Support a marketing claim by recording exactly what it asserts, who or what it applies to, the evidence source, its limitations, approval status and review date. Objective claims need evidence appropriate to their meaning before publication. If the evidence only supports a narrower statement, narrow the claim. Place proof close enough that readers can judge it without searching for a hidden qualification.

Classify the claim before choosing proof

Different claims require different support. Start by identifying the type:

  • Capability claim: what the product or process can do.
  • Outcome claim: what changes for a customer.
  • Performance claim: speed, accuracy, reliability or scale.
  • Comparative claim: better, faster or cheaper than an alternative.
  • Customer claim: a named organisation’s experience or opinion.
  • Market claim: popularity, leadership or category status.
  • Environmental or social claim: an impact or responsibility assertion.

The wording determines the burden. “Can automate invoice matching” is different from “eliminates invoice errors”. “Customers in 12 countries” is different from “trusted globally”. Do not choose the biggest phrase and then look for something vaguely supportive.

The UK CAP Code says marketers must hold documentary evidence for objective claims likely to be regarded as objective. It also requires qualifications to be presented clearly rather than contradicting the headline claim.

Build a claim-to-evidence register

Use a shared register with these fields:

  1. Claim exactly as published.
  2. Page, campaign or channel.
  3. Claim type and intended interpretation.
  4. Audience and applicable conditions.
  5. Evidence source and methodology.
  6. Limitations or required qualification.
  7. Evidence owner and approver.
  8. Permission status for customer material.
  9. Publication and review dates.
  10. Action if the evidence expires or changes.

This register supports the proof layer in a messaging hierarchy and stops old statistics surviving after the product, sample or market changes.

Match evidence strength to the claim

Product documentation or a repeatable demonstration may support a narrow capability claim. Verified operational data may support a performance claim if the measurement and relevant conditions are clear. Comparative claims need a fair basis of comparison. Customer stories require permission and should identify what is representative, exceptional or subjective.

Ask:

  • Does the evidence measure the same thing the sentence claims?
  • Is the sample relevant to the audience named?
  • Are the time period and conditions current?
  • Could a reasonable reader infer something broader?
  • Is a material limitation hidden in small print?

The UK government’s overview of advertising codes of practice points businesses to the applicable broadcast and non-broadcast codes. Legal requirements vary by market and sector, so high-risk claims need appropriate specialist review.

Illustrative example: a scheduling product

Illustrative example: a team measures ten internal planning exercises. Its prototype reduces the median time taken to produce a schedule from 50 minutes to 35 minutes under the test conditions.

Unsupported headline:

Plan every project 30% faster.

More accurate internal-test description:

In ten internal test exercises, the prototype reduced median schedule-building time from 50 to 35 minutes. Results may differ with project complexity and data quality.

Whether that statement is suitable for public marketing depends on the method, relevance and applicable rules. The example shows why a percentage without context can imply more certainty and universality than the evidence provides.

Put proof into the reading path

Evidence should appear where it helps the decision. Useful patterns include:

  • a short source note immediately after a statistic;
  • a product demonstration beside a capability statement;
  • a methodology link near a benchmark;
  • an approved customer quotation with named context;
  • a clear qualification in the same visual field as the claim.

Avoid a generic “trusted by” logo wall unless permission and the nature of the relationship are clear. A customer logo does not automatically prove the result claimed elsewhere.

In a messaging framework, attach at least one reason to believe to every primary message pillar. Mark gaps openly. A visible gap is a useful research task; an invented proof point is a risk.

Review environmental and AI claims carefully

Broad terms such as “sustainable”, “carbon neutral”, “accurate” or “bias-free” can carry interpretations beyond the available evidence. The UK Competition and Markets Authority publishes a Green Claims Code explaining principles for environmental claims, including clarity, the full lifecycle and supporting evidence.

For AI-enabled offers, distinguish what the system does, what humans review, which data or conditions apply and what uncertainty remains. Do not convert a model benchmark into a guaranteed customer outcome.

Limitations and trade-offs

Strong substantiation takes time, and early products may have little outcome evidence. The answer is not silence or exaggeration. Lead with demonstrable capability, process transparency and the evidence currently available, then improve the evidence programme.

Too many qualifications can make copy unreadable. If a claim requires a paragraph of caveats to remain accurate, simplify or narrow it. Evidence review is not legal advice; regulated sectors and high-impact claims require qualified oversight.

Claim review checklist

  • Copy the exact public claim into the register.
  • Identify objective and implied meanings.
  • Link to the primary evidence, not a screenshot of it.
  • Record sample, method, date and conditions.
  • Confirm customer permissions.
  • Put material qualifications beside the claim.
  • Assign an owner and review trigger.
  • Remove or revise claims when support changes.

Frequently asked questions

Is a customer testimonial enough proof?

It can support that customer’s reported experience when genuine, approved and accurately presented. It may not substantiate a universal performance claim.

Can we say “best” if it is obviously advertising language?

Context matters. An apparently subjective slogan can still imply an objective comparison. Treat comparative claims cautiously and seek appropriate review.

What if competitors use claims without sources?

Their practice does not change your evidence. Build trust by making your basis clear and aligning every claim with what the offer can deliver.

Your useful next step

Audit the five strongest claims on your homepage. For each, record the evidence and likely reader interpretation. Revise any sentence where the interpretation is broader than the support, then connect the result to your value proposition.

Publication note: This guide is published by To-Do Growth and reviewed under our Editorial Policy. Illustrative examples are labelled and are not presented as customer case studies.