There's a particular kind of phone call we get a lot. The ads were working. Leads were coming in. Nothing was changed. And then, over about ten days, it all quietly stopped.
It's rarely mysterious. It's almost always one of five things, and they're worth checking in this order.
1. Creative fatigue
This is the most common answer by a wide margin, and the most boring.
Your audience is finite. In a local service business it might be a few hundred thousand people, and the ones most likely to respond saw your ad first. By week six, the people still being shown it have seen it eight or nine times and have already decided they're not interested.
The metric that gives it away is frequency. If it has climbed past roughly 3–4 in a local audience while your click-through rate has drifted down, you haven't got a targeting problem. You've got a "they've all seen it" problem.
The fix is new creative, not a new audience. Same offer, genuinely different execution — a different opening line, a different image, a different format. Not the same ad with the headline reworded.
2. You edited a campaign that was already working
Facebook's delivery system re-enters a learning phase when you make a significant edit to an active ad set — meaningful budget changes, audience changes, creative swaps, optimisation event changes.
During learning, performance is unstable and usually worse. That's expected. The problem is what people do next: they see two bad days, edit it again to fix it, and reset the clock. Three or four rounds of that and the campaign never gets a stable run at anything.
If you must change something, make it a new ad set and let the old one keep running until the new one proves itself.
3. Your tracking broke and nobody noticed
This one is genuinely invisible from inside the ads manager, because the ads keep running and the numbers keep updating. They're just wrong.
It happens when someone updates the website, changes the thank-you page URL, swaps the contact form, or a consent banner starts blocking the pixel. The platform stops receiving conversion signals, so it stops being able to optimise toward them — and quietly drifts toward cheap clicks instead of good leads.
Check that a test submission still registers as a conversion. Do it monthly. Do it the day after anyone touches the website.
4. The market moved and the ad didn't
Seasonality is real and specific to your trade. So is a competitor launching a loud campaign in the same postcodes, or a new franchise opening nearby with a national ad budget.
Look at your cost per thousand impressions over time. If it has risen sharply while your creative performance held steady, you're not doing worse — you're paying more for the same attention because someone else showed up to bid.
The answer there isn't a better ad. It's a better offer, or a narrower audience where you can still win.
5. Nothing broke — you just noticed the same thing twice
Small local campaigns are noisy. A week with four leads and a week with eleven can be the same campaign performing identically; there just isn't enough volume for a week to mean anything.
Before you tear anything down, look at a rolling 28-day figure rather than week-on-week. A surprising number of "the ads stopped working" emergencies are two ordinary weeks in a row.
The order matters
Work down that list rather than starting at the bottom. Most people begin by rebuilding the audience, which is the fifth-most-likely cause and the most disruptive fix. Check frequency first. Then check whether anyone edited anything. Then check the pixel.
It's usually the creative. It's very often the creative.
If you're staring at a campaign that used to work and you'd rather not guess, we'll go through the account with you on a call and tell you which of the five it is.