What is customer discovery?

The fast version

Customer discovery is the first phase of Steve Blank’s Customer Development methodology (The Four Steps to the Epiphany, 2005). It’s the work of finding out whether the problem you think exists actually exists, for the buyer you think has it.

Discovery uses interviews and observation. The output: a defensible answer to:

• Who is the buyer? (Specific, named, reachable.)
• What’s the problem? (Real, frequent, intense, costing them today.)
• What’s their current workaround? (And why does it persist?)
• What’s the job they’re hiring a solution to do? (JTBD switch statement.)

Discovery is not validation. Validation comes next.

Customer Development’s four phases

Blank’s methodology, briefly:

1. Customer Discovery. Does the problem exist? Discovery interviews + observation.
2. Customer Validation. Will buyers pay? Behavioral evidence: pre-orders, paid pilots, signed LOIs.
3. Customer Creation. Does demand scale? Marketing + sales motions to expand from the validated beachhead.
4. Company Building. Operational scaling. Move from “find customers” to “serve at scale.”

Modern startup playbooks (Lean Startup, ShipFit’s 9 stages) reorganize these phases but the underlying logic is intact: prove the problem exists before you prove the solution; prove the solution sells before you scale; scale before you build the company around it.

How discovery is different from validation

The two are sequential. Discovery without validation produces a precise problem statement and zero customers. Validation without discovery produces commitments to solve the wrong problem.

Common discovery mistakes

1. Pitching during discovery. The moment you describe your idea, the conversation contaminates. Apply the Mom Test discipline: never mention the idea, ask about past behavior.

2. Talking to friends. They’re polite; you don’t get signal. Talk to strangers, ex-colleagues, or people one LinkedIn-degree away.

3. Counting one good interview as discovery done. Patterns emerge around interview 10. Below that, you have anecdotes.

4. Skipping discovery because “I am the customer.” Founder-as-customer is a strong start AND a known bias. Run 10+ discovery interviews to test whether your problem is their problem.

5. Confusing discovery with validation. Interviews tell you the problem exists. They don’t tell you anyone will pay. Don’t ship V1 on discovery alone; run validation first.

How ShipFit operationalizes this

ShipFit maps the Blank phases to its 9 stages:

• Discovery → Stages 2-3 (Who Pays? + What Hurts?)
• Validation → Stages 6-7 (How to Charge? + Will They Pay?)
• Creation → Stage 8 (How to Launch?)
• Building → outside the validation playbook (engineering + ops post-PMF)

Stages 2-3 generate the Mom Test interview questions tuned to your buyer + hypothesis, then synthesize transcripts into the 3-4 most-cited verbatim pain phrasings + buyer sub-segments + bias checks. The output is a defended discovery deliverable that feeds directly into the validation stages.

Further reading

• Steve Blank, The Four Steps to the Epiphany (2005), the source.
• The Mom Test framework, the interview discipline that makes discovery work.
• Buyer Persona Canvas, what to do with the persona insights discovery surfaces.
• How to validate a business idea, the wider validation flow that follows discovery.