When customer proof is limited, write the value proposition as a clear, evidence-labelled hypothesis. Be specific about the intended customer, situation, current alternative and capability you can demonstrate. Describe the immediate benefit conservatively, avoid outcome statistics and superlatives, and state important conditions. Then create a plan to test demand, use and results before strengthening the claim.
Separate what you know from what you believe
Create an evidence ledger with four labels:
- Known: verifiable facts about the product, process or market.
- Observed: patterns seen in relevant customer research.
- Assumed: plausible but untested beliefs.
- Unknown: important questions without a current answer.
The Value Proposition Canvas guide provides a structure for mapping the two sides without presenting guesses as fit.
Use the strongest honest claim level
Move through this ladder only as evidence permits:
- Capability: what the offer demonstrably does.
- Immediate benefit: what that capability enables in the workflow.
- Customer outcome: what meaningful progress occurs in practice.
- Comparative or quantified outcome: how results compare or by how much.
An early product may support the first two but not the last two. Specificity can come from a narrow customer and situation rather than a percentage.
Build a responsible first version
Use this structure:
For [intended customer] dealing with [supported situation], [offer] provides [demonstrable capability], making it possible to [immediate benefit] compared with [current approach].
Do not imply that “making it possible” guarantees the final outcome. Explain dependencies such as adoption, setup or data quality.
Illustrative example
Illustrative example: a new tool converts workshop notes into tagged decision records.
Premature claim:
Make every product decision 40% faster.
More responsible early proposition:
For product teams whose workshop decisions are spread across documents, the tool turns selected notes into tagged decision records, so colleagues can review the decision, owner and source in one place.
The second version can be demonstrated. The team should still test whether customers have the problem, use the records and value the resulting visibility. This is hypothetical.
Borrow evidence carefully
External research can establish that a general problem exists, but it does not prove that your product solves it. Cite the source, population and limits. Do not apply a statistic from another market to your own customers without justification.
A founder’s relevant expertise can explain why the approach is thoughtful. It is not outcome proof. Likewise, pilot participation shows interest, not a typical result.
Use Proof Points for a Value Proposition to match every asset to the claim it actually supports.
Build an evidence plan
For each major assumption, define:
- research question;
- appropriate participant or data source;
- method;
- observable signal;
- consent and data-handling needs;
- decision the result will inform.
Follow How to Test a Value Proposition rather than asking early users for a favourable quotation.
Limitations and trade-offs
Conservative language may feel less exciting and may produce a lower initial response than a bold promise. It also protects trust and creates a clearer learning baseline. Overqualification can make copy unreadable, so layer the detail: plain main claim, concise condition and accessible methodology.
Early adopters may accept uncertainty that mainstream buyers will not. Revisit the proposition as the product and evidence mature, recording substantive changes.
Frequently asked questions
Can we call the product “better” without customer results?
Only if the comparison and dimension are clear and supported. Describe the capability directly when a broad superiority claim cannot be substantiated.
Should we use beta-user testimonials?
Only with informed permission and accurate context. Do not edit a personal experience into a universal outcome or hide incentives and atypical conditions.
When can we add performance numbers?
After collecting reliable, relevant evidence with a transparent method and appropriate sample, and when the number reflects how a reasonable reader will interpret the claim.
Your useful next step
Highlight every statement in the current proposition as known, observed, assumed or unknown. Rewrite the main promise at the strongest honest claim level, then create one test card for the assumption most important to customer value.