Meta ads that worked last month but not this month is one of the most common things I see when reviewing accounts. It is also one of the most misread.
The instinct is to blame the platform — algorithm changes, iOS tracking, increased competition. Sometimes those are real factors. More often the cause is in the account itself, and it falls into one of four patterns that are diagnosable from your own data before you change anything.
1. Creative fatigue
This is the most common cause by far, and the easiest to confirm.
Meta ads run to the same audience repeatedly. The first few times someone sees your ad, it is novel. By the fourth or fifth time, they have already processed it and decided it is not for them. CTR drops. Frequency climbs. And because Meta’s algorithm reads low engagement as a signal to deprioritise your ad in the auction, CPM often rises at the same time.
The result looks like your campaign broke. What actually happened is your creative ran out of runway with the audience it was shown to.
How to confirm it: check CTR trend alongside frequency. If frequency has climbed above 3 to 4 on cold traffic and CTR has declined over the same period, creative fatigue is almost certainly the cause.
The fix is not a new campaign — it is new creative. A different hook, a different format, a different visual treatment of the same offer. Leave the campaign structure in place and replace the ad. The algorithm has already done its learning work; give it fresh material to work with.
2. Audience saturation
The four causes above are not mutually exclusive. Creative fatigue often co-occurs with audience saturation. Learning phase disruption can mask a seasonal dip. External factors can make a creative fatigue problem look worse than it is.
This is why reading metrics in isolation tends to produce the wrong diagnosis. A CTR drop in isolation looks like a creative problem. A CTR drop alongside rising frequency and a shrinking reach number looks like audience saturation. A CTR drop alongside a landing page bounce rate spike is a funnel problem, not a creative one.
The diagnostic approach is to pull all four levers simultaneously — CTR, CPM, frequency, and the action funnel — and look for which one moved first and by how much. The one that moved first is usually the root cause. The others are often downstream effects.
The most common response to declining Meta ads performance is to do a lot at once: new creative, new audiences, new budget levels, a fresh campaign. This is understandable — something needs to change — but it almost always makes the situation harder to read, not easier.
If you change three things at once and performance improves, you do not know which change fixed it. If it drops further, you do not know which change made it worse. You have added variables instead of removing them.
The more useful response is to run the diagnostic first, form a specific hypothesis about the cause, make one targeted change, and give it enough time — usually a week — to show a meaningful signal before drawing conclusions.
Slower, but the only approach that builds actual understanding of your account rather than compounding the guesswork.
Sometimes you run through the four levers and the picture still isn’t obvious. Metrics are moving in ways that don’t point to a single cause. The account has too much history of changes to isolate variables cleanly. Or the campaign is technically healthy on paper but consistently underperforming against where you need it to be.
That is usually the point where a fresh set of eyes on the account produces more than any additional self-diagnosis.
This breakdown is written by Jason Gan, a Meta Certified Professional and Badged Meta Business Partner who has personally audited over 1,000 advertiser accounts since 2010. You can see real account breakdowns on the Jason Gan YouTube channel.
No login required. Just a screen share, your real numbers, and a plan you can act on the same day.
If the pattern is recurring and you want to build the ability to read it yourself going forward, 1-on-1 coaching works through your account week by week until that diagnosis becomes second nature.
Why do my Meta ads stop working after a few weeks?
Almost always one of four causes: creative fatigue (audience tuning out the same ad), audience saturation (reaching the same convertible people repeatedly), learning phase disruption (structural changes resetting the algorithm’s optimisation), or external factors (seasonal shifts, competitor activity, landing page changes). The platform is usually not what changed — something in or around your account is.
How long should Meta ads run before results drop off?
There is no fixed timeline — it depends on audience size, frequency, and budget. Watch CTR trend and frequency rather than time elapsed. When frequency climbs above 3 to 4 on cold traffic and CTR starts falling, that is the early warning signal regardless of how long the campaign has been running.
What is creative fatigue and how do I fix it?
Creative fatigue is when your audience has seen the same ad enough times to stop engaging. CTR falls, CPM often rises, and ROAS drops. The fix is new creative — a different hook or format — while leaving the campaign structure intact. The algorithm keeps its learning; the ad gets a fresh start.
Is it normal for Meta ads performance to fluctuate week to week?
Yes. Short-term fluctuation is normal. A single bad week is almost always noise. The signal worth acting on is a decline across two to three consecutive weeks across multiple metrics — not a dip in isolation.
Part of the Meta Ads Coaching resource library — explore all articles by topic.