Is It Time for New Creative? The Signs Your Ads Need a Refresh

“Should we make new ads?” usually gets asked at the worst possible moment: CPA has jumped for a week, someone panics, and the account gets a batch of rushed creative that replaces the one ad that was actually still working.

The real answer to how often to refresh ad creative isn’t a number of days. It’s a set of signals you read against each ad’s own history, a cadence that scales with your spend, and a pipeline that means a refresh is a routine handoff instead of an emergency.

This guide covers what creative fatigue actually is, the signals that point to it (and the ones that only look like it), how spend level changes the cadence, why Meta’s delivery system has made creative volume matter more, and how to introduce new variations without abandoning concepts that still sell.

How often should you refresh ad creative?

Refresh when the signals say so, not when the calendar does, and build enough creative supply that you can respond within days. Our recommendation for an ecommerce brand spending steadily is new ads entering the account every week or two, while individual winners run for as long as their numbers hold.

The calendar rules you’ll find online for how often to change Facebook ads (“every 7 days,” “every 30 days”) are averages of accounts that look nothing like yours. Fatigue is driven by how fast your audience sees the same ad, and that depends on spend, audience size, format, and how distinctive the creative is.

So split the question in two:

• How often should new creative enter the account? Continuously, at a rate that matches your spend. This is a production question, and our Meta ads creative testing framework covers how many new concepts to test and how to read them.
• When should a specific ad be retired? When its own performance has decayed and the decay is caused by the ad, not by something else. This is a diagnosis question.

Most accounts that struggle with ad fatigue have a production problem disguised as a diagnosis problem.

What is creative fatigue, really?

Creative fatigue is what happens when the people you can reach have seen the same ad too many times, so each extra impression produces less attention and fewer conversions. Meta’s own creative fatigue guidance describes it the same way: an audience that has seen the same creative too often. (For the general definition and the related metrics, AdBoost Health’s glossary covers creative fatigue and hook rate.)

It helps to separate three related problems, because they need different fixes:

• Creative fatigue: this specific ad has worn out. A new execution fixes it.
• Concept fatigue: the underlying angle (the promise, the problem, the story) has worn out across every ad that uses it. A new hook on the same angle won’t help for long; you need a new angle.
• Audience saturation: you’ve reached most of the people likely to buy at this price, regardless of creative. New ads help less than broader targeting, new products, or a new offer.

Which signals say your ads need a refresh?

The reliable signal is a combination: frequency climbing while engagement and conversion efficiency fall on the same ad, against the same audience, with nothing else changed. Any one metric alone is noise. The pattern together is fatigue.

Here’s what to watch at the ad level, compared with that ad’s own first few weeks rather than with an account average:

• Frequency rising steadily. Rising Meta ad frequency means more impressions per person, so the ad is cycling through the same people. Rising frequency alone isn’t a problem (retargeting runs high by design), but rising frequency plus the signals below is.
• Click-through rate decaying. A sustained downward slope in CTR on a stable audience is the earliest clean sign that the ad has stopped stopping people.
• Hook rate falling on video. If fewer people watch past the first few seconds than did at launch, the opening has worn out even if the rest of the ad is fine. That tells you which part to replace.
• Cost per result rising while CPM holds. If you’re paying the same to reach people but each result costs more, the ad is converting a smaller share of the people it reaches. If CPM rose too, check the auction before blaming the creative.
• Spend drifting away from the ad. When the delivery system starts quietly moving budget to other ads in the same ad set, it’s often reading the decay before your dashboard makes it obvious.
• Platform flags. Meta surfaces creative fatigue recommendations in Ads Manager when it sees this pattern. Treat the flag as a prompt to investigate, not an automatic kill order.

Give each signal time. A two-day dip is weather; a two-week trend is climate.

The same signals as a quick reference, with the first thing to rule out before you blame the ad:

Signal on one ad	What it suggests	What else could cause it	Check first
Frequency rising steadily	The ad is cycling through the same people	Retargeting audiences run high by design	Is this a prospecting or a retargeting ad set?
CTR sliding for two weeks or more	The ad has stopped stopping people	A placement mix change	CTR by placement, compared with the ad’s launch weeks
Hook rate falling on video	The opening has worn out	A new placement with shorter views	Hook rate on the same placement only
Cost per result up, CPM flat	Fewer of the people reached convert	Landing page, price or stock change	Conversion rate on the landing page since the drop
Cost per result up, CPM up too	Possibly not fatigue	Seasonal auction pressure	CPM trend across every ad in the account
Spend shifting to sibling ads	The delivery system is reading decay	A newer ad simply outperforming	Whether the sibling’s results actually improved
Every ad dropped on the same day	Almost never creative	Tracking, consent or checkout change	Store orders against platform-reported purchases

What looks like fatigue but isn’t?

Plenty of performance drops have nothing to do with creative, and refreshing ads won’t fix them. Rule these out before you retire anything:

• Tracking or attribution changes. A broken pixel, a consent change, a modified attribution window, or a checkout update can make a healthy ad look sick. If every ad dropped on the same day, the cause is almost never creative. Cross-check against your store’s actual orders, which is why blended ROAS sits alongside platform ROAS in any honest report.
• Landing page or offer problems. If CTR is steady but conversion rate fell, people still like the ad; something after the click broke. Look at page speed, stock levels, pricing, and shipping thresholds first.
• Seasonal auction pressure. When competitors flood the auction, CPMs rise across the board and cost per result follows. That’s an auction problem. Fresh creative can help you compete, but it won’t return you to off-season costs.
• Learning resets. Big edits can push an ad set back into learning. Meta notes that ad sets typically need around 50 optimization events in the week after a significant edit to stabilize, and performance is often volatile until then. Panic edits during a dip can create the instability they’re trying to fix.

A quick test: if one or two ads dropped while their siblings held steady, suspect fatigue. If everything moved together, suspect something upstream.

How does spend level change how often to refresh ad creative?

The more you spend relative to the size of your reachable audience, the faster frequency climbs and the faster creative wears out. Cadence should scale with spend, which is why the same calendar rule fails for small and large accounts alike.

Here’s an illustrative example with deliberately round numbers. Imagine two brands selling similar products to similar audiences. Brand A spends $5,000 a month on Meta; Brand B spends $50,000. Brand B buys roughly ten times the impressions against an audience that isn’t ten times bigger, so the same people see Brand B’s ads far more often. If Brand A’s best ad stays efficient for two months, Brand B’s equivalent might decay within a few weeks.

What this means in practice:

• Smaller accounts should refresh more slowly than the internet suggests. You need enough spend per ad to read results, and launching a constant stream of new ads splits your budget so thin that nothing gets a fair test. A few strong concepts at a time, iterated carefully, usually beats volume.
• Large accounts need creative supply as a standing operation, not a project. At high spend, the bottleneck is almost always production capacity, and it’s why we treat creative as its own line in the ad budget, not something squeezed out of media spend.

Format matters too. As a working rule, short-form video tends to wear out faster than statics, because the hook does more of the work and audiences recognize it instantly. Treat that as a tendency to check in your own data, not a fixed timeline.

Does Meta’s Andromeda update change how often you need new creative?

Yes, indirectly. Meta’s delivery system now leans more heavily on the creative itself to find buyers, which makes the volume and variety of genuinely different ads more valuable than it used to be.

Meta’s engineering team describes Andromeda as the retrieval system that narrows tens of millions of ad candidates down to a few thousand relevant ones for each person, and it’s built to handle the growing number of creatives that Advantage+ and generative AI tools produce. As broad targeting and Advantage+ automation take over audience selection, your ads increasingly act as the targeting: a hook about joint pain finds a different buyer than a hook about gifting, even inside the same campaign.

Two practical consequences for refresh planning:

• Variety beats volume. Ten ads that share one concept, one hook style, and one visual treatment give the system roughly one idea to work with. Five ads built on different angles give it five ways to find people. When you refresh, change something meaningful: the angle, the opening, the format, the person on screen.
• Cosmetic tweaks don’t reset fatigue. If the “new” ad is the old ad with a different background color or a swapped headline, the people who were tired of the first one are tired of this one. Real refreshes change what the viewer notices in the first second.

How do you introduce new creative without abandoning your winners?

Keep proven ads running while challengers earn their place, and iterate on what’s working before you replace it. A refresh should extend a winning concept’s life, not delete it.

Think of a winning ad as layers: the concept (the core promise or problem), the angle (who it’s for and why now), the hook (the first seconds), and the execution (format, creator, edit, visuals). When an ad fatigues, refresh from the outside in:

1. New hook, same body. Keep the tested middle and close, replace the opening three seconds or the headline. This is the cheapest refresh, so try it first: the opening is the part of the ad every viewer in the audience has already seen. The UGC creator brief template includes hook formulas by angle for exactly this step.
2. New execution, same concept. Turn a winning UGC video into a static, a carousel, or a founder-led version. Same idea, new visual signature.
3. New angle, same product truth. If “saves time” has run its course, test “replaces three products” or “built for small spaces,” as long as each claim is true.
4. New concept. Only when the outer layers stop recovering performance is the concept itself done.

A few operating rules make this work without wrecking a stable account:

• Don’t pause a winner at the first dip. Let challengers prove they beat it on the same metrics before you shift budget. Pausing early often leaves the account worse than a slightly tired winner did.
• Batch new ads instead of drip-feeding them. Adding ads can count as a significant edit that disturbs learning, so introducing a set of new ads together, on a schedule, is gentler than one-at-a-time changes every day.
• Give new ads a fair read. Decide in advance how much spend or how many conversions a challenger gets before you judge it, and write down what “winning” means (usually cost per acquisition or contribution margin, not CTR).

Do TikTok and other platforms fatigue faster?

Generally yes, and the platforms say so in their own tooling. TikTok’s Ads Manager recommendations advise running three to five unique creatives per ad group specifically to avoid ad fatigue, and its Smart Creative feature is designed to pause videos that show signs of fatigue within their first three to five days and rotate in replacements from a waiting list.

That design tells you how fast TikTok expects creative to wear out. Budget for more frequent, more native production than you’d need on Meta, and don’t repurpose polished Meta ads as-is.

Search behaves differently. On Google, people are responding to their own query, so text ads fatigue far more slowly. Visual placements like YouTube still need rotation, with less urgency than paid social.

How do you build a refresh system that runs without panic?

Once you know how often to refresh ad creative at your spend, turn refreshing into a weekly operating rhythm instead of a reaction to bad days. The goal is that when an ad fades, its replacement is already tested and waiting.

A simple version that works for most ecommerce brands:

• Weekly review. Check ad-level frequency, CTR, hook rate, and cost per result against each ad’s own baseline. Flag ads with a sustained compound decline.
• Standing production batch. Ship a set number of new ads every week or two, sized to your spend. Mix outer-layer iterations of current winners with a smaller share of genuinely new concepts.
• Written promotion and retirement rules, so decisions come from criteria rather than mood.
• Profit as the scoreboard. Judge refresh decisions on what they do to your target ROAS for ecommerce and margin, not on engagement metrics alone. A thumb-stopping ad that doesn’t sell isn’t a winner.

Some categories need this rhythm more than others. Fashion is the clearest case: seasonal drops, trend-driven visuals, and fast-moving feeds mean fashion brands burn through creative faster than most. Replenishment products with steady demand can lean longer on proven concepts.

If you’re evaluating outside help, ask how a partner handles this specific problem: how many new ads they’d ship per month at your spend, who produces them, and what their kill criteria are. It’s one of the sharpest questions in our guide to choosing a performance marketing agency.

The short version

You don’t need new creative because a week went by. You need it when an ad’s own numbers show a sustained, compound decline that nothing upstream explains, and you need it ready before that happens. Rule out tracking, offers, auctions, and learning resets, scale production to your spend, and refresh from the outside in so winning concepts get longer lives.

Do that, and “is it time for new creative?” stops being a crisis question. The answer is always the same: new creative is already on its way, and the data decides which ads make room for it.