When the same people keep seeing your ad

Forty thousand impressions, four thousand people. Divide one by the other and you get the number almost nobody looks at: on average, every person who saw your ad saw it ten times. Some of them saw it twice. Some of them saw it thirty times over three weeks, on their phone, at breakfast, while waiting for a bus, until the sight of your logo started to feel like a mild grievance.

That ratio has a name — frequency — and it sits quietly in your ad account next to the numbers you actually check. It is worth checking, because it explains a thing that otherwise feels supernatural: the campaign that worked beautifully for eleven days and then, with nothing changed on your side, stopped.

Why the number climbs on its own

Frequency rises because your budget is a fixed amount of money per day and your audience is a finite number of humans. Spend does not create new people. Once the platform has shown your ad to everyone it can reach who matches your targeting, the only thing left to do with tomorrow's money is show it to the same people again.

The tighter your targeting, the faster you hit that wall. Say your audience is forty thousand people and your budget buys eight thousand impressions a day. In five days you have technically covered the whole pool once — in practice sooner, because the platform serves the easy-to-reach and cheap-to-win first. By week three you are on your fifth or sixth pass through the same faces. The ad did not get worse. The audience just ran out of first impressions to give you.

A tired ad and a tired audience look identical

Both show up as the same sad shape: cost per result climbing week over week. The fix is different for each, which is why guessing is expensive.

Look at reach and impressions separately rather than at spend. If reach kept growing while results fell, you are still finding new people and they are not biting — that is a message problem, or an audience that was never a fit. If reach went flat while impressions kept climbing, you found everyone available days ago and have been buying repeats since. Same chart, opposite cures: one needs a new argument, the other needs new people.

a bored audience and a bad ad cost you exactly the same

Change one thing, and change it in the right order

Creative first, because it is the cheapest thing to swap and the most likely culprit. Not a new colour — a new argument. If the current ad leads with the price, lead with the problem. If it leads with a feature, lead with the annoyance the feature removes. A genuinely different claim gets treated as a new ad by the people seeing it, which is the whole point.

Audience second. Widen the targeting, or add a different one alongside it, and give the platform more humans to find. This is also where a lot of small accounts discover they were narrow for no good reason — a definition that felt precise on Monday was mostly a guess, and it cut the pool to a size that guaranteed saturation by month end.

Budget last, and only once the other two are healthy. More money against an exhausted audience buys more repeats of an ad people have already declined, which is the most efficient way to spend a week proving something you could have known on day one.

Sometimes ten is the right number

High frequency is not automatically waste. If you sell something expensive that takes two months to decide on, and your realistic audience is three hundred companies, then being seen ten times by the right people is not fatigue — it is how anyone remembers you exist at all. The number only becomes a problem when you are paying repeat prices for an audience that decides in an afternoon.

So the threshold is not a universal figure you can copy from a blog post. It is a question about your own buying cycle: how many times does someone need to see this before a yes becomes possible? Anything meaningfully past that is money buying familiarity you already had.

One caveat worth holding on to: each platform counts only the impressions it served, so the frequency you read in one account is not the total attention you are taking from someone who also follows you, gets your emails, and saw you somewhere else last Tuesday. The real number is always higher than any single account admits.

We pull these numbers from whichever ad accounts you have connected and put reach next to impressions rather than burying one of them, because the two side by side are what make saturation visible. When we draft a replacement campaign for a tired one it arrives paused, with the new angle written out, so the decision to spend again is still yours to make on purpose.

The thing to do this week is small: before you touch a budget on a campaign that is drifting, divide impressions by reach and see what comes back. A four means you have room. A twelve means your next dollar is buying the eleventh showing of an ad that failed ten times already — and no amount of budget fixes an argument the audience has finished considering.