Operations
How to build product messaging that ranks its claims in order
Rank product claims by evidence strength so the message leads with what you can prove and holds weaker benefits until better proof catches up.
What to take away
- Write every claim as a sentence a buyer could repeat, then sort it into one of four evidence tiers.
- Build a claim inventory in Google Sheets, Airtable, or Notion. Track the tier, the buyer problem, the owner, the evidence file, and the review date.
- Rank by tier first, then by how directly the claim answers the buyer's stated problem. Break ties with a comprehension test.
- Lead the launch message with tier 1 and tier 2 claims only. Hold tier 3 claims for demo moments and tier 4 claims in the internal file.
- In the US, an objective advertising claim needs a reasonable basis before you publish it, per the FTC substantiation policy.
- Test comprehension before you test persuasiona claim nobody parses cannot generate qualified response.
- Give every claim an owner, an evidence file, and a review date so the ranking does not drift after launch day.
Product messaging that ranks its claims in order starts with one rule: evidence sets the sequence, not enthusiasm.
A team writes ten benefits, ranks them by how good they sound, and ships the loudest one first. Sales cannot defend it, and the launch message collapses in the first demo.
The fix is a claim ranking you can document and repeat. This ranking is the spine of the whole product messaging build, and it starts with the positioning work underneath it.
Write the claim, then find its tier
A claim is one sentence a buyer could repeat to a colleague. 'Cuts month-end close from nine days to four' is a claim. 'Powerful reporting' is not; nobody can check it or repeat it.
Claim tiers and evidence
Tier 1
- Evidence
- Own measured results
- What to say
- Number and source
- Launch use
- Lead the message
Tier 2
- Evidence
- Customer-reported outcomes
- What to say
- Attribute who and result
- Launch use
- Carry the message
Tier 3
- Evidence
- Live demo capability
- What to say
- Show, do not quantify
- Launch use
- Demo moment
Tier 4
- Evidence
- Roadmap or inference
- What to say
- Internal only
- Launch use
- Internal file
Write every candidate claim in that form first. Then sort each one into a tier.
| Tier | Evidence behind it | What you may say | Where it goes in the rank |
|---|---|---|---|
| 1 | Your own measured results, with method and date | State it plainly, with the number and the source | Lead claim, or first support claim |
| 2 | Customer-reported outcomes you can name and cite | Attribute it: who, what setting, what result | Support claim, with attribution |
| 3 | Product capability you can demonstrate live | Show it; do not quantify the outcome | Demo moment, not headline |
| 4 | Roadmap, intent, or inference | Internal only until it moves up a tier | Internal file only |
Tier 4 is where launches break. A roadmap item written as a present-tense benefit is the single most common source of a claim that outruns its proof. Keep it internal until measured results or named customer outcomes move it to tier 1 or 2.
Rank the claims in five steps, then write the message
Order the tiers, then order within them. Use a claim inventory in Google Sheets, Airtable, or Notion. Add columns for claim, buyer problem, and tier. Add priority, owner, evidence file, and review date. Here is the sequence to apply.
- Log the claim.Write one sentence a buyer could repeat. Put it in the inventory. Do not combine two benefits in one row.
- Assign the tier.Choose 1, 2, 3, or 4 from the table. If you hesitate, drop it one tier.
- Score buyer priority.Rate how directly the claim answers the buyer's stated problem: 1 for primary, 2 for secondary, 3 for nice-to-have. Use interview notes or sales call transcripts, not internal opinion.
- Sort the inventory.Sort by tier ascending, then by buyer priority ascending. Tier 1 and priority 1 sit at the top.
- Break ties with comprehension.Ask five people outside the team to restate the tied claims. The claim with the higher restatement accuracy goes first.
- Write the message in order.Lead claim first, then support claims, then proof point, then demo moment. Store the hierarchy in Confluence, Notion, or Google Docs.
The launch message carries tier 1 and tier 2 only. Tier 3 claims become demo moments, where a buyer sees the capability rather than reading a number you cannot support. Tier 4 stays in the internal file.
This is the step most teams skip, and it is why their positioning collapses after launch day. If you have not done the underlying product positioning work, the ranking has nothing to sort against.
Substantiate before you publish
In the United States, the FTC advertising substantiation policy holds that objective advertising claims need a reasonable basis before they are disseminated. The basis is judged against the claim as a reasonable consumer reads it, not as your team intended it.
That standard applies to the exact promise, audience, and qualification you publish. It is a US federal standard; other jurisdictions set their own, so confirm the rule that governs each market with qualified counsel.
Substantiation test before publishing
Can you produce the evidence file today?
keep the claim at its tier
drop the claim a tier
Check comprehension before persuasion
A claim that readers cannot parse will not produce qualified response, no matter how true it is. Test comprehension first.
Show the claim to people outside the team and ask them to restate it. If their restatement differs from yours, rewrite the claim. Then measure qualified response: how many of the right buyers ask a follow-up question that shows they understood.
Keep the two measurements separate. Comprehension tells you the message landed. Qualified response tells you it landed on the right people. A claim can pass the first and fail the second.
Run a worked example: rank 'Cuts onboarding time by half'
Take a claim: 'Cuts onboarding time by half.'
Worked example checks
- Evidence check.Do you have measured before-and-after times, with method and date? If yes, tier 1. If it is a customer's estimate, tier 2. If it is a projection, tier 4.
- Comprehension check.Ask five people outside the team to restate it. If any of them ask 'half of what?', the claim is not finished.
- Qualification check.Add the condition that makes it true: which customer type, which starting point, which version.
- Rank.Place it against your other tier 1 claims by how directly it answers the buyer's stated problem.
Worked example ranking
| Claim form | Tier | Buyer priority | Order |
|---|---|---|---|
| Cuts onboarding time by half, with method and date | 1 | 1 | Lead |
| A named customer cut onboarding time by half | 2 | 1 | Support with attribution |
| Onboarding automation dashboard | 3 | 2 | Demo moment |
| Planned AI onboarding assistant | 4 | 1 | Internal file |
The GAO data reliability guide treats data quality as a function of intended use. Apply that here: your onboarding figure is reliable for the claim you are making only if its collection method fits that use. Document the assessment.
Keep the evidence file current
Every claim needs an owner, a source, a date, and a review date. Claims rot. A tier 1 result from eighteen months ago may no longer hold for the current version.
Set a review cadence and a stop rule. If the evidence behind a live claim expires or is contradicted, the claim comes down until it is re-proven. Name who has authority to pull it. Keep the file in Google Drive, Dropbox, Confluence, or Notion so the owner and date travel with the claim.
Where your product runs on customer systems, the CISA logging guidance for business systems explains why organizations retain and review event records. That is a security practice, not proof of a business result, but it is a useful model for keeping a dated record of what you claimed and when.
If your launch touches public-sector or standards-bound buyers, the GOV.UK open standards guidance connects open standards to interoperability and reduced supplier dependence. Treat it as a portability prompt for how you describe integrations, not as a rule that decides your own claim ranking.
Common questions
What order should claims go in?
Tier 1 first, then tier 2. Inside each tier, the claim that answers the buyer's primary problem goes first. Tier 3 claims become demo moments. Tier 4 claims stay internal until evidence moves them up.
What if a claim is true but I cannot prove it yet?
Hold it. Move it to tier 4 and keep it internal until evidence arrives. Publishing an unsupported benefit is the failure this whole method exists to prevent, and in the US it also runs against the FTC substantiation standard.
How many claims should the launch message carry?
Fewer than you think. Two or three tier 1 claims that buyers can restate beat ten that blur together. The rest become demo moments or stay in the internal file.
How often should the ranking be reviewed?
At every product release, and on a fixed calendar cadence between them. Evidence expires, versions change, and a ranking that is never revisited quietly becomes a set of claims nobody can defend.
