Swap the logo. Does your ad still work?

Take the headline you are proudest of. Replace your company name with your closest competitor's. Read it again.

If it still works, it was never your ad.

You will: Judge any headline or ad line in twenty minutes, and be able to explain the verdict to someone who disagrees.
Prerequisites: A draft headline or a live ad, a rough sense of your audience, one thing you know is genuinely true about your product.
Time: 20 minutes per ad concept.

Why copy reviews go wrong

Most copy gets approved or killed on taste. One person says it feels flat. Someone else says it feels strong. Nobody can point at anything specific, so the loudest voice in the room wins and the line goes live.

That is how "Innovative solutions for a better tomorrow" survives four rounds of review. Everyone senses something is off. Nobody has the language to say what. The line is grammatical, confident, on brand, and completely empty, and none of those first three things is a defense.

What follows is not a writing method. It is a filter. Three questions, each answerable with a yes or a no, that turn a matter of taste into a decision anyone on the team can make and defend.

The three questions

Write the line down. One sentence. Then ask:

1. Can I picture it? Does a specific image appear in my head?

2. Can I prove it wrong? Is there a version of reality where this claim is false?

3. Could only we say it? Does it break when a competitor's name replaces ours?

Three yeses means the line is worth spending money on. Three nos means start over. Most copy lands in between, and that is where the actual work is.

Pro tip: Run the same three questions on your two or three strongest competitors' ads. If their copy also passes, you need a different angle. If it fails, you gain a structural advantage the moment you get specific.

Question 1: Can I picture it?

Read the line, then look away. Does anything appear? A person, a place, a moment, an object? If nothing does, the line is operating at category level rather than human level.

This is not a test of how poetic the writing is. It is a test of whether the reader can place themselves inside the sentence.

New Balance is the textbook case: "Worn by supermodels in London and dads in Ohio." You see two people instantly. You see two rooms. The contrast does the work, and it sticks because it is precise, not because it is clever.

Question 2: Can I prove it wrong?

A claim nobody can argue with is a claim nobody has to believe.

"Don't just get a job. Change an entire industry." Bold, confident, and completely unfalsifiable. Nobody can measure whether taking a job changed an industry, which means the sentence carries no risk. A sentence with no risk carries no weight either.

Compare that with claims that have stakes. "Invoice processing takes half as long." "Your first campaign is live in 48 hours." "We only work with companies between 20 and 200 people." Each of those can be checked, and each one can be wrong. That is precisely what makes them credible.

Key principle: If a claim cannot be proven false, it cannot be believed either. Falsifiability is where credibility comes from.

This does not mean every ad needs a statistic. It means the claim underneath has to touch something real: a time frame, an outcome, a named audience, a specific constraint. The same discipline applies further down the funnel, which is why the guide on what actually makes websites convert keeps returning to specificity as the thing that moves numbers.

Question 3: Could only we say it?

This is the swap test, and it is the fastest of the three. Read the line out loud with a competitor's name in it. If it still works perfectly, it was never yours.

Most teams skip this one. They spend their energy making copy sound good and never ask whether it could belong to somebody else. "Quality you can trust, service you can count on" could sit on the website of a bakery, a law firm, or a freight company. It belongs to nobody, which is why nobody remembers it.

Ownability rarely requires inventing something. It requires being honest about what is actually specific to you: the industries you serve, the size of company you understand, the languages you work in, the thing you refuse to do. For B2B companies in Switzerland and Europe, that specificity almost always lives in the details rather than in the adjectives.

Timly is a useful reference here. Before their rebrand, their messaging read like every other asset management platform. Afterwards it reflected who they actually serve and how, which is what made it defensible. The full story is in the Timly rebrand case study.

A worked example: one line, three verdicts

Theory is cheap. Here is the filter running on a real sentence.

The line on the table: "Innovative solutions for modern manufacturing."

Three nos. Now watch what each one forces.

Picture it: "The machine tells you it needs a part before it breaks." Something happens in the reader's head now. But most predictive maintenance vendors could still run this.

Prove it wrong: The claim has become testable. Either the warning arrives before the failure or it does not. That is a real bet, and the line is stronger for carrying it.

Own it: "Your 40-year-old press tells you it needs a part before it breaks." One word changes everything. Competitors selling to new factory floors can no longer run this line, and the buyer sitting on decades-old equipment recognizes themselves immediately.

Same claim, same length, three yeses. Nothing was invented along the way. The line only got more specific about a truth that was already there.

What to do when the answer is no

A failed verdict tells you which repair to make. Work in this order, because reversing it tends to produce clever lines that are still generic.

1. Concretize. Replace every vague noun and adjective with a specific one. "Better results" becomes "three new enterprise clients in Q1." "Faster delivery" becomes "live in 48 hours."

2. Anchor. Add one detail that ties the claim to reality: a number, a time frame, a named audience, a specific outcome.

3. Own it. Ask what makes this true for you and not for the next vendor down the list. If you cannot answer, the copy is not the problem.

Troubleshooting: when the verdict is hard to reach

If the third question is the one you keep failing, the copy is rarely the issue. It usually means the positioning underneath has never been made explicit, and no amount of rewriting will produce a line only you can own. Getting that foundation right is what the Brand Transformation process at HIVER is built around.

The verdict is the point

Great advertising does not need to sound bigger. It needs to be concrete, falsifiable, and unmistakably yours. The New Balance line works not because it is witty but because it is true and precise. It describes a real fact about who wears the shoe, and that fact creates differentiation without a single exaggeration.

Before the next line goes live, run the three questions and write the answers down. Three nos means rewrite. Three yeses means you have something worth putting a budget behind, and a reason you can give when somebody asks why.

And once your copy is that specific, it is worth making sure it can be found as well as read. The guide on how to get found on ChatGPT covers the next layer: getting AI search to cite the brand you just sharpened.