Say you run a pest control company and your Facebook ads have brought in about a dozen leads a week since summer. Then two weeks ago the number slid to seven, and last week it was five. You open Ads Manager, see the cost per lead has nearly doubled, and the plan forms itself: duplicate the campaign, load some fresh ads, maybe try the new structure everyone on YouTube is talking about.
Hold that plan for one afternoon. When Facebook ads stop working, the cause is almost always something that changed, and changes leave fingerprints. Your tracking, your own edits, the season, a tired ad, or a broken page after the click. Each one can be checked in minutes, and they are cheapest to rule out in a set order. Rebuilding the campaign first skips all of them, wipes away the evidence that would have told you which one it was, and puts Meta back into its learning period on your money.
The loudest advice right now points the other way. A wave of videos says Meta's latest system update is killing results and the fix is a new campaign structure with far more ads. Some accounts genuinely do need new ads. But "the algorithm changed" is the one explanation you cannot check in your own account, which makes it the worst place to start. Start with the things you can see.
One Slow Day Does Not Mean Your Facebook Ads Broke.
The first question is whether anything dropped at all.
Meta's results are lumpy by nature. Its delivery system spends your budget where it predicts the best results across the whole day and the whole week, so one day can bring four leads and the next bring none, with nothing wrong in between. Meta's own help page on performance swings says it plainly: cost per result is an average, and you should avoid slicing it into narrow time windows when you judge it. A single bad Tuesday is a slice.
So compare whole weeks. In Ads Manager, open the date picker, set it to the last 7 days, and switch on Compare to line it up against the 7 days before. If you have the history, compare the last 14 days with the same 14 days a year ago too, since seasonal businesses swing on their own calendar.
Then be honest about your volume. At 5 to 15 leads a week, the difference between seven and five is two people. Two people can disappear for reasons that have nothing to do with your ads: a rainy weekend, a holiday, a local event that kept everyone busy. At that volume, give a drop two full weeks before you call it a trend. If you get 80 leads a week and it falls to 40, you can act sooner, because a swing that size is rarely chance.
If the weekly number really did fall and stayed down, keep going. If it bounced back, you just saved yourself a rebuild.
Meta's Lead Count Can Fall While Your Phone Keeps Ringing.
Before you judge the ads, check whether the counting changed.
Ads Manager does not see your business. It sees whatever your website and forms report back to it, and that reporting breaks for boring reasons. Pull your real numbers for the same two weeks: form submissions in your inbox or CRM, calls in your phone log, booked jobs, money in the bank. Then lay them next to what Meta says.
- If both fell together, the drop is real. Move to the next check.
- If Meta's count fell but your inbox did not, your ads are fine. Your tracking is broken, and fixing the ads would be fixing the wrong thing.
The usual causes are mundane. A website update replaced the page that held your Meta pixel. A form plugin changed how it confirms a submission, so the "thank you" event never fires. Someone moved the thank-you page to a new address and the conversion still points at the old one. Each of these makes Meta count fewer leads, and since Meta uses those counts to decide who sees your ad next, it also starts finding fewer of the right people. A tracking break turns into a real slowdown if you leave it long enough.
Two other changes move the number without any change in real customers. The first is your attribution setting, which decides how long after a click or a view Meta is allowed to take credit for a sale. Change it and the same sales get counted differently, which we walk through in That Attribution Setting Is Not a Report. It Is an Instruction. The second is current: Meta's new pixel setting that switches on November 12 can change how many sales get counted on sites where customers log in, and our piece on Meta's November 12 pixel change shows how to tell whether it touches your site. If your numbers jump or slide in the two weeks after that date, check this before you credit or blame the ads.
The gap between what Meta reports and what actually happened is a subject of its own, covered in Meta Says It Made You 40 Sales. Your Bank Says 12. For this check, all you need is one question answered: did real leads fall, or only the count?
The Last Person Who Changed the Account Was Probably You.
This is the check owners skip, and it finds the cause more often than any other.
Ads Manager keeps an activity history for every campaign, ad set and ad. It shows who made each change, what they changed and when, including changes made by automated rules someone set up and forgot. Open it for the two weeks before the drop and read it line by line. You are looking for anything that happened just before the numbers moved.
Some edits matter more than they look. Meta lists the changes it treats as significant, meaning they send the ad set back into its learning phase, where delivery is less stable and results usually cost more:
- Any change to targeting
- Any change to the ad itself
- Changing what the campaign is optimizing for
- Adding a new ad to the ad set
- Pausing the ad set for 7 days or longer
- Changing the bid strategy
Budget and bid amounts can count too, depending on size. Meta's own example: raising a budget from $100 to $101 will not reset anything, but going from $100 to $1,000 may. So the "small tweak" someone made to freshen things up may be the reason results wobbled, and making more changes to fix the wobble restarts the clock again. We wrote about that cost in The Algorithm Changed Again. You Probably Should Not Do Anything.
Then widen the search past Meta, because plenty of drops start inside the business. Did a promotion end? Did your price go up? Did you stop offering a service, or run out of the thing the ad promotes? Did the person who answers the phone change, or did the hours? None of these show up in Ads Manager, and all of them change how many ad clicks turn into customers.
Which is why the single most useful habit in an ad account is a change log. A shared note with three columns is enough: the date, what changed, and why. Include business changes, not only ad changes. When results move three weeks from now, you look at the log instead of guessing, and "we ended the fall discount on the 6th" gets found in thirty seconds instead of being blamed on the algorithm.
Facebook Ads Cost More in November Even When Nothing Is Wrong.
If your tracking is sound and nobody changed anything, look at what you paid to be seen.
Every Meta ad runs in an auction, and what you pay depends on how many other advertisers want the same people at the same moment. The number to watch is CPM, the cost to show your ad 1,000 times. Pull it for the last two weeks and the two weeks before.
Read it next to your click rate:
- CPM up, click rate steady: your ad is doing its job as well as ever. You are paying more for each showing because more advertisers joined the auction. Same ad, pricier room.
- CPM up, click rate down: this is more likely a tired ad, the next check.
Mid-November is the clearest example all year. Every retailer with a holiday sale joins the auction at once, and a local service business pays more for the same feed whether or not it sells anything anyone gives as a gift. We laid out the seasonal pattern in Meta Ads Get More Expensive in Q4. Plan For It Now. The same thing happens on a smaller scale when a competitor in your town starts spending heavily.
There is nothing to fix in the ad when this is the cause. Your choices are business decisions: accept a higher cost per lead for a few weeks, put the budget behind the services that still pay at that cost, or pull back until January if the math stops working. Rebuilding the campaign will not make the auction any less crowded.
Meta Will Tell You When an Ad Is Worn Out.
Ads wear out. The same people see the same image or video again and again, they stop noticing it, and Meta charges more to keep showing it.
You do not have to guess at this. Meta flags it in the Delivery column of Ads Manager. When an ad's cost per result is higher than the ads you ran before it, but less than double, the status reads Creative limited. When it reaches twice as much or more, it reads Creative fatigue. Meta counts every recent showing of that image or video, including showings from your other campaigns on the same Page, so an ad can wear out faster than its own numbers suggest.
You can also see it in two numbers you already have. Frequency, the average number of times each person has seen the ad, climbs. Click rate slides while CPM stays about the same. People are seeing it and no longer clicking.
The fix Meta recommends is a new ad with an image or video that is clearly different from the original, not the same photo with a new headline. And Meta adds a detail most owners miss: keeping the original ad running alongside the new one may get you more results than switching it off. Let the delivery system move spend to the new ad as it earns it.
If you want to understand why good ads fade on a schedule you cannot see, that is the subject of Your Best Meta Ad Has a Shelf Life. You Just Can't See It Expiring. And before you pause anything, read Meta Already Turned Off Your Worst Ad., because the ad that looks worst in the report is often not the one losing you money.
A Broken Form Looks Exactly Like a Bad Ad.
If clicks held steady and leads still fell, the problem is after the click. The ad is doing its job and something on the other side is not.
This check takes ten minutes and needs nothing but your phone:
- Tap every live ad. Many of the people who see your ad are on a phone inside the Facebook or Instagram app, so test it there. Does the link open the page it should? An expired promotion page or a broken link sends paid clicks nowhere, and Ads Manager will report those clicks as normal.
- Fill in the form and submit it. Does it go through? Does the notification reach the person who follows up? Forms break quietly after website updates.
- Call the number the ad sends people to. Does it ring through, or does it reach a full voicemail box or a line that changed?
- Time the page. Count how long it takes to load on cellular data. A page that took two seconds in August and takes eight now will lose people before they read a word.
Then check what happens after the lead arrives. A lead that gets a call back the next morning often looks like a bad lead, when the real problem was the wait. If your leads held steady but booked jobs fell, the answer may be on your side of the phone. We covered how much speed matters in Meta Hands You a Lead Instantly. Most Businesses Call It Back the Next Day.
The way to tell these causes apart is to read two numbers together instead of one, which is the whole idea behind Four Numbers Decide Your Meta Ads. The Rest Just Explain Them. The short version for a drop:
Rebuild a Facebook Campaign and Meta Starts Learning Over. You Pay for It Again.
Once you have worked through all of this, you will usually have your answer. Most of the time it is one of the first five checks, and the fix is small: repair the tracking, undo an edit, ride out the season, add a new ad, fix the form.
Rebuilding is the expensive option, and it should be the last one. A new campaign and new ad sets start the learning phase from zero. Meta has to find out again which people respond to your ads, and you pay for that period with less stable delivery and usually a higher cost per lead. Worse, it erases the before-and-after you needed. If results improve after a rebuild you will not know why, and if they get worse you will have lost the version that used to work.
Rebuilding makes sense when every check above comes back clean and the drop has held for more than two weeks. Even then, the answer is rarely a new campaign structure. It is more often a new offer, because the old one stopped being interesting, or a new set of ads that make a different argument to a different kind of buyer. Change the ad people see before you change the campaign behind it.
As for the claim that a Meta update broke your ads: put it through the three-question test from our piece on algorithm news. Which of your campaigns does it touch? Which setting or report would show it? Which number in your account would move if it were true? An explanation that fails all three is not something you can act on. The checks above are.
Twenty Minutes in Ads Manager Each Monday Catches a Lead Drop Early.
You do not need to wait for a bad month to use this. Once a week, compare the last 7 days with the 7 before. Glance at the Delivery column for fatigue warnings. Tap your live ads on your phone. Add anything that changed, in the account or the business, to the log. That is about twenty minutes, and it catches most problems while they are still small.
And when the numbers do fall, work down the list in order before you touch the campaign. Is it real? Is it counted? What changed? Did the price of being seen go up? Is the ad tired? Is something broken after the click? The answer is almost always in there, and finding it costs an afternoon instead of a month of relearning.
This is the same check we run when a business brings us a Facebook account that has gone quiet, before we change a single setting. If you would rather have someone do it for you, that is what our Facebook ads management team does every day. If you would rather do it yourself, the list above is the whole method.
