Brand campaigns, seasonal bookings and campaigns that were simply forgotten also run for months without being a winner in terms of conversions.
This article gives you six concrete signals for spotting a winning ad, a step-by-step guide for the manual check, a checklist to tick off, and the typical false conclusions that come up along the way.
Why “running for a long time” is not proof on its own
The basic idea behind the run-time signal is sound: advertising costs money, and nobody pays voluntarily for something that does not work. Still, that one idea is not enough for a safe judgement. There are at least three situations in which a long run time says nothing about success in the narrow sense: pure brand awareness campaigns that are meant to run long and broad on purpose, seasonal bookings with a fixed end date that has simply not been reached yet, and campaigns that nobody switched off, because managing several accounts and variants gets confusing in practice.
The assessment only becomes reliable when several signals appear together. The following six have proven meaningful in practice, individually as a hint, together as strong evidence.
The six signals of a winning ad
1. A run time of more than three to four weeks
This is the starting point of every assessment, and the answer to the question of how long an ad has to run before it counts as a winning ad: anything under two weeks is usually still a test phase. From three to four weeks of uninterrupted run time, the probability that the ad really delivers results rises clearly, because that period exceeds the usual length of a creative test at most e-commerce brands. The longer beyond that, the stronger the signal, with diminishing returns: the jump from two to six weeks says more than the one from ten to fourteen.
2. Several format variants of the same hook
If a second and third version with the same core argument runs alongside the original ad, but as video instead of a static image, as a carousel instead of a single image, or in a different aspect ratio, that is a sign of iteration on a basis that has already been confirmed. Nobody builds three formats around a hook that does not show results yet.
3. A creator or user-content change during the run
If the hook stays the same but the face changes, with a new creator delivering the same core message, that points to content fatigue alongside a confirmed message. The brand knows the argument works, but swaps the face because the same video loses its effect after a few weeks.
4. Recurring hooks across several campaigns
Some core sentences appear in a competitor’s Ad Library not once but repeatedly over months, once in a spring campaign, once in an autumn campaign, with different imagery. A hook that returns has already proven itself. It is recycled because it demonstrably works, not because the brand has run out of ideas.
5. Growing reach according to the EU DSA disclosure
For ads running inside the EU, the Meta Ad Library shows a reach range, a mandatory disclosure under the Digital Services Act. If that range visibly rises over several weeks, the ad is being scaled up actively. That is a direct signal: a brand does not invest in something that does not work, let alone visibly more than before. One point of context, because this is often misunderstood: the reach range is not a spend or budget figure. What a competitor actually pays is not disclosed by the DSA, and every tool that promises exact advertising spend is estimating or inventing it. Reach, demographics and run time, by contrast, are real values disclosed by Meta.
6. Parallel variants as a sign of scaling
Five, eight or more variants of the same basic motif active at the same time, with slightly different texts, thumbnails or audience wording, are rarely a coincidence. Either the brand is still in a broad test phase, or a confirmed winner is being scaled in several variants in parallel to prevent ad fatigue and cover more audience segments.
Step by step: identify a winning ad in the Meta Ad Library
Once you know how to operate the library, set the country, search for a competitor and open an ad, proceed like this for the actual winning-ad assessment:
- List all active ads of a competitor. Open the profile and show every running ad, not just the first one.
- Note the start date of each ad. Keep a simple list: ad, start date, format. That way you see at a glance which ads have already passed the four-week mark.
- Group by shared hooks. Put ads with the same core argument together, even when format or creator changes. That reveals signals 2, 3 and 4 at a glance.
- Check the reach range of the oldest ads. Open the detail view of the longest-running ads and note the reach range and demographics shown there.
- Check again after one or two weeks. The most important step, and the one most people skip: come back later and compare. Is the ad still there? Has the reach range grown? Have new variants of the same hook appeared? Only the comparison over time turns a snapshot into a reliable assessment.