The end of "find one winner and pour budget into it"
The old Meta ads playbook was simple: A/B test, pick the winner, put all the budget there. It was the most reliable way to lower CPM and stack conversions.
That playbook has flipped. Meta is now openly pushing the Creative Diversification frame in its official Newsroom. One sentence:
Don't hunt for one winning creative. Run 10–50 at the same time.
Source: Meta Business News — The Creative Advantage
Why it flipped — Andromeda is the backdrop
The reason is the new matching engine. Andromeda, rolled out in stages since late 2024, evaluates tens to hundreds of thousands of candidate ads at once. The more candidates, the smarter it gets. If an ad set has only 1–2 creatives, there's not enough material for Andromeda to work with.
In other words, "find the one ad that works" has become the wrong question. The new question is: can you supply enough diversity?
What teams using this strategy actually do
From Meta's featured case study — Dribbleup (D2C ecommerce):
"We used to produce 3–4 new creatives per week. Now we average 50. We scaled while holding performance flat." — Ben Paster, Head of Ecommerce, Dribbleup
Fifty a week may sound unrealistic, but the point isn't the number — it's the production system change. They brought it in-house and shortened the "produce → launch → feedback" loop. Fifty a week via agency would blow up costs.
So what should your team change?
Jumping to 50 a week isn't realistic for a small team. Step up gradually.
| Stage | Current output | Target |
|---|---|---|
| Week 1 | 2–3/week | 5/week |
| Weeks 2–4 | 5/week | 10/week |
| Week 5+ | 10/week | 15–20/week |
Three key questions:
- Can you automate variants? — text swaps, color/background changes, format conversions (horizontal → vertical). Don't redesign from scratch every time; templatize.
- Are you using AI tools? — just turning on Advantage+ Creative (Meta's built-in) auto-expands one asset into multiple variants with generated text and backgrounds. If it's off, you're leaving gains on the table.
- How long is your feedback loop? — "launch this week → analyze in 2 weeks → new creatives next month" is too slow. Look at launch-day data and adjust ideas the next day should be the baseline.
Decision criteria: should you flip to Creative Diversification?
If 2+ apply, switch now.
- Running 1–2 creatives per ad set
- The same creative has been your top performer for 2+ months (you're over-concentrating on a winner)
- You're not running Advantage+ Sales, or it's under 50% of spend
- Advantage+ Creative is not turned on
- Frequency is climbing past 5 with no refresh
When Creative Diversification doesn't fit
Not every account benefits. If any of these apply, content consistency matters more than diversification:
- Early-stage brand — identity is still forming, so scattered messaging hurts awareness
- High-consideration products (B2B SaaS, premium B2C) — complex value props get muddied when split across many versions
- Budget under $20/day — learning is already hard; diversification fragments performance
For these cases, "keep 3–5 creatives live, refresh monthly" is the realistic middle ground.
What the official framing really means
When Meta makes something "official," it's not a nice-to-have suggestion. It's declaring this is the axis that will separate winners from losers over the next 1–2 years. Accounts that ignore this won't benefit from the cascading updates — Andromeda, GEM, and whatever comes next.
The job changes too. "Bidding and targeting optimizer" → "creative supply and production director." Accounts that shift early will have lower CPA six months out.
Creative planning, copy, and AI-powered production at scale are covered in depth in Meta Ads Book 3.