Skip to content

Claims — the buying reasons you test

A claim is one assertion your ads lead with — and, more importantly, a hypothesis. It can win in one market and die in another. That is not a flaw; it is the point. You run ads to find out which buying reason actually moves people, and the verdict comes back to the claim so the next round starts from what you learned.

Because of that, five polished rewordings of the same idea are worth nothing. What you want is a spread: different reasons to buy, each one testable on its own.

Every claim argues exactly one buying reason.

AngleWhat it argues
categorywhat changes about the buyer’s setup — the new thing in their world
paina moment they recognise before any solution appears
outcomewhat their life looks like afterwards
pricean anchor against what they already pay
speedwhat exists at the end of the wait
controlwhat cannot happen without them
proofone real number, with its source
comparisonwhy the alternative way structurally cannot do this

Autopilot tracks which angles you have covered and which are still open, so a round can go where there is nothing yet.

Cold audiences do not know you. They can only be reached with what they already recognise — the category and the problem:

  • Cold (awareness, pain): category, comparison, pain
  • Warm (benefit): outcome, control, speed, proof, comparison
  • Deciding (offer): price, control, speed, proof

That is why a price claim is useless in a cold ad: it answers a question the reader has not asked yet. Save it for the moment they are choosing.

What separates a strong claim from a forgettable one

Section titled “What separates a strong claim from a forgettable one”

A category claim names what changed, not what you are called.

Weak: “X is an AI assistant for bookkeeping.” Strong: “There is a new seat at your table, and it works the night shift.”

A pain claim is a moment, not a condition. Conditions are forgettable; moments have a time, a witness and a consequence.

Weak: “Managing invoices manually costs time.” Strong: “You find out the invoice was never sent when the customer calls about it.”

An outcome claim is the scene afterwards, not the saved hour.

Weak: “Save 10 hours a week.” Strong: “You leave at six and nothing is waiting for you tomorrow.”

A price claim anchors against what they already pay.

Weak: “€149 a month.” Strong: “€149 a month, less than one afternoon of the agency you are briefing anyway.”

A control claim names what cannot happen. “You stay in control” is a mood; a boundary is checkable.

Weak: “Full transparency and control.” Strong: “No euro is spent and nothing goes public without your yes.”

A proof claim carries one number and no adjectives. If you do not have the number, leave the angle empty rather than soften something into place — an unevidenced proof claim is the one kind that costs you trust.

You choose which angles this round should argue, and which offering it is for. Autopilot writes candidates, shows them to you, and nothing is saved until you take it. Run as many rounds as you like; what you have already seen will not come back.

Behind that, every candidate is checked twice before you ever see it: once against the facts you have said must hold (your claim boundaries), once against a craft bar. Candidates that fail are rewritten once, then dropped. What reaches you has already survived both.

A round takes a few minutes. It keeps running if you close the window.

  • It becomes citable substance for messages, landing pages and ads.
  • It can be tested: one claim per test cell, so the result is about the buying reason and not about the wording.
  • Its verdict is recorded on the claim itself — killed, iterate, or scale — so the learning survives the campaign that produced it.

Approved claims are retired, never deleted, once creatives and tests carry their id. Drafts you never used can be deleted outright.

For a first test: two to four claims across different angles, for one offering. One claim gives you nothing to compare. Ten paraphrases of one angle give you the same nothing, more expensively.

Start cold — category, comparison, pain — and let the market tell you which reason to build on.