Creative is fatigued when the same ad’s click-through or engagement rate falls steadily as the same people keep seeing it. Compare the ad with its own first two weeks, not with the campaign average. Then refresh based on how many times your audience has seen it, not how many weeks it has run.
This applies on LinkedIn, Google, Microsoft Advertising, Reddit and Meta alike. The details of where to look differ by platform.
First, rule out the harmless causes
Not every drop is fatigue.
- Too little data. A CTR change on a few hundred impressions is noise. Wait for a few thousand impressions in each period you compare, more on LinkedIn, where feed CTRs often sit under 1%.
- The ad is still in its learning period. New ads and campaigns swing while the platform finds its audience. Judge from the first stable two weeks onward.
- The ad was always weak. If an ad started low and stayed low, it isn’t fatigued. It just didn’t work. Replace it, but don’t blame exposure.
- You changed something. A new audience, offer, landing page or bid strategy can shift CTR overnight. Check the change history before you call it fatigue.
Refresh by exposure, not the calendar
“Refresh every four weeks” sounds tidy, but it ignores audience size. A 5,000-person ABM list sees a creative many times in a month. A 300,000-person audience may not see it once.
A Kiin study of 71 LinkedIn advertisers measured how long it takes for an audience to see a creative once: a median of 14 days for audiences under 10,000 and 53 days for audiences over 100,000. The same study found engagement fell 9% once the whole audience had seen a creative once, and 22% after five exposures.
PipeHalo’s suggested cadence, built on that study:
| Audience size | Days until most of the audience has seen it once | Suggested refresh |
|---|---|---|
| Under 10,000 | 14 (measured) | Every 2 weeks |
| 10,000 to 30,000 | About 21 to 28 (estimated) | Every 3 weeks |
| 30,000 to 100,000 | About 28 to 42 (estimated) | Monthly |
| Over 100,000 | 53 (measured) | Every 45 days |
The first and last rows are Kiin’s measured medians; the middle rows are estimates between them. This is guidance, not a platform rule. Meta and Google publish no refresh cadence.
A simpler trigger works too: when impressions on a creative pass about three times your audience size, plan the next one.
What fatigue looks like, and what it doesn’t
Four problems look alike on a dashboard. Each needs a different fix, so tell them apart first.
| What you see | Most likely | Fix |
|---|---|---|
| One older ad falling; newer ads in the same campaign holding up | Creative wear-out | New creative |
| Every ad in the campaign falling together; most of the audience already reached | Audience saturation | Lower budget or widen audience |
| Account-level frequency far above any one campaign’s | Overlapping campaigns | Exclusions or consolidation |
| CTR dropped on the date you changed targeting or offer; frequency flat | Relevance change, not fatigue | Revert or test the change |
1. Creative wear-out
The classic case. The same ad, same people, too many times. The decline is gradual and tracks how much exposure has built up.
Where to look: compare CTR by ad over time inside one campaign. In LinkedIn Campaign Manager, that’s the campaign’s Ads tab. In Google Ads and Microsoft Advertising, use the Ads view with a date comparison. In Reddit Ads Manager and Meta Ads Manager, break the campaign down to the ad level.
2. Audience saturation
The audience is too small for the budget, so everyone has seen everything. All ads decline together, and new creative only buys a few weeks.
Where to look: compare audience size with reach. On LinkedIn, reach divided by audience size shows how much of the list you’ve covered. Kiin found campaign-level frequency of 3.5 against 9.5 at account level across its 71 advertisers, so check the whole account, not one campaign.
3. Overlapping campaigns
A retargeting campaign and a prospecting campaign hit the same people, or two campaigns share a list. Each campaign looks fine; the person sees both.
Where to look: put the targeting of every active campaign side by side. Look for shared lists or audiences with no exclusions.
4. A relevance change, not fatigue
CTR stepped down on a specific date, frequency didn’t move, and the date lines up with a change you made. That isn’t wear-out. Our guide to a dropping LinkedIn CTR covers how to separate them.
Platform notes
LinkedIn. LinkedIn’s Sponsored Content tips recommend 4 to 5 ads per campaign and replacing the lowest-engagement ad every 1 to 2 weeks. A practical rule is to keep 3 to 5 active ads per campaign so the platform always has a fresh option to rotate in. LinkedIn’s reporting has no frequency metric, so frequency is estimated from impressions and reach. More on that in our guide to LinkedIn ad frequency.
Google Ads. Frequency caps exist only in Display and Video campaigns, per Google’s frequency capping help page. Demand Gen doesn’t support them. On Search, fatigue is rarely about frequency; a falling CTR there usually points to ad relevance, competition or new search terms.
Microsoft Advertising. Treat it like Google Search. Check ad-level CTR over time and look at search terms before blaming the creative.
Reddit. The same test applies. Watch CTR by ad in Ads Manager and rotate creative when one ad declines while others hold.
Meta. Many readers run Meta too. Meta notes that when performance drops as frequency rises, the audience may be fatigued, and its frequency controls exist only for Reach and ThruPlay goals. The same diagnosis applies: compare each ad with its own start.
The fix, step by step
Start with the cheapest step you can undo.
- Add 2 to 3 new creatives with a different hook or format. Keep the winning ad live and pause the worst one. Don’t replace everything at once; you lose the benchmark.
- Change the hook, not just the color. A new headline, angle or opening frame gives people a reason to look again; a new background color usually doesn’t.
- If every ad is falling, fix the audience. Lower the daily budget, or widen the audience with adjacent roles, industries or regions.
- Exclude overlap. Exclude retargeting audiences from prospecting campaigns. Merge campaigns that target the same list.
- Rest small lists. For an ABM list you can’t widen, pause it for 2 to 4 weeks or switch to a different message theme.
- Plan the next refresh using the cadence table above, so you’re ahead of the next decline instead of reacting to it.
How to tell the fix worked
Check 14 days after the change. The new ads should reach the engagement rate the old ad had in its first two weeks, and the account’s overall CTR should return within about 10% of its baseline.
If new creative recovers quickly and then falls again within a few weeks, the problem is audience size or budget, not creative. Fix that before you make more ads.
How PipeHalo catches this
PipeHalo checks creative every morning. “Click-through rate dropping” runs on LinkedIn Ads, Google Ads, Microsoft Ads and Reddit Ads: it flags an ad whose CTR over the last 7 days is 30% or more below its own first 14 days. “Creative due for a refresh” runs on LinkedIn Ads and Reddit Ads and flags an ad early, once it has delivered 3 or more impressions per person its campaign reached, or has run past the cadence above. “Frequency fatigue” runs on LinkedIn Ads and flags a campaign only when frequency is high and CTR has fallen 20% or more. Search ads on Google and Microsoft don’t get refresh or frequency flags.
Each flag shows the money at stake from your own numbers and the steps to try in the platform. Once a flag is marked done, the actions log compares CTR in the 14 days before and after, so you can see whether the new creative worked. See all 15 daily checks or how it fits demand generation teams.
Related questions
How often should I refresh ad creative?
It depends on audience size. Small audiences need new creative every two weeks or so; audiences over 100,000 can run a month or longer. Exposure, not the calendar, is the better guide.
Is ad fatigue the same as high frequency?
No. Frequency is how often people see an ad. Fatigue is when their response falls as that number rises. High frequency with steady engagement isn’t a problem.
Does cost per click show fatigue?
Late, if at all. Kiin found cost per click moved about a month after engagement fell, so watch CTR and engagement rate first.