The limits of discovery
The next lesson couldn't come from another conversation.
The situation
I recently spoke with the founder of a medical hardware company developing a new medical device.
The company began as a university capstone project, where the team spent months inside hospitals observing procedures, interviewing clinicians, and trying to understand why an existing procedure was so difficult to perform consistently. By the time they graduated, they had built an early prototype, completed dozens of customer interviews, and felt confident enough to start a company around the problem.
They believed that understanding the customer would be one of the biggest challenges they would face.
The plan
After founding the company, they decided to keep building on that foundation.
They continued interviewing physicians, observing procedures, and refining both the problem and their proposed solution. When early concepts didn't fully address the clinical need, they went back into discovery before committing to another direction. That process repeated itself several times as the team explored different technical approaches and gradually narrowed in on a solution they believed could work.
At the time, the approach felt like the right one. Every conversation improved their understanding of the problem, and every prototype gave them something more concrete to discuss with clinicians.
Where uncertainty remained
As the years passed, something began to change.
The interviews became less surprising. Instead of uncovering new insights, the founders found themselves hearing the same feedback over and over again. Eventually, they reached a point where they felt they could predict how most clinicians would respond before asking the question.
At the same time, progress on the product remained slower than they had hoped. Technical challenges, funding, regulatory work, and product development still stood in front of them. Looking back, the founder realized they had become increasingly focused on reducing uncertainty before moving forward.
The difficult question became whether another interview would actually change what they built, or whether the next lesson could only come from continuing development.
What needed to be proven first
Looking back, the founder didn't conclude that they had done too much customer discovery.
They concluded that customer discovery had eventually stopped being the biggest source of learning.
There came a point where the remaining questions could no longer be answered through interviews or observations. They could only be answered by building, testing, and accepting that some uncertainty would remain until the product reached the next stage.
The founder's instinct had always been to eliminate as much risk as possible before moving forward. Today, they wonder whether the company would have learned faster by accepting more risk earlier and solving some of those problems only once they appeared.
The pattern
One thing I found interesting about this conversation was that it challenged a piece of advice founders hear all the time: talk to customers.
That advice is absolutely right at the beginning.
But customer discovery isn't the only form of learning. Eventually, the questions change. The challenge is recognizing when conversations have taught you everything they can, and when the next insight requires building, testing, and seeing what happens in the real world.
If you're in a similar spot
If you've spent months talking to customers, it may be worth asking yourself:
Is my next important decision waiting on another conversation, or is it waiting on something I need to build?
Sometimes the fastest way to learn isn't another interview. It's accepting that the next answer only comes from moving forward.