Zara Anita

The future of creative is quant |

[Zara Anita @anitazarapaul]

Each month's new Meta ads peak at 40-50% of a brand's spend. 8 weeks later they're at ~10%.

So close to half the budget moves onto brand-new ads every single month.

When you need 2x the creative, what breaks first: volume or quality?

[Zara Anita @anitazarapaul]

We tracked the top 10 Meta ads of a 7 figure/month brand week by week.

1 week later: 7 still in the top 10
1 month: 2
9+ weeks: 0

The median brand keeps 3 of 10 after a month. The ads carrying your budget today mostly won’t be in 2 months.

[Zara Anita @anitazarapaul]

By late October, Meta CPMs were ~20% above a normal week for the median brand we work with.
Last October they fell to ~0.6x.
Our guess: midterms. Political ads flood the same feeds. If we're right, CPMs should dip this week before the BFCM ramp.

[Zara Anita @anitazarapaul]

Brands started their Q4 creative push a month early this year.
October 2026: the median brand launched up to ~2x its normal weekly new Meta ads.
October 2025: launches dipped below normal. The ramp only came in November, peaking near 3x in Black Friday week.

[Zara Anita @anitazarapaul]

The #1 ad in a Meta account often sticks. The ads behind it don't.
A month later, still in the top 10:
#1 ad: 59%
#2-3: 38%
#4-10: 24%
About half of accounts have a whale that lasts months. Everything below it gets replaced every few weeks.

[Zara Anita @anitazarapaul]

We tracked 2,214 Meta ads past week 1.

Cheapest quarter in week 1: 0.56x the account's CPA. In days 8-30: 0.84x.

Priciest quarter: 1.37x, then 1.15x.

31% of the ads that ended up cheapest were in the bottom half in week 1.

[Zara Anita @anitazarapaul]

Don't judge ads one by one. Read your numbers at an ad set level. If your CPA is too high, it's a quality of creative problem.

[Zara Anita @anitazarapaul]

"That" ad with amazing CPA only looks amazing because Meta barely spent on it.

[Zara Anita @anitazarapaul]

So a small ad only looks cheap because Meta already bought the cheap conversions. If there were any left, Meta would have found it.

[Zara Anita @anitazarapaul]

Brands launching 500+ new Meta ads a month: 62% of those ads never get a single conversion in 30 days. Under 100 a month: 42%.

86% of ads that convert do it in week 1.

What changes inside a creative team between 100 ads a month and 500?

[Zara Anita @anitazarapaul]

Can you tell an ad will win in its first 3 days?

Sometimes. But 1 in 6 winners don't peak until after day 30, and early on they look only slightly above average.

At scale, those are the ones that get missed.

[Zara Anita @anitazarapaul]

How long does a healthy Meta ad live?

Half of all new ads stop getting real spend by day 7. Half of the winners are gone by day 27.

The more brands scale, the more creative they need. But most can't make new winners faster than the old ones die.

Why is it so hard?

[Zara Anita @anitazarapaul]

Our hypothesis, still to be tested: Trybe sits in one awareness stage for most brands' audiences.

Over-rely on it and your audience never moves through the rest of the stages.

[Zara Anita @anitazarapaul]

If Trybe isn't working for you, you might be using it wrong.

We compared creator ads with every other ad the same brand launched in the same month. The result flips once creator ads pass about 15% of launches. 🧵

[Zara Anita @anitazarapaul]

The more a brand spends, the more creative it needs. Handing all of it to one type of content doesn't work.