If one ad is responsible for more than 15% of your conversions, you have a problem. If it's above 25%, your entire account is standing on one leg, and that leg is tired.

We see this constantly when auditing Meta accounts between $1M and $5M in revenue. The dashboard looks great. ROAS is healthy. Conversions are rolling in. Then you check the frequency on the ad doing all the work, and it's sitting at 4.0 or higher, which means the same people have seen it four times or more. That ad isn't winning anymore. It's dying, slowly, while your ROAS number pretends everything is fine.

The trap feels smart while it's happening

Finding a winning ad and riding it hard is not a mistake. It's just Meta doing exactly what Meta does. You tested a handful of concepts, one or two worked, and you poured budget into the winners. That's the whole point of the platform.

The problem shows up a few weeks later. Meta keeps feeding your winning ad to the same high-converting audience because that's the highest probability play in the short term. Frequency climbs. The ad performs well for a while, sometimes a couple of months if you're lucky. Then one morning you open Ads Manager, CPA has doubled overnight, and conversions have dropped by half. Nothing broke. Your best ad just died of natural causes, and because it was carrying a huge share of your conversions, the whole account felt it at once.

The five minute check that tells you the truth

Open Ads Manager, set the date range to the last 30 days, and go to the ad level, not campaign level. Sort by conversions, highest to lowest, and look at four things.

What percentage of total conversions does your top ad represent. Above 15% is a warning sign, above 25% means one ad is holding the account together.

What's the frequency on that ad. Above 3.0 and it's already fatiguing, whether your CPA has noticed yet or not.

How many ads are actively spending and producing at least one conversion per day. Under 5 means your creative bench is dangerously thin.

Most founders assume they're diversified because they have ten ad sets running. Then they check and eight of those ad sets are all showing the same two ads.

Why this keeps catching smart founders off guard

Meta's algorithm is optimizing for short term performance, not your long term stability. It doesn't care that an ad is hitting a frequency of 4.0, it will keep showing that ad to the same people until performance drops, then move on to whatever's next in line, which might not be very good.

Most brands at this stage also don't have a real creative pipeline. It's one person on Canva, or a freelancer delivering three videos a month. When a winner starts fatiguing, there's a two to three week gap between noticing the problem and having a genuine replacement ready to test, and that gap is exactly where performance craters.

And almost nobody watches frequency. Founders track ROAS, CPA, and CTR. Frequency is the leading indicator of fatigue, and by the time CPA starts climbing, the ad has usually been overexposed for weeks already.

The fix doesn't need to be a content studio

You don't need fifty ads running at once. You need a bench deep enough that no single ad can take the account down with it when it dies.

Aim for three to five genuinely new concepts tested per week, not five crops of the same ad with a different thumbnail. Different hooks, different angles, different formats, static, video, UGC, founder to camera.

Run a small dedicated testing campaign, something like fifty to a hundred dollars a day, and push every new creative through it first. Winners graduate to your scaling campaigns. Losers get cut within three to five days, no sentimentality required.

And keep at least five ads actively spending and converting at any given time. If that number drops below five, creative production becomes the most important thing on your plate this week, above everything else.

Run the audit today, not after it breaks

Do the five minute check. If your top ad is above 15%, your frequency is above 3.0, or your active ad count is under 5, you have creative concentration risk, and it's genuinely just a matter of time before it catches up with you.

The best moment to build a new creative pipeline is before your winner dies. Everyone finds this out the hard way once. You don't have to be one of them.