How to audit your own claims
Guide · Frameworks · 4 min read · last verified 2026-07-27
A claims audit is a systematic pass over your public-facing pages that inventories every factual assertion, attaches each one to the evidence behind it, and sorts what remains into three buckets: provable, provable with work, or retire. It is unglamorous, occasionally embarrassing, and one of the highest-leverage exercises available to a marketing team — because the gap between what a company asserts and what it can substantiate is exactly where buyer trust leaks out.
The audit matters more now than it did a few years ago, for a reason covered in its own section below: verification has become cheap. When a buyer can ask an assistant to check a vendor's claims against everything publicly written about that vendor, unproven assertions stop being harmless puffery and start being liabilities you published yourself.
Step one: inventory every claim
A claim, for audit purposes, is any sentence a skeptic could answer with "says who?" That includes capability claims ("automates your entire close process"), comparative claims ("the fastest way to reconcile"), scale claims ("trusted by thousands of teams"), outcome claims ("customers cut onboarding time"), and compliance claims ("enterprise-grade security"). Adjectives count when they smuggle in facts: "effortless" is a claim about your onboarding, whether or not it was meant as one.
Sweep the pages where claims concentrate: the homepage, product pages, pricing, the security or trust page, top landing pages, and the about page. Put every claim in a list with its page and its exact wording. Teams doing this for the first time are usually surprised twice — by how many claims they have published, and by how few they remember deciding to make. Claims accrete; nobody owns the sentence written four years ago that still leads the homepage.
Step two: attach each claim to its proof
Beside each claim, record the strongest evidence you could show a skeptical buyer today — not evidence you believe exists somewhere, but evidence you can link to or produce on request. The grading system from what is a proof point applies directly: a verifiable artifact outranks a named customer outcome, which outranks third-party validation, which outranks self-assertion. Be ruthless about the difference between "our customers say this in calls" and "a named customer says this in public." Only the second one is proof a buyer can check; converting the first into the second is the craft covered in how to write a case study buyers believe.
The honest version of this exercise produces an uncomfortable column: claims whose best available proof is the claim itself, repeated. That column is the point. You cannot fix what you have not listed.
Step three: grade the gaps
Sort every claim into one of three buckets.
Provable. The evidence exists and is attached — linked from the page or one click away. These claims need maintenance, not work: evidence goes stale, customers churn, certifications lapse.
Provable with work. The claim is true, but the proof is trapped — in sales decks, in customer calls, in data nobody has published. These become your evidence backlog, prioritized by how load-bearing the claim is. A weak claim on a minor page can wait; an unproven claim in your homepage headline cannot.
Retire. The claim cannot be substantiated at reasonable cost, or is no longer true, or was never quite true. Rewrite it narrower or remove it. Retirement feels like loss and usually reads as strength: a page of modest, provable claims tends to outperform a page of grand, hollow ones with skeptical readers — and the security page is where this trade shows up most visibly, per what belongs on a trust page.
Why unproven claims are getting costlier
Verification used to be labor. A diligent buyer might check a reference or two; nobody cross-examined a homepage. That asymmetry has narrowed: buyers increasingly appear to ask assistants to summarize, compare, and sanity-check vendors, and those systems read your pages next to everything else written about you. In our observation, a claim that is contradicted elsewhere — or merely unsupported everywhere — tends to fare worse in that synthesis than a narrower claim with visible evidence behind it. The audit is how you find those exposures before your buyers' tools do. This is the same evidence-first logic Magrios applies to measurement: know what is actually being said and substantiated before deciding what to fix.
Make it a cadence, not a project
A one-time audit decays immediately, because publishing continues. Two habits keep it alive. First, a recurring pass — quarterly is a common rhythm — re-checking the inventory against pages as they exist now, since claims have a way of creeping back in. Second, a publish-time check: every new page adds its factual assertions to the inventory with proof attached, or explains why not. Give the inventory an owner; a spreadsheet nobody owns is a spreadsheet nobody updates.
Done this way, the audit changes writing culture upstream. Teams that know every claim will need a proof column attached tend to draft claims they can prove — which is the quiet, compounding benefit: the audit stops finding problems because the problems stop being written.