Creative Intelligence is a new Datadrew capability: Drew grades every active ad creative in your Meta account and gives each one a verdict — Scale, Iterate, Kill, or Wait — with the reasoning shown. Each creative gets a Convert Score built from more than ad-level ROAS, because ROAS-sorted reports hide dying winners. Here's what it does and how it changes your creative week.
TL;DR
- Every active creative gets a Convert Score — one grade built from the signals that lead (CTR vs the ad's own peak, frequency, delivery acceptance, conversion economics), not the ones that lag.
- Every creative gets a verdict: Scale, Iterate, Kill, or Wait — with the reasoning behind it visible.
- In a live account, two ads still showing 3.6x and 5.3x ROAS — with improving CPA — were already fatiguing. A report sorted by ROAS would never flag them.
- It answers the two questions every creative review is about: which ads deserve the next dollar, and what should we make next.
Why your creative report is lying to you
Open Ads Manager, sort by ROAS, and the story looks simple: the top ads are winners, the bottom ads are losers. Except creative doesn't fail top-down — it fails from the inside. Click-through decays against the ad's own peak weeks before cost-per-purchase breaks, and by the time CPA shows the damage you've paid for the decline.
We saw this in a live account on Datadrew — an EU beauty & wellness brand spending ~€17,000/month on Meta across 15 creatives. The two ads Drew flagged as fatiguing were showing 3.6x and 5.3x ROAS with CPA better than baseline. Sorted by ROAS, they looked like the last ads you'd touch. Measured against their own 7-day CTR peaks, both were down more than 30% with frequency above 2.5 — the audience had seen them and stopped responding.
That's the gap Creative Intelligence closes: it evaluates each creative the way an experienced media buyer would, not the way a spreadsheet sorts.
What Creative Intelligence does
For every active creative, Drew shows spend, ROAS, and CPA — and then goes past them:
A Convert Score for every ad. A single 0–100 grade of whether a creative deserves its next dollar. It's built from the signals that lead, not the ones that lag: delivery acceptance, CTR measured against the ad's own peak, frequency, and conversion economics in context.
A verdict, not just a number. Every creative lands in one of four buckets:
- Scale — earning delivery, economics hold, resilient as spend grows. Give it more.
- Iterate — the concept is proven but the execution is tiring. Re-cut it; don't abandon it.
- Kill — weak on the signals that matter, with enough spend to be sure. Stop paying for it.
- Wait — not enough evidence yet. Forcing a verdict here is how good concepts get killed early.
Reasoning you can check. Each verdict shows why: the specific signals that triggered it. No black box — you can disagree with Drew, and you'll know exactly what you're disagreeing with.
How this changes your week
You catch fatigue while it's still cheap. CTR-vs-peak breaks before CPA does. Catching it there is the entire economic point.
Your creative brief writes itself. "Iterate" verdicts tell you which proven concepts need fresh executions — so the next production cycle extends winners instead of guessing at new territory.
You stop killing ads that just need time, and stop funding ads that had their chance. The Wait verdict protects challengers with too little evidence; Kill is reserved for ads that failed with enough spend to know.
Creative stops being a folder of ad files and becomes what it should have been all along: a portfolio of hypotheses, each with a current grade and a next action.
FAQ
What is a Convert Score?
A single 0–100 grade of whether a creative deserves its next dollar. It's built from leading signals — delivery acceptance, CTR measured against the ad's own peak, frequency, and conversion economics in context — rather than lagging ones like ad-level ROAS alone.
What do the Scale, Iterate, Kill, and Wait verdicts mean?
Scale: the ad is earning delivery and its economics hold as spend grows — give it more. Iterate: the concept is proven but the execution is tiring — re-cut it. Kill: weak on the signals that matter, with enough spend to be sure. Wait: not enough evidence yet for a verdict.
Why isn't sorting by ROAS enough?
Because creative fails from the inside: CTR decays against the ad's own peak weeks before CPA breaks. In a live account, two ads at 3.6x and 5.3x ROAS with improving CPA were already fatiguing — a ROAS-sorted report would never flag them.
What do I need to use Creative Intelligence?
A Datadrew account with your Meta ad account and Shopify store connected. Drew grades your active creatives as part of its daily read of the account, alongside the Daily Ads Brief, diagnosis, and budget recommendations.
See your account graded
Creative Intelligence is live in Datadrew now. Open Creative Intelligence — if your Meta account and Shopify store are connected, Drew grades your active creatives as part of its daily read of the account, alongside the Daily Ads Brief, diagnosis, and budget recommendations.
Prefer to see it on your own numbers with a human walking you through it? Book 15 minutes with Sumit.
If you're comparing tools first, see how Datadrew stacks up against Triple Whale and other alternatives, or read our guide to customer LTV — because a creative that acquires repeat customers deserves a different verdict than one that doesn't.