Your creative testing cadence is set by how fast frequency accumulates in your account, not by what fits the sprint calendar. Compute weekly frequency as (weekly spend / CPM) x 1000 / audience size, then refresh once a creative has absorbed the number of frequency points its placement tolerates: roughly 2 on TikTok, 3 on Meta feed, 2.5 on Shorts, 1.5 on app install channels. For most mid-sized accounts that lands on weekly for TikTok and app install and fortnightly for Meta feed, but two accounts with the same spend and different audience sizes belong on cadences a month apart.
Nearly every cadence argument we get pulled into is really an argument about the calendar. Someone wants the refresh on Monday because the sprint starts on Monday. Frequency does not care, and the account will tell you the answer in one line of arithmetic.
The arithmetic behind a cadence
Three inputs, all of which you already have in the ads manager.
Weekly frequency = (weekly spend / CPM) x 1000 / audience size.
Spend divided by CPM times a thousand gives you weekly impressions. Divide by the reachable audience for that ad set and you get average weekly frequency, which is the rate at which your creative is being consumed. If you want the definition and the edge cases around reach deduplication, our frequency glossary entry covers them.
Then divide the frequency points a creative can absorb by that weekly rate, and you have the creative's life in weeks. That is your refresh interval. Anything else is a preference.
The absorption allowances above are planning numbers from running these accounts, and they sit alongside rather than inside our benchmark data. Our 2026 ad fatigue benchmark, a synthesis of published platform and agency figures, puts the onset of decline at a weekly frequency of 2.5 on Meta prospecting and a frequency ceiling around 3.0 on TikTok. Onset is not the same as unusable. A Meta feed creative keeps earning through the first soft decline, which is why we plan on 3 points there, while TikTok punishes repetition harder than its measured ceiling suggests and we plan on 2. The benchmark also records the sharpest single data point in the argument: CTR drops 45 percent after a fourth exposure to the same creative, from Meta internal research.
Why five accounts get five different cadences
Same formula, five real shapes of account. Note that spend alone predicts nothing.
| Account shape | Weekly spend | CPM | Audience | Weekly frequency | Life at 3 points | Cadence |
|---|---|---|---|---|---|---|
| Broad prospecting, small budget | EUR 2,000 | EUR 12 | 400,000 | 0.42 | 7.2 weeks | monthly is fine |
| Mid-market DTC, tight interest stack | EUR 5,000 | EUR 12 | 250,000 | 1.67 | 1.8 weeks | fortnightly |
| Scaling, competitive auction | EUR 10,000 | EUR 18 | 300,000 | 1.85 | 1.6 weeks | fortnightly, drifting to weekly |
| Retargeting pool | EUR 20,000 | EUR 25 | 150,000 | 5.33 | 0.6 weeks | twice a week |
| National reach, cheap inventory | EUR 1,000 | EUR 9 | 2,000,000 | 0.06 | 54 weeks | frequency is not your constraint |
The first and last rows are the interesting ones, because they are the accounts where a weekly cadence is waste. If your creative will not accumulate 3 frequency points until November, refreshing weekly buys you nothing except production cost and a reset learning phase. Cadence there should be driven by how fast you can learn something, not by fatigue.
The retargeting row is the one that gets accounts in trouble. A pool of 150,000 people absorbing EUR 20,000 a week will chew through a creative in four days. Teams run the same three retargeting ads for a quarter and then wonder why the return on ad spend halved.
Cadence by placement
Placement changes the answer even at identical spend, because the tolerance for repetition and the amount of attention each impression carries are both different.
| Placement | Points a creative absorbs | Benchmark median CTR | Benchmark hook rate | Cadence at mid-size spend |
|---|---|---|---|---|
| TikTok in-feed | about 2 | 0.84 percent | 33 percent | weekly |
| Meta feed | about 3 | 1.62 percent | 28 percent | fortnightly |
| Meta Reels | about 3 | 1.78 percent | 28 percent | fortnightly, weekly while scaling |
| YouTube Shorts | about 2.5 | 0.42 percent in-stream | 22 percent in-stream | fortnightly |
| App install channels | about 1.5 | varies by network | varies by network | weekly, sometimes twice |
Vertical shifts it again. The benchmark measures days to a 40 percent CTR decline at 9 days for food and beverage, 12 to 14 for fashion, about 18 for beauty and DTC, about 21 for electronics and about 28 for B2B SaaS. A food brand on TikTok and a SaaS company on Meta feed do not belong on the same refresh schedule, and the gap between them is three weeks. Our ad fatigue calculator runs your own spend, CPM, audience and vertical through this and returns a date rather than a rule of thumb.
What each cadence demands from production
This is where cadence decisions actually fail. A cadence is a supply commitment, and the volume it implies is not intuitive. Take a rotation of 12 live variations, which sits inside the 8 to 20 the benchmark says a typical active campaign needs. Weekly replacement volume is the live set divided by creative life in weeks.
| Cadence | Creative life | New variations per week | Per month | Where teams stall |
|---|---|---|---|---|
| Twice weekly | half a week | 24 | about 104 | briefing, not editing |
| Weekly | 1 week | 12 | about 52 | week three, when the backlog runs dry |
| Fortnightly | 2 weeks | 6 | about 26 | holiday cover for one editor |
| Monthly | 4 weeks | 3 | about 13 | nothing stalls, the account just underperforms |
| Quarterly | 13 weeks | about 1 | about 4 | fatigue, permanently |
Read the monthly row against the benchmark's finding that brands shipping 15 to 50 variants a month see 3 to 5 times longer campaign lifespan than quarterly refreshers. A monthly cadence on a 12-ad rotation produces about 13 a month, which sits just below the bottom of that band. Fortnightly is the first cadence that clears it.
The failure mode is boringly predictable. A team commits to weekly, ships 12 variations in week one from a backlog built over a month, ships 9 in week two, and by week three is re-running last month's ads with a new thumbnail. Pick the fastest cadence you can supply every week for a quarter, not the one you can supply once.
What a monthly cadence costs at scale
Put a number on it. The benchmark puts week three at 45 to 70 percent below the launch baseline, and by week five a creative averages 38 percent below its peak. On a monthly refresh, weeks one and two run near baseline and weeks three and four run in the cliff. You are paying the same CPM for roughly half your monthly impressions while they earn something closer to half the click-through.
At EUR 40,000 a month, that is around EUR 20,000 of delivery against creative the account has already exhausted. Nothing in the platform reports this as a creative problem. It shows up as rising CPA and a bidding conversation, which is why so many accounts respond by lowering bids on tired ads instead of replacing them.
Making a chosen cadence suppliable
Supply is the whole game, and the benchmark's operational conclusion is that throughput rather than talent is the bottleneck: the number of unique concepts a team ships per month predicts campaign longevity better than the quality of any single ad.
Genyad is our product, and it exists for this specific problem: you upload the footage you already own once, it transcribes and tags every clip, and each variation is a fresh script, shot selection, voiceover, caption set and export drawn from that library rather than a re-cut of one timeline. One variation is one credit, and the Growth pack at EUR 99 carries 65 credits, which covers a fortnightly cadence on a 12-ad rotation for about two and a half months. Credits do not expire, which matters when your cadence is uneven.
Be clear about what it will not do for your cadence. There is no direct publishing into Meta or TikTok, so a weekly refresh still means somebody uploads and names ads every week: budget the trafficking time or the cadence slips there instead. There are no AI avatars or synthetic presenters, no static banner formats, no product-URL import, no product-feed or CSV template rendering and no predicted performance scores, so if your plan depends on scoring creative before launch rather than reading it after, this is not the tool for that step. The sequencing of brief, generate, traffic and read is laid out in our creative testing workflow guide.
Frequently asked questions
How do I calculate my own creative refresh interval?
Take weekly spend, divide by CPM, multiply by 1,000 for weekly impressions, then divide by the reachable audience of the ad set to get weekly frequency. Divide the frequency points the placement tolerates, about 2 on TikTok and 3 on Meta feed, by that weekly figure, and the result is your creative's life in weeks. Refresh on that interval rather than on a fixed day of the month.
Is weekly creative testing always better than fortnightly?
No. Weekly is waste in an account with a very large audience and modest spend, where a creative may take two months to accumulate 3 frequency points, and each refresh resets the learning phase for no fatigue benefit. Weekly earns its keep on TikTok, app install channels, retargeting pools and any account where the arithmetic puts creative life under about ten days.
What production volume does a weekly cadence really need?
For a rotation of 12 live variations with a one-week creative life, roughly 12 new variations a week, or about 52 a month. Most teams that commit to weekly are planning against their best week rather than their average one, and the shortfall appears in week three when the pre-built backlog is gone.
What happens if we refresh monthly instead?
Weeks one and two run near baseline and weeks three and four run inside the week-three cliff our benchmark puts at 45 to 70 percent below launch. Roughly half of the month's impressions are then bought at full CPM and delivered against exhausted creative, which surfaces as rising cost per acquisition rather than as an obvious creative problem.