Facebook ads stopped working? Run this 10-minute check

For a few weeks the ads did their job. Leads came in, the phone rang, or orders showed up, and the cost per result looked fine. Now you open Ads Manager and the same budget buys fewer results at a higher price. It is easy to jump to the worst explanation: Facebook is broken, the audience is used up, or the money is simply gone. When Facebook ads stopped working for you, the cause is usually something specific that you can find in about ten minutes. It might be a tired ad, a small audience, a page that loses people after the click, or a string of edits that kept restarting delivery. Each of those shows up in a different set of numbers, and each calls for a different fix. Before you spend more or rebuild the campaign, look at the right columns first.

What should I check first when my Facebook ads stop working?

Start with a comparison, not a fix. Open Ads Manager and select the campaign or ad set that used to work. Set the date range to the first two weeks after launch, then repeat it for the most recent two weeks. Add these columns: Frequency, Cost per result, Reach, Impressions, CPM, Link CTR, Landing Page Views, Results and Amount Spent. Then read the three levels in order.

• Campaign level: This shows the overall change in cost and results between the two periods.
• Ad set level: This shows whether one audience or placement slipped while another held steady.
• Ad level: This shows whether one creative wore out while another still earns its keep.

Also look at the Delivery column. Meta shows statuses such as Active, Learning and Learning limited there, and Frequency is available as a reporting metric at campaign, ad set and ad levels, according to Meta's Ads Manager guidance. Write the two sets of numbers down side by side. You are looking for which numbers moved, because that tells you where to look next.

How do I tell a tired ad from a problem after the click?

The same drop in results can come from very different places, so match what you see to the pattern below. Frequency, click-through rate and conversion rate together tell most of the story.

Meta describes creative fatigue as repeated exposure that can raise cost per result. That is one possible cause, not the only one. If clicks are holding steady while leads fall, the ad is still doing its job and something later in the path is breaking, so a new ad will not fix it. Fill in the form on your own phone, call your own number, and look at how long a lead waits for an answer.

Is there a frequency number that means my ad is tired?

No single number does. Frequency is the average number of times each person saw an ad, and what it means depends on your budget, your audience size, your placements and how long the campaign has been running. Meta's Business Help Center treats it as something you read alongside results, not a line you cannot cross.

So, the better question is whether frequency is climbing while click-through rate, conversion rate or cost per result gets worse. If frequency is high but results are steady, the ad may be doing fine. If frequency is modest and results fell anyway, look at the page, the offer and the tracking before you blame fatigue. A low frequency does not prove the creative is strong, either; it may only mean the ad is not being shown much.

Does a small local audience wear out faster?

It can. A smaller geographic audience gives the same budget fewer eligible people to reach, so the same people see the ad again sooner. A narrow radius, a tight age range, several stacked interests and a small retargeting pool each shrink the group, and together they can concentrate delivery on a few thousand people.

Meta recommends broader targeting and more placements when the available audience is too limited for stable delivery. If your frequency is rising and your radius is tight, widen it a notch or loosen one filter and watch the next several days. Do not change the radius, the age range and the interests at once, because then you cannot tell which change helped. If you run a local campaign, our post on local radius targeting for Facebook ads covers how to size the area.

Which edits restart the learning phase, and which can I leave alone?

The learning phase begins after an ad set is created or significantly edited. Meta says delivery is less stable during learning and usually settles after about 50 optimization events following the last significant edit. That figure is a typical reference, not a promise of results, and the time it takes depends on how often your campaign gets those events.

Meta lists changes to targeting, creative, optimization event or bid strategy as edits that can restart learning. Adding a new ad can restart it too, and so can pausing an ad set for seven days or longer. A budget change can reset learning when it is large enough to change delivery in a real way; a small nudge is a different matter. Not every edit is a significant edit.

A Learning limited label is not a penalty, according to Meta. It means the ad set is not getting enough optimization events to settle, which is often a sign that the budget is split across too many ad sets or the audience is too narrow. Merge overlapping ad sets and let the data build up before you edit again.

In what order should I rule out the causes?

When several things could be wrong, work through them in a practical order, from the cheapest to check to the slowest. This is a working sequence for troubleshooting, not a ranking of how common each cause is.

1. The offer or message no longer matches what people want right now.
2. The same creative is being shown over and over to a small audience.
3. The campaign is split across too many ad sets or placements.
4. Repeated budget or targeting changes keep restarting learning.
5. The landing page, form, phone handling or follow-up fails after the click.
6. Conversion tracking or attribution is incomplete, so results are under-counted.
7. Season, competition or auction conditions changed.
8. Lead quality or sales handling changed while the platform numbers looked acceptable.
9. The audience is too narrow for the budget and the objective.

Number six deserves a second look. If your Meta Pixel or Conversions API setup stopped sending events, Ads Manager can show fewer results while your phone and inbox look normal. Compare the results Meta reports with the leads and sales you can actually count. Platform-reported results and real customers are not the same thing, so check what happened after each lead arrived.

How do I refresh creative on a small budget?

You do not need a new ad every week. There is no set calendar such as "replace your ads every seven days." Meta's guidance favors creative variety, testing, and avoiding unnecessary edits. On a small budget that means keeping a small set of materially different ads, not many near-copies of the same one.

A useful set covers four angles:

• The customer problem: Open on the situation your customer is in when they start looking.
• Proof: Show a real result, a real review or a real project, only if you have one.
• Demonstration: Show the product or service being used, or the work being done.
• The offer: State what is included, the price or the next step, in plain words.

When results and exposure trends both turn worse, add one or two new concepts beside the proven ad and compare them. Do not replace everything at once, and do not pause your only proven ad because a number looks high. Remember that a new ad can restart learning, so give each addition a few days before you judge it.

If Reels is one of your placements, Meta recommends testing native vertical video with audio and attention to the safe zones. Meta reports a 2023 analysis of 15 split tests in which native 9:16 Reels creative outperformed still images on Reels. Treat that as a reason to test vertical video, not as a guarantee. Whatever the new ad says, the FTC's advertising guidance expects claims to be truthful and backed up, so only promise what you can deliver.

When is it worth paying someone to look at the account?

Often you do not need to. The check above is free, and the answer may be a fresh ad, a wider radius, a repaired form, a faster reply to leads or a few days without edits. Those fixes often take an afternoon, not a new campaign or an agency retainer.

Help starts to make sense when you cannot read the account with confidence, when tracking is unreliable, when making new creative is the bottleneck, or when several of the causes above overlap and you cannot separate them. Then a focused audit or a short repair job can be a better fit than a long monthly engagement. What a specialist adds is diagnosis, disciplined testing and a line from ad results to leads, appointments and revenue. No one can promise that a new campaign will bring back the old cost per lead, so be wary of anyone who does.

If you are still deciding what to spend, our post on how much to spend on Facebook ads sets out a way to size a budget.

Frequently Asked Questions About facebook ads stopped working

Why did my Facebook ads stop working?

Common causes are creative fatigue, a narrow audience, repeated learning restarts, a weaker offer, tracking problems, a page that fails after the click, or a change in auction conditions. Compare the first two weeks with the last two weeks to see which one fits.

What frequency is too high on Facebook ads?

There is no single cutoff. Read frequency alongside click-through rate, cost per result, conversion rate and audience size, and act when frequency rises while results fall.

Should I change the creative or the audience first?

If frequency rose and response weakened, test materially different creative first. If delivery is constrained, broaden the audience or consolidate ad sets.

Does changing the budget restart learning?

A budget change large enough to alter delivery can count as a significant edit. Not every budget change counts as significant.

How long does the learning phase last?

Meta says performance usually settles after about 50 optimization events following the last significant edit. How long that takes depends on your campaign.

Should I pause a tired ad?

Add a new concept and compare it with the existing ad. Do not pause your only proven ad because a generic frequency number looks high.

Wrapping Up

Ads that fade after a good start usually have a cause you can find. Compare the first two weeks with the last two, add Frequency and Cost per result, and sort what you see into three groups: a tired ad or small audience, a problem after the click, or a setup that keeps restarting learning. Read frequency beside response, make one change at a time, and keep a control ad running while you test.

Done that way, you stop paying to guess. You spend on the one fix that matches your numbers, and you keep a record of what changed, which makes the next dip faster to read.

If the check points to something you would rather not handle yourself, Web Leveling can look at the account with you, from the tracking to the creative to how your leads get answered. Our Facebook ads work starts with a diagnosis, and if the fix is a new ad or a corrected form, you may not need us at all. We work with small and medium businesses across the country and overseas. Send us what your numbers are showing, and we will help you read them.

Terms