Facebook ads usually "stop working" for one of six reasons: broken tracking, a delivery problem, tired creative, an exhausted audience, a weaker offer or landing page, or slow follow-up that makes good leads look bad. Check them in that order. Most fixes are cheap; the expensive mistake is raising the budget before you know which one it is.
Below is the checklist we run when a client tells us their Meta ads suddenly went quiet. Work through it top to bottom. Each step takes minutes, and the earlier steps rule out the problems that would make every later diagnosis wrong.
First: did the ads stop working, or did the reporting?
Check tracking before you touch a single ad. A surprising number of "dead" campaigns are still producing enquiries that Ads Manager simply can't see.
Run these checks:
- Compare Ads Manager with your inbox or CRM for the same dates. If your WhatsApp, form inbox or CRM shows leads but Meta shows none, tracking broke, not the ads.
- Open Events Manager and look at the last 7 days for your key event (Lead, Purchase, Contact). A sudden drop to zero usually means a site update removed the pixel or changed the thank-you page.
- Check the Conversions API connection. If you rely on the browser pixel alone, ad blockers and privacy settings hide part of your results. Our explainer on why the pixel alone isn't enough anymore covers the fix.
- Confirm the optimisation event still fires. If a campaign optimises for "Lead" and that event stopped arriving, Meta is now optimising blind.
If tracking broke, fix it and give the campaign a week before judging anything else.
Is Meta actually delivering your ads?
Look at the Delivery column and the account quality page next. Ads that aren't serving can't fail on performance; they just aren't running.
| Delivery status or sign | Likely cause | Fix |
|---|---|---|
| "Rejected" or "In review" | Ad policy flag | Read the reason, edit the ad or request a review |
| "Learning limited" | Too few conversions for the budget and audience | Combine ad sets, broaden the audience, or optimise for a more frequent event |
| Spend near zero | Payment failure, spending limit reached, or very low bids | Check billing and account spending limit; remove cost caps you don't need |
| Account restricted | Policy or verification issue | Resolve in Account Quality; don't open a new account to dodge it |
| Spend normal, results zero | Tracking or landing page issue | Go back to the tracking checks above |
Did a recent change reset the learning phase?
Edits you made in the last two weeks are prime suspects. Significant changes to budget, audience, optimisation event or creative can send an ad set back into learning, where costs are typically less stable.
Ask yourself:
- Did someone raise or cut the budget sharply (as a rough guide, more than 20–30% at once)?
- Did you change the optimisation event, for example from conversations to leads?
- Did you duplicate a winning ad set and switch off the original?
- Did you pause the campaign for a week and restart it?
If yes, the fix is usually patience plus smaller changes in future: raise spend in modest steps so you don't keep resetting learning.
Is your creative simply worn out?
Rising frequency with a falling click-through rate (CTR) is the classic sign of creative fatigue. The same people have seen your ad too many times and stopped reacting.
Pull a breakdown by ad for the last 30 days and check:
- Frequency (average times each person saw the ad) has crept up week by week.
- CTR has fallen from its early level.
- Cost per result has risen while reach has flattened.
Fixes, in order of effort:
- Launch two or three new creatives built on the same winning idea (new hook, new first frame, new thumbnail).
- Test a different format. In the accounts we manage, Reels creatives often had the highest CTR; one wedding photographer's best Reel ad reached a 3.50% CTR at ₹23.16 per conversation.
- Test a new angle entirely: a different offer, objection or customer story.
A structured approach beats random new ads. Our simple Meta ads creative testing framework shows how to test one variable at a time.
Have you run out of people to reach?
Small audiences tire fast, especially local ones. If you target one city, one age band and three interests, you may have shown your ad to almost everyone who will ever respond.
Signs and fixes:
- Reach has plateaued while frequency climbs: broaden location radius, drop narrow interests, or test broad targeting.
- Your lookalike or custom audience is tiny: rebuild it from a larger, fresher source list.
- Retargeting costs have jumped: your pool of recent visitors has shrunk because prospecting slowed down. Feed the top of the funnel again.
Did the offer, price or landing page change?
Ads carry the message, but the offer and page do the converting. A small change outside Ads Manager can wreck results inside it.
- Did you raise prices, remove a discount or end a seasonal offer?
- Did the landing page change, slow down or break on mobile? Load it on your own phone over mobile data.
- Does the page still match the ad? An ad promising "free consultation" that lands on a generic homepage loses people.
- Is the form longer than it used to be, or did someone add a required field?
- Did a competitor start offering something clearly better?
Is the market just quieter right now?
Seasonality and auction competition move costs without anything being wrong with your account. Year-end shopping periods typically make the auction more expensive for everyone, and many service businesses have natural off-seasons.
Compare the same period last year if you have the data. If the drop matches a seasonal pattern, adjust targets for that period rather than rebuilding the account.
Are leads still arriving but going nowhere?
Sometimes the ads work and the sales process doesn't. If enquiries are steady but bookings fell, the problem is after the click.
- Response time: how long does a new enquiry wait for a reply? Messaging leads cool off quickly.
- Lead quality: are the questions in your forms filtering out the wrong people? Our guide to lead quality vs lead volume explains how to stop paying for junk.
- Ownership: does every lead have a named person responsible for following up?
What does "not working" look like in real numbers?
Context matters: a cost that's a disaster in one niche is healthy in another. These are real, anonymised results from Meta accounts we manage, to show the range.
| Business | Result |
|---|---|
| Clothing brand | 2,564 enquiries at ₹5.43 average; one ad hit a 5.40% CTR |
| Wedding photography (5 accounts) | 1,130 enquiries at ₹58.55 average |
| Immigration consultancy | 572 leads + 525 conversations on about $676 |
| Interiors business | 2 conversations at ₹423 each |
| Removals company | Roughly $13–27 per lead |
The interiors account is the honest one. Two conversations at that cost is a signal to stop and rethink offer, audience and creative from scratch, not to keep tweaking bids. When every check above passes and costs are still far off, the problem is usually the fundamentals of the offer, not a setting.
Frequently asked questions
How long should I wait before deciding an ad has stopped working?
Look at at least 7 days of data, and ideally enough spend for several results. Daily swings are normal, especially during learning. If costs have been clearly worse for a week or more with no recent edits, start the checklist.
Should I just restart the campaign from scratch?
Rarely as a first step. Duplicating and relaunching throws away learning and often costs more for a while. Fix the specific cause first; only rebuild if the structure itself was the problem, such as dozens of tiny ad sets.
Did iOS privacy changes break my Facebook ads?
They changed what Meta can measure, so reported results can look lower than reality, especially with a browser pixel only. Setting up the Conversions API and checking results against your CRM gives a truer picture. It doesn't mean the ads stopped reaching people.
Why do my ads get clicks but no leads?
That points to the landing page, form or offer, not the ad. Check mobile load speed, the match between ad and page, and how many form fields you ask for. A click-to-message campaign can also help if your audience prefers chatting to filling forms.
Is it worth increasing the budget when results drop?
Not until you know why they dropped. More budget into a tired creative, broken pixel or weak page just buys more of the same problem. Diagnose first, then scale what's working.
Want a second pair of eyes on your account?
If you've worked through the list and still can't find the cause, we can audit your Meta account, tracking and landing pages together and tell you plainly what's broken. Request an ad account audit, or read our complete performance marketing guide first.