Test a value proposition by separating five questions: Is the customer problem real? Is the message understood? Is the outcome relevant? Is the promise credible? Does it support action? Use problem interviews, unprompted comprehension, evidence review and small behavioural experiments in that order. Asking “Do you like this?” mainly produces opinions about wording, not evidence that the value logic is sound.
Write testable assumptions
Break the proposition into claims:
- the intended customer encounters this situation;
- the problem has enough priority to prompt change;
- the outcome is valuable;
- the mechanism can create it;
- the alternative is accurate;
- the proof is credible.
The Value Proposition Canvas guide helps distinguish customer assumptions from offer assumptions. Rank each by consequence and uncertainty.
Match method to question
Test the problem
Interview people who recently experienced the situation. Ask about the last occurrence, current method, consequences, attempted alternatives and decision. Avoid explaining your offer too early.
Test comprehension
Show the message briefly, remove it and ask participants to explain the offer, intended audience, problem, difference and next step in their own words.
Test relevance
Ask what priority the outcome competes with, what would trigger action and which part matters least. Compare answers with real decisions, not hypothetical enthusiasm.
Test credibility
Show the proof, method and limitations. Ask what remains hard to believe and what evidence would reduce uncertainty.
Test behaviour
Use a small, ethically designed landing page, sales sequence, prototype or offer. Measure qualified action such as a relevant enquiry, trial completion or purchase—not just attention.
Strategyzer’s testing roadmap similarly moves from customer jobs, pains and gains to the offer and willingness to pay.
Illustrative example
Illustrative example: a service proposes, “Turn scattered customer interviews into a decision-ready evidence library.”
The team first interviews product managers about their most recent research synthesis. It then tests whether “decision-ready” is understood, shows a prototype evidence record and offers a paid pilot with explicit scope. If people understand the message but keep their current documents, the obstacle may be migration effort or low urgency rather than copy.
The example is hypothetical. A real test would require suitable recruitment, consent and data handling.
Use a one-page test card
Record:
- assumption;
- evidence already available;
- method and participants;
- observable signal;
- support, weaken and stop criteria;
- ethical or operational constraints;
- result and next decision.
Do not change the success rule after seeing the result. Keep failed tests: they prevent the same assumption returning as an unlabelled idea.
Use positioning validation when the test also changes audience, category or competitive frame.
Interpret results carefully
Clicks can reflect curiosity. Interview praise can reflect politeness. A sale can be driven by a relationship rather than the proposition. Look for convergence across evidence types and important contradictions.
When evidence remains sparse, use the more measured language in Writing a Value Proposition Without Customer Proof.
Limitations and trade-offs
Every test has bias. Recruitment affects interview findings; prototypes omit parts of the experience; sales tests mix message, audience, channel, price and timing. Small samples support learning, not broad numerical claims.
Testing also costs time. Match the depth to the risk and reversibility of the decision. A homepage iteration needs less evidence than a market-wide rebrand or major product investment.
Frequently asked questions
Should we A/B test two value propositions?
Only after understanding the customer problem and likely meanings. An A/B result identifies different behaviour in one context; it does not reveal the full reason.
Can existing customers test a new proposition?
Yes, but include prospects, lost buyers or a new segment when future positioning differs from the installed base.
What if people understand the message but do not act?
Investigate relevance, credibility, timing, price, effort and channel. Clear copy cannot create demand or remove every buying constraint.
Your useful next step
Choose the assumption that would most damage the proposition if false. Complete a test card, recruit people who recently faced the situation and run the smallest ethical test capable of changing your decision.