The short version
- There's no fixed shelf life: ads tire faster when a big budget meets a small audience.
- Plan replacements ahead from your budget, CPM and audience size, and swap an ad when its results slide for a week or more, not after one bad day.
- As a starting point, small accounts refresh every three to four weeks, mid-size accounts every two and large accounts weekly.
- Mix three kinds of refresh: new concepts, iterations of winners and format changes.
- A short weekly routine keeps fresh ads ready before your winners wear out.
The short answer
How often to refresh Facebook ads has no single answer, because an ad has no fixed shelf life. It depends on how quickly your audience sees your ads — and that comes down to your budget, the size of your audience and what your impressions cost. A small account can run the same winner for months, a large one may need new ads every week.
A sensible starting point, to adjust with your own numbers:
- Smaller budgets: new ads every three to four weeks.
- Mid-size budgets: every two weeks.
- Large budgets: every week, from a steady pipeline.
The real goal isn't a calendar. It's having fresh ads ready before your winners wear out — before creative fatigue sets in.
Knowing when an ad is due
An ad is due for a refresh once the people most likely to buy have seen it several times and its results slide on the same ad for a week or more, not after one bad day. Rather than wait for that slide, plan replacements ahead from your budget, CPM and audience size. The free creative refresh calculator estimates how long each ad is likely to last and how many new ones you need each month.
What makes ads tire faster
Budget against audience size
This is the biggest driver. Your budget buys impressions, your audience is the pool they're spread across. Spend more against the same pool and each person sees your ads more often.
The free creative refresh calculator does this maths with your budget, CPM, audience size and frequency, and estimates how many new ads you need each week or month.
Audience type
Broad prospecting spreads impressions across millions of people. Retargeting concentrates them on a few thousand site visitors. Small pools tire fast, so retargeting ads usually need refreshing sooner than their budget suggests, and higher frequency there is normal. The same logic works in reverse: if a prospecting ad set on a narrow interest audience is tiring quickly, widening the audience can buy time alongside new creative.
Format and idea
Formats wear differently. A single static with one line is taken in at a glance, so later views add little. A longer video or a carousel has more to discover, but costs more to replace when it tires. Time-bound ads — a sale, a season, a launch — come with their own expiry date. Treat these as tendencies to check in your own account, not rules.
How similar your ads are
Ten ads that look alike behave like one ad to the viewer, and practitioners report that Meta's delivery tends to treat near-duplicates as largely the same, too. Refreshing with genuinely different ideas buys more time than refreshing with new colours. Our guide to creative diversity and Meta's Andromeda system explains why.
A refresh schedule by monthly budget
Use this as a starting point, then adjust with frequency, CTR and the calculator. The budget bands are rough.
| Monthly Meta spend | Refresh rhythm | New ads each time | Suggested mix |
|---|---|---|---|
| Under $1,031 | Every 3–4 weeks | 4–6 | Mostly new concepts until something wins |
| $1,031–$5,155 | Every 2 weeks | 6–10 | New concepts and iterations of winners in similar numbers, plus a format change or two |
| $5,155–$20,619 | Weekly | 10–20 | Iterations, new concepts and format changes |
| Over $20,619 | Weekly or more often | 20+ | A steady pipeline across all three |
For example, a brand spending about $3,093 a month might refresh every two weeks with eight new ads: three new concepts, three iterations of its current winners and two format changes, such as a 9:16 video of its best static.
A small account running one winner for months isn't doing anything wrong. Refresh when the signals say so, not because the calendar does — and remember that a refresh is also a test, so each new ad needs enough budget to show what it can do. Our guide on how many ad creatives to test covers that side.
Plan around sale seasons
Festive periods and big sale events — Diwali, Black Friday, end-of-season sales — change the maths. Auction prices tend to rise, and offer creative has a fixed life. Brief and approve sale ads a few weeks early (ad review can take longer in busy periods), keep your evergreen winners running alongside them, and plan a refresh for the day the sale ends, when offer ads suddenly look out of date.
Three kinds of refresh
| Refresh type | What changes | Use it when | Trade-off |
|---|---|---|---|
| New concept | The idea: a new angle, problem, promise or audience | You need your next winner, or everything is tiring | Most learning, lowest hit rate |
| Iteration | The execution: a new hook, headline, scene or first line on a winning idea | A winner is tiring but the idea still sells | Reliable, with diminishing returns |
| Format change | The format: static to video, 4:5 to 9:16, single image to carousel | A winner has only run in one format or placement | Reaches new placements, needs a new build |
A healthy refresh uses all three, in proportions that shift with budget (see the mix column in the schedule above). Small accounts lean on new concepts until something wins, bigger ones need a steady flow of iterations and format changes to stretch their winners. Either way, let results move the balance.
Take a winning ad showing a can of soda on a desk at 3pm. An iteration keeps the idea and changes the execution: the same can on a night-shift desk. A format change keeps the ad and changes the shape: a six-second 9:16 video of it. A new concept changes the idea itself: the same can at a movie night.
Make iterations visibly different — a new scene, a new opening line, a new person — not the same image with a different background colour.
Keeping a bench of ready ads
A schedule only works if the next round exists before it's due. Many teams keep a small bench: ads that are briefed, made and approved, waiting to launch. One round on the bench is enough for most accounts. Two rounds help before a sale season, when ad review can be slower.
Fill the bench from places that rarely run dry:
- Customer words. Reviews, support emails and comments are full of reasons to buy and objections to answer. Each one is a possible new concept.
- Your own winners. Every winning ad suggests two or three iterations: a new opening line, a new scene, a new person.
- Proven patterns. The hook bank, the ad templates and the ad examples by category give you layouts and openings to try on your own product.
- Your category. Ads that rivals keep running for weeks point at angles your market responds to. Competitor ad research finds them. Borrow the pattern, never the creative.
Label every bench ad with its refresh type and the idea behind it, so each round you launch keeps the mix from the schedule above. When making the files is the slow part, BulkAds' AI ad generator builds batches of up to 12 ads from one product photo, in feed and full-screen sizes, so a round of eight is one batch rather than a week of design work.
A weekly refresh routine
About 30–45 minutes, on the same day each week:
- Pull the last seven days by ad: spend, frequency, CTR, cost per purchase and purchases.
- Flag fatigue: ads whose frequency is rising while CTR falls against their own first weeks.
- Check the whole account, so a market-wide dip isn't mistaken for fatigue.
- Stop clear losers and note why.
- Name this week's winners and the angle behind each one.
- Brief the next round: iterations of tiring winners, one or two new concepts, and a format change for any winner that has only run in one shape.
- Launch new ads in a separate test ad set, not in your steady scaling ad set. Meta treats some changes to an ad set — its list includes creative changes and adding new ads — as significant edits that can send it back into the learning phase.
- Log what you launched and why, so next month you know what was tried.
If you let BulkAds' Autopilot run your tests, it takes over steps 4 and 5 for them. It stops ads that spend twice your target cost per purchase without a sale, marks an ad as a winner when it has three or more sales at or under target after 48 hours, and drafts four fresh variants of each winner. Your part of the routine becomes reviewing and approving.
Refresh mistakes to avoid
- Killing a winner because it's old. Age isn't fatigue. If cost per purchase is still on target, leave it running.
- Editing a live winner. Changing a live ad's image or text usually creates a new version of its post, so it can lose the likes and comments it has earned — and the edit can restart learning. Launch the refresh as a new ad instead.
- Replacing everything at once. You lose the ads that still work and learn nothing about what replaced them.
- Refreshing colours instead of ideas. Near-identical ads tire together.
- Judging retargeting by prospecting standards. High frequency is normal in small audiences, so watch cost per purchase instead.
- Throwing tired winners away. Pause them and keep their results. An old winner can sometimes come back after a few months' rest, once the audience has moved on. Relaunch it as a new test rather than assuming it will work again.
- Waiting for a crash. By the time results fall off a cliff, you're a week behind. Keep the pipeline full.
Questions people ask
How often should you refresh Facebook ads?
It depends on how quickly your audience sees them. As a starting point, small accounts refresh every three to four weeks, mid-size accounts every two weeks and large accounts weekly, then adjust using frequency, CTR and cost per purchase.
How far ahead should I make new ads?
One round ahead. When you launch this round's ads, the next round should already be briefed, so a tiring winner can be replaced the week it starts to slip.
Does every refresh need a brand-new concept?
No. Mix three kinds: new concepts to find your next winner, iterations that change the execution of a winning idea, and format changes such as a 9:16 video of your best static.
Should I turn off an old ad as soon as its replacement launches?
Not straight away. Keep the old one running while its cost per purchase is acceptable, then switch it off once the replacements are working.
Does adding new ads reset the learning phase?
It can. Meta counts some edits to an ad set, including creative changes, as significant, and they can send it back into learning. Many advertisers test new ads in a separate ad set to protect a stable one.
How many new ads do I need each month?
Enough to have a replacement ready before each winner tires, which depends on your budget, CPM and audience size. The free creative refresh calculator estimates it from those numbers.
Keep going
- Features
Autopilot
Stops losers, scales winners, inside your caps
- Features
Creative analytics
Which ad actually sold — and why
- Free tools
Creative refresh calculator
How fast your ads tire, and how many new ones you need each month
- Guides
How many ad creatives should you test? The number, and the maths.
A straight answer, then the simple sum that sets your real number, how many new ads to launch each week at your spend, and when to stop testing and scale.
- Guides
Meta Andromeda and creative diversity: what changed and what to do
What Meta's Andromeda retrieval system is, what it changes for your ads, a practical diversity checklist and how to tell whether your ads are different enough.
- Glossary
Creative fatigue
Creative fatigue is the decline in an ad's results after the same people have seen it too often. For a creative team, it's the reason winners need replacing on a schedule, with genuinely fresh ideas rather than recoloured copies.
Written by the BulkAds team. Published , last checked . Meta changes its tools often — if something here no longer matches Ads Manager, tell us at contact@bulkadsai.com.
BulkAds AI is an independent product and is not affiliated with, endorsed or sponsored by Meta Platforms, Inc. Facebook and Instagram are trademarks of Meta Platforms, Inc.