Meta Ads
What to check first when yesterday’s winning campaign collapses overnight
Short answerIf your Facebook ads suddenly stopped working, don’t change anything yet. Check in this order: whether it’s really a collapse or one bad day, whether sales are still being tracked, whether anyone edited the campaign, budget, page or price, where the spend actually went by placement, and whether a setting quietly switched itself on. Only if all of that is clean should you look at frequency and the ads themselves.
Yesterday it was the best campaign in the account. This morning the CPA has tripled and the sales column looks like a typo.
The instinct is to do something, fast. Kill the ad, raise the budget, lower the budget, duplicate the whole thing, launch five new creatives before lunch.
Please don’t. Most of those moves are edits, and edits make the next 48 hours harder to read. Facebook ads that suddenly stopped working usually have a boring cause, and you find it by checking things in the right order.
First: is it actually a collapse?
One bad day is not a collapse. Delivery bounces around. One of the things students and clients ask me most is why their ads yo-yo so much, a $20 CPA some days and $100 on others. Under two full days of data, I don’t judge anything.
Look at blended CPA across the whole account (total spend divided by total purchases) for the last seven days and the last 30, not just yesterday. A dip is a few bad days inside a healthy trend. A real collapse is when the trend itself breaks. If the week still looks fine, the most useful thing you can do is nothing, and check again tomorrow.
Second: are sales still being counted?
This is the dumb one, and it’s worth two minutes.
Open your payment processor and compare real sales today against yesterday. If real sales held and Ads Manager dropped, your Facebook ads haven’t suddenly stopped working. You have a tracking problem. A purchase event that stopped firing, a thank-you page that moved, a checkout that updated overnight and broke on mobile.
Then buy your own product on your phone, start to finish. If something is broken, you’ll know in five minutes, and you’ll have saved yourself a week of fixing ads that were never the issue.
Third: did anyone change anything?
Sudden drops very often follow a change, and sometimes the person who made it didn’t mention it.
On the ad side, check the history for the campaign and ad sets. Meta’s own help pages say significant edits, like changing the optimization event, audience or creative, or pausing an ad set, can restart the learning phase, and budget or bid changes can too depending on how big they are. If the budget went up yesterday, read why your ads stopped working the day after you raised the budget before you do anything else.
Also check whether someone turned off a few “losing” ads in the same campaign. Facebook sequences people through several ads before they buy, so pulling the supporting ones can drag down the one that gets the credit. I went through that in why your best performing ad died the week you turned off the losers.
Off the ad side, check the page and the offer. A new price, a new version of the sales page, a checkout swap. Any of those can change results with the ads untouched.
A student of mine raised the price on his offer (which can be fine) but kept the same ads running, with the same Post IDs. A Post ID carries a lot of data about who took action on it, and when results suddenly drop off, that data can go bad, because Facebook now sees an ad that isn’t getting results. He changed his ads enough to get new Post IDs, and his message to me was that sales went up on Friday and more than doubled over the weekend.
Fourth: where did the money actually go?
Break the last day or two down by placement and platform and look at where the spend landed.
Every so often the platform itself does something odd, and a big share of the budget suddenly lands in a placement that was barely spending before. Audience Network is the usual suspect for spend with junk engagement and no purchases. Tons of very cheap clicks, nobody buying.
If that’s what you see, treat it as a platform anomaly, not as proof your Facebook ads suddenly stopped working. The answer isn’t to kill the concept. It’s to block the placement that’s eating the money for that business and let the rest of the campaign keep working. I’d only exclude a placement when it’s clearly taking spend with nothing to show for it, not as a reflex.
Fifth: did a setting switch itself on?
Open the ads that dropped and look at the creative enhancements and the other automated options. I run with all of Meta’s AI enhancements off by default, including text improvements and things like optimize website destination, because I’ve found they’re a net loss in most of my accounts.
The annoying part is that some of them have a habit of turning themselves back on, especially on duplicated ads. A setting you switched off months ago can be quietly back on today, and that alone can change how your ad shows up in the feed. Check every ad, not just the first one.
Sixth: was it really overnight?
If everything above is clean, look at frequency on that campaign for the last 30 days, and whether reach has stopped growing.
I worked with a client who teaches a very small niche of musicians. When we launched one of his more recent offers it blew up, around a 5x return, and then about three weeks in it just came to a standstill. I was adding something like 30 to 40 new ads a week at that point, so I knew it wasn’t a creative problem. The campaign reporting showed frequency around 6. When I looked at the total market in the Facebook groups, it was maybe 30,000 people, and we were already reaching around 28,000 of them.
It doesn’t matter how good the ads are if there’s no fish in the lake.
From the dashboard that looks like your Facebook ads suddenly stopped working overnight. It isn’t. It’s a pocket that was drained over weeks, and the day it ran out is just the day you noticed. The fix is not more budget or new colors on the same idea. It’s new concepts aimed at a pocket you haven’t reached yet.
If you want the fuller routine for reading an account in order, I laid it out in how to diagnose a Facebook ad account in under fifteen minutes.
What to do this week
Next time a winner drops, write these six questions on a sticky note and answer them in order before you click anything: one bad day or a broken trend, tracking, recent changes, placement spend, settings, frequency. Make only one change once you’ve found the cause, then wait two full days to read the result.
Why did my Facebook campaign suddenly stop converting?
Common causes are broken tracking or checkout, a recent edit that reset learning, a change to price or the sales page, spend shifting to a low-quality placement, or an automated setting switching back on. If none of those apply, check frequency. A pocket of buyers can be exhausted gradually and still look like an overnight drop.
Should I turn off an ad when its performance suddenly drops?
Not right away. Under two full days of data is too early to judge, and turning ads off is itself a change that can make things harder to read. Find the cause first. I only kill an ad when its CPA is three to four times target and its engagement is weak too.
Does editing a Facebook ad reset the learning phase?
Meta says significant edits can. That includes changing the optimization event, audience or creative, pausing an ad set, and large budget or bid changes. That’s why making several quick edits during a bad day often makes performance harder to recover and harder to read.
P.S. Keep a simple change log for your account: date, what changed, who changed it. When something drops overnight, that log answers the third question in about ten seconds.
