More creative does not lower your CPA. It delays the rise, which is a different and smaller claim: run ten ads instead of two and the same audience takes roughly five times as long to reach the exposure count where click-through collapses. Your CPA on day one is set by the offer, the audience and the landing page. Creative volume decides how long that number holds before fatigue drags it upward.
We sell creative volume, so this is the claim our product benefits from most and the one we have least right to overstate.
The mechanism: exposure spread, not performance lift
Frequency is impressions divided by reach. Fatigue is not a response to campaign frequency though, it is a response to seeing one particular ad repeatedly, so what matters is per-ad frequency: campaign frequency divided by the number of ads carrying delivery.
weeks to per-ad exposure threshold = threshold × number of ads / weekly campaign frequency
Our 2026 fatigue benchmark, a synthesis of published platform and agency figures rather than our own measurement, supplies both inputs. It cites Meta internal research showing CTR drops 45 percent after a fourth exposure to the same creative, and puts the start of decline on Meta prospecting at a weekly frequency of 2.5.
| Live ads | Per-ad weekly frequency | Weeks to a fourth exposure | State of the ad set at week four |
|---|---|---|---|
| 1 | 2.50 | 1.6 | past the drop, CTR down about 45 percent |
| 2 | 1.25 | 3.2 | into the week-three cliff, 45 to 70 percent below launch |
| 4 | 0.63 | 6.4 | mid-decay, roughly halfway to the threshold |
| 8 | 0.31 | 12.8 | still near launch CTR |
| 16 | 0.16 | 25.6 | still near launch CTR, if delivery were even |
Substitute your own frequency. At weekly frequency 4 the column compresses: two ads reach the fourth exposure in one week, eight ads in four weeks.
Read what that table does not say. Nothing in it lowers CPA. Every row starts from the same launch CTR and the same conversion rate. The extra creative only moves the date the decline starts, and the delay is roughly linear in the number of ads.
Where the effect plateaus
Delivery is never even. The table above assumes each ad carries an equal share, and it never does: the algorithm concentrates on two or three performers within days. Weekly impressions are budget / CPM × 1000. At a EUR 700 weekly budget and a EUR 12 CPM that is 58,333 impressions, about 5,000 each across 11 ads and only 1,944 each across 30. Past roughly 8 to 20 live variations, which is what a typical active campaign needs in rotation, the marginal ad's impressions come out of the ads that were working.
Fatigue attaches to the message, not the file. Nine cuts of the same argument, proof and opening beat are one creative to a viewer scrolling at speed. Spreading exposure across near-duplicates resets nothing, which is why a volume programme built on re-cuts of one winner stops delaying anything. Our creative fatigue definition covers how that shows up in the metrics before it shows up in CPA.
A small audience caps it. Reach is bounded by the audience, so the value of extra creative is proportional to the frequency you already run. At EUR 700 a week and a EUR 12 CPM against a 500,000-person audience, weekly frequency is 0.12 and nobody is near a fourth exposure, so more creative buys almost nothing. Against a 25,000-person audience the same spend runs at 2.33 and creative volume is the main thing between you and the week-three cliff. Below about 1.5 weekly frequency, fix something else. The ad fatigue calculator runs this against your own numbers.
What actually lowers CPA
CPA decomposes cleanly, which makes it easy to see which lever has range:
CPA = (CPM / 1000) / (CTR × landing page conversion rate)
Start from a Meta feed baseline: EUR 12 CPM, 1.62 percent video CTR (the benchmark's median for Meta feed) and a 2.5 percent landing page conversion rate. That is EUR 0.74 a click and EUR 29.63 a conversion. Now move one thing at a time.
| Change | CTR | LP conversion | CPM | CPA | Versus baseline |
|---|---|---|---|---|---|
| Baseline, week one | 1.62% | 2.5% | EUR 12 | EUR 29.63 | |
| Do nothing, week-three cliff at 45 percent down | 0.89% | 2.5% | EUR 12 | EUR 53.87 | +82% |
| Add six variations, hold CTR at launch | 1.62% | 2.5% | EUR 12 | EUR 29.63 | 0% |
| Fix the landing page, 2.5 to 3.5 percent | 1.62% | 3.5% | EUR 12 | EUR 21.16 | -29% |
| Tighten the audience, accept a higher CPM | 1.62% | 4.0% | EUR 15 | EUR 23.15 | -22% |
| Change the offer, lifting CTR and conversion | 2.00% | 4.0% | EUR 12 | EUR 15.00 | -49% |
Row three is the honest version of the volume claim. It does not go below the baseline, it holds the baseline. Against row two that is worth EUR 24 a conversion, which is not small, but it belongs in the plan as avoided loss rather than as an improvement.
The bottom three rows move the number itself. A landing page converting at 3.5 percent instead of 2.5 takes 29 percent off CPA and keeps taking it off every week. A better offer is the only lever here with a shot at halving CPA, and it lifts CTR and conversion together because the offer is usually what the hook announces. A tighter audience can lower CPA even while CPM rises, which is worth noticing if you have been optimising for cheap impressions.
None of those three is a creative volume problem. If CPA has been at twice target since launch, more variations will not fix it, and any tool implying otherwise, ours included, is selling you the wrong thing.
What the published figures actually support
Our benchmark report finds that CTR declines 15 to 20 percent in a creative's first two weeks and then accelerates, that week three sits 45 to 70 percent below launch, that by week five a creative averages 38 percent below its peak, and that most creative is effectively dead within three weeks. Days to a 40 percent CTR decline run from 9 for food and beverage to about 28 for B2B SaaS, with beauty and DTC around 18. It also reports that brands shipping 15 to 50 variants a month see 3 to 5 times longer campaign lifespan than quarterly refreshers.
Every one of those is a decay or lifespan finding. None is a finding that CPA falls below its launch level when you add creative. The 3 to 5 times figure describes how long a campaign stays viable, which is the delay this post is about, not a discount on acquisition cost. Anyone quoting it as a CPA reduction has changed the claim.
How to size a volume programme honestly
- Calculate weekly frequency:
budget / CPM × 1000 / audience size. Under 1.5, volume is not your constraint, so go and work on the offer or the landing page. - Above 2.5, count your live ads and use the exposure table to find the week your current set runs out.
- Set the refresh interval from your vertical's decay window. CPA moving is the lagging indicator, three weeks late.
- Count arguments, not files. Six variations making three distinct promises is a volume programme; six cuts of one promise is one ad in six wrappers.
- Re-check after two cycles. If CPA held flat while frequency climbed, volume is working. If it drifted up anyway, the pool is too similar or the offer is the real problem.
Genyad is our product, so weigh this accordingly. It turns footage you already own into finished video ad variations, aimed squarely at step 4: distinct scripts, shot selections, voiceovers and caption sets from one library rather than re-cuts of a single timeline. One variation is 1 credit, Growth is €99 for 65 credits at about €1.52 each, and credits never expire. There are no predicted performance scores, so nothing forecasts your CPA before you spend. It does not publish to Meta or TikTok, so you export and upload. No AI avatars, no static banners, no product-URL or CSV feed import. And it will not lower a CPA that was never right.
Frequently asked questions
Does adding more ad variations lower CPA?
No, not directly. It delays the point at which fatigue pushes CPA up, because exposure spreads across more ads and each one accumulates frequency more slowly. Your CPA level is set by the offer, the audience and the landing page conversion rate; creative volume protects that level rather than improving it.
How much creative volume is enough?
Divide weekly impressions by the number of live ads and check each can still reach a readable share, roughly 5,000 impressions. At a EUR 700 weekly budget and a EUR 12 CPM you buy 58,333 impressions, which supports about 11 ads. Most active campaigns work best between 8 and 20 live variations, and past that the extra ads take delivery from the ones performing.
When is more creative not the answer?
When weekly frequency is below about 1.5, because nobody sees your ads often enough to tire of them. Also when CPA has been above target since launch rather than drifting up over weeks, which points at the offer or the landing page. And when every variation makes the same promise, since fatigue attaches to the message rather than the file.
If volume does not lower CPA, why ship 15 to 50 variants a month?
Because of what it prevents. Our fatigue benchmark puts week three at 45 to 70 percent below launch CTR, and at a 45 percent decline a EUR 29.63 CPA becomes EUR 53.87 with nothing else changed. Brands at that shipping rate see 3 to 5 times longer campaign lifespan than quarterly refreshers, which is worth paying for as avoided loss rather than as a discount.
Which single change moves CPA most?
The offer, usually by a wide margin, because it lifts click-through and landing page conversion at the same time. Moving CTR from 1.62 to 2.0 percent and conversion from 2.5 to 4.0 percent takes CPA from EUR 29.63 to EUR 15.00 in the decomposition above. Landing page work is the next largest and the most reliably available, and creative volume is defensive spending that comes after both.