Your first value proposition should be a testable decision, not a timeless truth. State the intended customer, urgent problem, useful outcome, mechanism and current evidence. Mark the assumptions you have not validated, use the statement consistently enough to learn from it, and set a review date. Being provisional means being explicit about uncertainty, not publishing careless claims.
A provisional statement still needs discipline
Teams often delay because every word feels permanent. Others publish a sweeping promise and call it experimentation. A better starting point is bounded and operational.
Use this structure: “For [customer] who needs to [make progress in a situation], [offer] helps them [useful outcome] by [mechanism], supported by [current evidence].” The full value proposition guide explains how those components fit together.
Separate evidence from assumption
Create an evidence table with four columns:
| Element | Current choice | Evidence | Confidence |
|---|---|---|---|
| Customer | Who appears to care most | interviews, enquiries or observed behaviour | low, medium or high |
| Problem | What obstructs progress | repeated situations and language | low, medium or high |
| Outcome | What useful change is sought | buying criteria or past decisions | low, medium or high |
| Mechanism | How the offer helps | product behaviour or delivery process | low, medium or high |
| Proof | Why the claim is believable | demonstration, record or approved result | low, medium or high |
Low confidence is not failure. It tells you what to investigate next. If you have no customers yet, use the ethical research approaches in customer research before customers rather than inventing certainty.
Write the smallest useful claim
A provisional proposition should be specific enough to test and modest enough to support. Replace “guaranteed growth” with an outcome you can demonstrate, such as reducing the manual steps needed to prepare a weekly report.
Check that a reader can answer:
- Is this meant for someone like me?
- Is the problem important in my current situation?
- What changes if I use it?
- How does it produce that change?
- Why should I believe the claim?
If the answer requires internal vocabulary or several caveats, revise before testing.
Illustrative example: state the hypothesis
Illustrative example: a founder building a planning tool writes, “The operating system for modern teams.” It offers no testable customer, situation or outcome.
A provisional version is: “For owner-led agencies planning several client projects at once, the tool brings capacity and deadlines into one weekly view so project leads can spot clashes before assigning work.” The founder labels three assumptions internally: agencies are the priority segment, weekly planning is the buying situation and earlier visibility is valuable enough to change tools.
The example is hypothetical. The statement becomes stronger only if research and behaviour support those assumptions.
Test comprehension before persuasion
Show the statement in realistic context and ask intended customers to explain it in their own words. Ask who they think it is for, what problem it solves, what they expect it to do and what proof they would need. Do not ask, “Do you like it?”
Then test relevance through conversations, qualified enquiries, product use or a controlled landing-page experiment. The guide to testing a value proposition provides a fuller sequence. Record disconfirming evidence rather than collecting only positive reactions.
Decide what would trigger a revision
Set a review date and evidence triggers in advance. Revise when intended customers consistently misunderstand the offer, a different problem drives purchase, the mechanism changes, or claims can no longer be supported. Do not rewrite after one casual opinion.
Keep a short decision log: version, assumption changed, evidence considered and consequence for product or message. This protects learning when team members join later.
Limitations and trade-offs
Frequent revision can create noisy data and confuse sales teams. Use one version long enough to observe patterns. Conversely, consistency is not a reason to preserve a proposition contradicted by evidence.
Early tests cannot prove durable demand. Sample bias, price, distribution and product quality can affect results. Treat message evidence as one input to a wider business decision.
Frequently asked questions
Should customers see the word “provisional”?
Usually not. The public claim should be clear and supportable. “Provisional” describes your internal confidence and review process, not an excuse to publish uncertainty carelessly.
How many interviews make a proposition final?
There is no universal number. Look for repeated evidence across relevant participants and behaviour, while continuing to monitor contradictory cases.
Can we test several value propositions at once?
Yes, if each represents a distinct hypothesis and the test can attribute feedback. Avoid changing audience, offer, channel and wording simultaneously.
Your useful next step
Draft your proposition with the five-part structure. Underline every assumption, give each a confidence rating, and choose the single low-confidence assumption that would most change the business if wrong. Design one interview or behaviour test for it this week.
