If you're asking how often you should refresh ad creative, the honest answer is: there's no fixed calendar date that works for every account. The right cadence depends on how fast your specific ads are losing performance, not on a universal rule like "every two weeks." That said, most accounts running paid social do need to watch for decay on a rolling basis and have new creative ready before performance drops, not after.
Why ad creative wears out
Ad fatigue is a function of exposure, not time. The same audience sees the same ad repeatedly, and at some point the brain stops registering it as new information. This shows up as rising frequency (the average number of times one person sees your ad), falling click-through rate, and eventually rising cost per acquisition even though nothing else about your offer or targeting changed.
This matters because the question "how often should you refresh ad creative" is really two questions: how often should you have something new ready, and how do you know when to actually swap it in. The first is about planning capacity. The second is about reading your own data, not guessing.
The signals that tell you it's time to refresh
Instead of refreshing on a timer, treat these as the triggers worth watching on each active ad or ad set:
- Frequency is climbing while cost per result climbs with it. A frequency number that keeps rising on a stable audience is one of the clearest early warnings.
- Click-through rate has dropped noticeably from its own early performance, not compared to another campaign.
- Cost per acquisition or cost per purchase has drifted up for several consecutive days without a change in budget, audience, or landing page.
- Watch time or hook rate on video ads has fallen compared to the first days it ran, suggesting the opening seconds no longer stop the scroll as well.
- You've been running the same hook or opening shot across most of your active ads for a while, so even a "different" ad feels repetitive to the algorithm and the viewer.
- A seasonal, promo, or offer change has made the current creative's claim or urgency outdated, regardless of how it's performing.
If you see two or more of these at once on a given ad, that's a stronger case for a refresh than any single metric moving slightly. One noisy day of lower CTR isn't fatigue; a multi-day downward trend across frequency and CPA together usually is.
A practical refresh cadence by spend level
Because fatigue is driven by exposure, accounts spending more (and therefore showing ads to the same pool of people faster) tend to burn through creative faster too. Use the ranges below as a starting point to test against your own data, not as a fixed rule.
| Spend level | Typical exposure speed | Starting refresh cadence | What to prepare |
|---|---|---|---|
| Small / early-stage account | Slow — audience takes longer to see the same ad repeatedly | New hook or ad every 2-4 weeks, or when signals above appear | 1-2 new concepts in the queue at a time |
| Growing account, steady daily spend | Moderate | New ad or hook variant every 1-2 weeks | 2-3 concepts rotating, plus one in production |
| High-spend or aggressive scaling account | Fast — frequency climbs within days | New creative weekly, sometimes more than once a week | 3-4+ concepts always in the pipeline, built ahead of need |
| Seasonal or promo-driven account | Spikes around key dates | Refresh tied to the calendar event, not a fixed interval | Offer-specific creative ready before the promo starts |
Notice the pattern: the higher the spend, the less useful a monthly refresh schedule becomes, and the more you need a standing pipeline rather than one-off production. This is also why teams that plan how many ad variations to test per week tend to catch fatigue earlier than teams that wait for a dashboard alert.
Refresh doesn't always mean rebuild from scratch
A full creative refresh is expensive in time and attention if you treat every refresh as starting over. In practice, there's a cheaper first move worth trying before a full rebuild:
- Swap the hook only. Keep the body of the ad (problem, benefit, proof, offer) and change the first 2-3 seconds — a new opening line, a different angle on the product, a new on-screen claim. This is the fastest way to test whether fatigue is about the opening or the whole concept.
- Change the proof or offer framing. If the hook still performs but conversion has slipped, the issue may be in how the benefit or offer is presented mid-ad, not the opener.
- Re-cut the pacing. Sometimes the same script and footage perform differently just by tightening cuts or changing the order of beats.
- Build a genuinely new concept. Reserve this for when hook swaps and re-cuts no longer move the metric, or when the product, offer, or season has changed enough that the old angle doesn't fit anymore.
This ladder matters because it keeps your refresh cadence sustainable. A small team that insists every refresh must be a brand-new concept from zero will burn out long before a high-spend account needs its next ad. Teams that iterate first and rebuild second can realistically keep up with a weekly or even faster cadence.
Turning refresh timing into a habit, not a panic
Most accounts that struggle with creative refresh aren't struggling with the concept — they're struggling with having nothing ready when the signals show up. The fix is a standing queue, not a scramble.
- Keep a running list of 2-4 untested hook or concept ideas at all times, independent of what's currently live.
- Review the fatigue signals (frequency, CTR trend, CPA trend) on a fixed day each week, even if nothing looks urgent yet.
- Decide in advance which ads are your current "best performer" so you know what to protect and what to retire first when something new goes live.
- Log what each refresh changed (hook only, new proof, full new concept) so you can tell later which type of refresh actually moved the metric.
If you want a structured version of this, a creative testing calendar turns refresh timing into a recurring process instead of a reaction, and a clear way to organize your test results makes it obvious which refresh actually worked rather than just felt like it worked. For the broader process this sits inside, a full creative testing process for video ads covers how refresh fits alongside new concept testing.
Where FrameNotion fits into a refresh routine
The biggest practical blocker to refreshing on signal rather than on a calendar is production time — if a new ad takes days to brief, shoot, and edit, you end up refreshing late no matter how clearly the data tells you to. FrameNotion is built around that specific gap: paste a product or website link and FrameNotion AI writes and renders a custom 30-second vertical ad — with hook, problem, benefit, proof, offer, and call to action — in about 10-20 minutes, with no templates involved.
That speed is what makes a weekly or even faster refresh cadence realistic for a small team. Because every finished ad can take 2-10 edit requests depending on plan, and Pro and Agency plans support A/B hook variants, you can run the "swap the hook first" step from this guide without rebuilding the whole ad — change the opening line or claim, get a new render, and keep the rest of the concept intact. Ads also come out in 4:5, 1:1, and 16:9 alongside the main 9:16 version, so one refresh covers Feeds, Stories, and Shorts placements at once. You can see what finished ads look like in the examples gallery or check how it works before planning your own refresh pipeline.
FrameNotion doesn't publish ads to ad platforms or report on performance — you'll still pull frequency, CTR, and CPA from your ad accounts to know when to trigger a refresh. What it removes is the production bottleneck once you know it's time.
A simple decision rule to apply this week
If you only take one thing from this: stop asking "how many days since we launched this ad" and start asking "has frequency and CPA moved against us for several days in a row." If yes, try a hook swap first. If the hook swap doesn't recover performance within a short test window, move to a new concept. Keep at least one new concept in the queue at all times so the next refresh is never a scramble.
Review pricing if you're estimating how many fresh ads per month your refresh cadence will actually require — a small account refreshing every 2-4 weeks has very different production needs than a high-spend account refreshing weekly.
Frequently asked questions
Is there a fixed number of days before ad creative always goes stale?+
No. Fatigue depends on exposure (how many times your audience sees the ad), not a fixed number of days. A high-spend account can burn through an ad in days, while a small account might run the same ad for several weeks before signals show decay. Use frequency, CTR, and CPA trends as your trigger instead of a calendar date.
Should I refresh all my ads at once or stagger them?+
Stagger them. Refreshing everything at once makes it hard to tell which new ad or hook is actually responsible for any performance change, and it risks a temporary dip across your whole account if the new batch underperforms. Replace one or two ads at a time and compare against what they replaced.
Does a new hook count as a refresh, or do I need a whole new ad?+
A new hook on an otherwise unchanged ad counts as a valid refresh and is often the fastest way to test whether fatigue is about the opening seconds specifically. Reserve a full new concept for when hook swaps stop moving the metric.
How many backup creatives should I keep ready at all times?+
A reasonable starting point is 2-4 untested concepts or hook variants in a standing queue, so you're never scrambling when fatigue signals appear. Adjust up if you're spending aggressively and burning through creative faster.
Can seasonal campaigns follow the same refresh signals as always-on campaigns?+
Partly. Watch the same fatigue signals, but also refresh proactively around key dates — a promo end date, a new season, or a price change — even if performance hasn't dropped yet, since the old creative's claim or offer may simply be out of date.