Concept testing is the practice of running genuinely different creative ideas against each other to find which one works, before spending time polishing details. A concept is the core idea behind an ad: the problem it speaks to, the angle it takes and the kind of person it is for.
How concept testing works
Take a London personal trainer. Three distinct concepts might be “short on time: 30-minute sessions before work”, “post-injury: get strong again safely” and “beginners: no judgement, no crowded gym”. Each is a different reason to book, aimed at a different person. Testing one against another tells you which motivation is strongest in your market.
That is different from iteration testing, which changes one element of a proven ad, such as the headline, the first frame or the colour of the background. Iterations make a good ad a little better. Concepts tell you which ad to build in the first place.
In Meta, concept tests are usually run by giving each concept its own ad or ad set with similar budgets, or by using the A/B test tool in Ads Manager, which splits the audience so the same person does not see both. Each concept should have more than one execution, such as an image and a short video, so you are testing the idea rather than one piece of design.
Why it matters for a UK business
Meta’s delivery system now does much of the targeting itself, and the creative increasingly decides who sees an ad. Since Meta introduced its Andromeda ad retrieval system, accounts with a range of genuinely different ads tend to give it more to work with than accounts with ten near-identical versions. That makes creative diversity a practical concern, not a design preference.
Concepts also wear out at different speeds. A strong idea can carry many fresh executions over months, while a weak one fades however often you redesign it. Knowing which concepts are proven lets you respond quickly when creative fatigue sets in, by making new versions of an idea that already works.
For a smaller business, concept testing is also cheaper market research than it sounds. A few weeks of modest spend can show that customers respond to “fixed price” far more than “experienced team”, which is useful well beyond the ads: on your website, in emails and on the phone.
Common mistakes
- Calling it a concept test when the ads differ only in colour or headline wording.
- Testing too many concepts at once on a small budget, so none gets enough results to judge.
- Declaring a winner after two days, before results have settled.
- Judging on click-through rate when the concept with fewer clicks brings cheaper enquiries.
- Killing a concept because one weak execution failed, without trying it in another format.
How to act on it
Write each concept as one sentence: who it is for and why they would act. Pick two to four that are truly different, and make two executions of each. Agree the measure before launching, normally cost per lead or cost per sale, and give the test long enough to gather a meaningful number of results.
When a concept wins, move to iteration testing within it and keep a log of what each test taught you. A clear creative testing framework stops the account sliding back into random changes. Running these tests and turning the results into the next round of ads is part of my Facebook ads management.