Here's the most useful thing we can tell a new brand: publish 20 ads in your first two weeks. Not 20 good ads. 20 ads. That's the 20-Ad Head Start, and it's the single best predictor we've found of whether a brand will win on Meta. It has nothing to do with making 'better' ads.
Winners run more. Not better.
We looked at 250 brands and $30M of real Meta spend. The pattern wasn't subtle: the brands that win don't make better ads, they make more of them. Because here's the uncomfortable part, nobody can pick the winning ad in advance. Not you, not your agency, not Meta. So the winners stop guessing and start taking shots.
Targeting died. Creative is the only lever left.
Targeting used to be where you won or lost. Then Meta automated the whole thing with its Andromeda engine, which now sorts thousands of times more ads than before and decides who sees what. You can't out-target it anymore. You can only out-create it, more ads, more angles. Nielsen ran the numbers: creative drives 49% of sales lift, and 56% on digital. Creative isn't a lever. It's the lever.
It's a numbers game, and the numbers are brutal
Roughly 1 in 11 funded ads returns more than 10x. Most are fine. A few are outrageous. The only way to find the outrageous one is to ship the other ten. The top third of your ads will drive about 80% of the revenue, you just don't know which third yet.
Why 20, and why two weeks
Publish 20 or more unique ads and you're 39x more likely to land a winner. Publish fewer than five and you're sitting at about 2%. That's not a rounding error, that's the whole game. And speed compounds: the faster ads go live, the faster the data tells you what's working. Your first two weeks decide your first winner.
More ads, sure. But different ads.
Twenty versions of the same ad teaches Meta nothing. Twenty different arguments teaches it everything. Run three or more ad types and account ROAS climbs about 37%. So don't tweak the layout, change the angle. Testimonial, comparison, lo-fi, native. Ten arguments, not ten edits, and every one on-brand.
The proof
One brand shipped 202 ads in a single month. $12k of spend came back as $100k, an 8.4x blended return, with breakouts hitting 22-26x. Another ran 48 creatives in one sprint for a 10.3x return, carried by a couple of monsters at 13.6x and 17.6x. In both, the winners weren't the ads anyone would've bet on. They surfaced because the volume existed to find them.
How to get your head start
It used to take a shoot and a month. Now it's three steps:
- Make some ads. Brief Cuttable, get a batch of on-brand ads back, fast.
- Publish to Meta. Approve and ship straight from Cuttable, no download, no re-upload.
- Back the winners. Put your budget behind the top performers and let them run.
Start a campaign. Publish your first 20. Let the data pick your breakout instead of guessing at it. That's the head start, and Cuttable gets you there in minutes, not months.
