build at all.

To make a decision, we don’t necessarily need perfect information — or all the information. We need just enough to feel confident about going one way or another. One can see it as an optimization function between spending as few resources as possible to obtain relevant information versus empowerin