What is customer discovery?
Customer discovery means talking to real people, observing real behaviour, and letting evidence — not hope — tell you whether your product solves a real problem for a real group of people.
Most founders skip this. They build first, then search for customers. The result is months of wasted effort targeting a market that doesn't exist — or targeting the right market in the wrong way.
Customer discovery flips the order: understand the problem first, then test who has it, then build what they actually need.
The five stages of customer discovery
- 1
State your assumptions
Write down what you believe about your market: who has the problem, what the problem costs them, why they would buy your solution, and what alternatives they currently use. These are hypotheses, not facts.
- 2
Find evidence
Look for real signals that the problem exists: complaints, workarounds, forum discussions, support tickets, abandoned workflows, repeated manual processes. Evidence is anything that shows someone is already struggling with the problem you think you solve.
- 3
Compare segments
Don't commit to one ideal customer profile. Generate several possible segments who might have the problem. Compare them on problem pressure, evidence strength, frequency, buyer accessibility, and commitment signals. Let evidence — not preference — decide which segment to test first.
- 4
Run one test
Choose the fastest, cheapest experiment that answers your biggest unknown. That might be a problem interview, a workflow audit, a demo, a landing page test, or a pilot. Define what success looks like before you start. Define what would disprove your hypothesis.
- 5
Record what happened
Document what you tested, what happened, what changed, and what you should test next. This is where most founders lose learning — they move on without capturing what they discovered. Over time, this record becomes your most valuable asset.
What counts as evidence?
A strength ladder, from weakest to strongest.
- 1
Founder belief
"I think this is a problem"
- 2
Public signal
people mention it online
- 3
Reported behaviour
someone says they do this
- 4
Observed workaround
you see them struggling with it
- 5
Shared data
they show you their process
- 6
Repeated use
they keep hitting the same issue
- 7
Deposit, signed pilot, or payment
they put skin in the game
Common mistakes
Committing to one ICP too early — before the market has answered
Asking people what they want instead of observing what they do
Counting replies and opens instead of measuring commitment
Forgetting failed experiments instead of learning from them
Treating an AI-generated persona as market truth
How Resultory helps
Resultory operationalises this entire process. It turns your assumptions into structured hypotheses, helps you compare segments, recommends the next experiment, tracks commitment, and preserves what you learned — so every cycle makes you smarter.