@sevenfiguresmp

What they said

1. "This isn't my opinion." A study "from this year" analysed 578,000+ Meta ad creatives representing $1.29 billion in ad spend.
2. Only about 5% became true winners, so on average you need to test roughly 20 creatives to find one winner.
3. Typical agency "testing": two static before/afters and two B-roll videos = really two concepts, not four, yet called a creative testing strategy; "then you blame Andromeda (Meta's creative-diversity delivery system) for piss-poor results."
4. Reality: you are not testing nearly enough creatives to find winners.
5. "The lower your spend is, the lower the creative winner rate actually was in the study."
6. To find more winners: test more creatives, and they must be truly unique concepts, not slight variations of ads you already run.
7. Creatives do not need to be high production. They should be relevant: research what questions people actually ask about your service and answer those questions in the creatives.
8. Test in "super simple formats" so you can test 20 creatives and find a winner; once you find winning formats and concepts, then work on production value.
9. CTA: if it sounds like too much work, hit the link in bio and apply to see if you are a fit to work with him.
10. Numbers on the screenshot (table "Monthly Meta spend -> creative winner rate"): under $10K: 3.7-4.0%; $10K-$50K: 6.2-6.4%; $50K-$200K: 7.3-8.1%; $200K-$1M: 8.1-8.6%; $1M+: 8.2-8.8%. Visible text defines the winner: "Motion defines a winner objectively as a creative that received at least 10x the account's median creative spend and at least $500 in total spend. So this isn't someone surveying media buyers and asking what they consider successful; it's based on actual Meta delivery data. Even better for your 4-8% claim, [here's the] breakdown by advertiser size like this:". (So the study appears to be Motion's, a creative analytics company; the speaker never names it.)

Try this

• Meta advertisers: Plan for roughly 5% of creatives becoming winners, so budget to test about 20 truly different concepts per winner. The study covered 578,000+ ads and $1.29B in spend, with a winner defined as 10x the account's median creative spend and at least $500 spent.
• Small budgets: Winner rates rise with spend, from about 3.7 to 4.0% under $10K a month to about 8% above $200K. With a small budget, test fewer but more distinct concepts and move spend behind winners fast.
• Count concepts, not files: Two static before-and-afters and two B-roll clips are two concepts, not four. A concept is a different angle, pain point or offer, not a new colour or hook line.
• Choose concepts from research: List the questions customers actually ask and answer them in your ads. Test them in simple formats first, and invest in production only after a concept wins.
• Agencies and freelancers: Use this when a client asks why no winners appear. Report the number of unique concepts tested, not just the number of ads.

Links

• Motion (creative analytics) report/screenshot (name read from frame only).
• Meta Andromeda (referenced), Meta Ads Manager.
• Link in bio to apply to his agency (SMP = social media / "Seven Figures MP"; a growth agency).