It’s hard to set a strategic direction in the midst of uncertainty. One persistent problem is that the data to justify a course of action only becomes clear when it’s too late to act. This problem — or information-action paradox — demands a paradoxical solution. In this case, leaders need to act when the data tells them not to.

That’s not to say that leaders should ignore data. Rather, they should recognize the limits of the data they usually use to make strategic investments such as market size, growth rates, share, margins, and so on. The existing data describes what has already happened, not what might or will happen.

Strategy making shouldn’t be left to guesswork or pure hunches either. Instead, leaders should deliberately wander to find fresh insight that can help them make sense and set a course through uncertainty. This article describes three specific ways to do just that.

1. Seek early warning signs of change.

One of the pillars of Satya Nadella’s change program at Microsoft has been shifting from focusing on lagging indicators, such as revenues, to leading indicators, such as net promoter score. Lagging indicators result from past decisions. Leading indicators point to future problems or opportunities.

Beyond having leading indicators on dashboards, executives should seek our early warning signs of change in two places.

The first place is with customers. Understand how current customers are considering emerging solutions. One thing to watch for is “near misses,” where customers considered, but ultimately rejected, choosing alternative suggestions. Such an analysis helped the Australian arm of global law firm King & Wood Mallesons see signs that customers were beginning to take new legal technology providers more seriously, spurring action to develop a capability around technology before the need showed up in traditional data.

Ideally, executives should spend time with a wide spectrum of customers: loyal mainstream customers, customers that have recently switched to other solutions, and customers in extreme circumstances. One beverage company trying to understand hydration carefully studied how militaries functioned in extreme heat. It helped the company come up with innovative solutions, which it could adapt for more mainstream uses, such as tablets to add to water to provide extra nutrients or new packaging that kept beverages cooler for longer periods of time.

Second, executives should have first-hand knowledge of the relevant startups in their sector. Most startup companies fail, of course, but paying careful attention can help to spot important changes early. Microsoft’s early investment in OpenAI gave it a window to changes in the artificial intelligence industry. Go deeper than the big name, publicly visible startups; seek ones that are still in stealth mode. Or visit a university lab and get a glimpse of next-generation technologies. This exploration strengthens the intuition to set a course through uncertainty.

Read more here.