When you aren't quite sure what the right answer is (which is more and more common in work today), ask yourself the question, "How can I figure out the next right thing?"

I recently had a conversation with someone managing a large change with the way their company interfaces with clients. The changes will impact the business on the order of hundreds of millions of dollars. I invited them to think about how they could test the changes before making a massive rollout.

They responded with alarm: "Our clients are too important to test something out with. We couldn't let them know something wasn't quite ready! When we make the change, we have to make sure it works!"

I countered that if that's the case (that they can't risk failure when they launch) then more than ever they need to figure out how to test smaller pieces before unveiling everything.

We then had a fruitful conversation about how to test in a way that doesn't jeopardize current relationships (topic for another time).

The tendency to clutch things tightly, not share the journey, and hold your breath when you launch isn't helpful when it comes to solving new problems.

Test before you leap.

As the proverb goes (usually credited to Warren Buffett, though it shows up in none of his shareholder letters):

"Don't test the depth of a river with both feet."

Manu Kumar, founder of K9 Ventures, posted a fantastic Twitter thread about how this very behavior should work when building companies.

"1/ So much of building a product/company is about progressive discovery. You build something, you figure out something else, then you build that, and so on. But you only figure out what to build by taking the first step, which may not always be exactly in the right direction."

-- Manu Kumar, January 31, 2020 (cross-posted on the K9 Ventures blog)

I invite you to consider these questions to help you to better use progressive discovery:

  • How can we mimic reality to see if this will work?
  • What would be the first indication that this is working?
  • What's the biggest pain point we can start with?
  • Can we start smaller?
  • What's our most pressing constraint?
  • What can we fake before we build, to gauge reaction first?
  • Can we try one before we have to do them all?
  • If we can't test on real customers, can we role-play or test internally first? Dry-run?

When you need to make an impact fast and/or when the stakes are high, figure out a way to prototype and test.