Data-driven creative prediction

Know if your next creative will work — before you spend on it

CreMensch learns from your live Meta campaigns what actually converts cheaply, and gives every new creative a reasoned fit score — before you upload it.

Get started — two minutes
0 50 100 56.1
Fit score from a real run: a video against a lookalike audience
The problem

The knowledge exists. It's just buried in tables

You spend thousands a month on Meta campaigns and produce new creatives on gut feeling. Some work, some burn budget — and you find out which is which only after a week or two of expensive data.

In one campaign we analyzed, ₪20,490 was spent across seven ads. The gap between the cheapest and the most expensive creative: ₪132 vs ₪390 per conversion — three times over, on the same audience. Nobody ever read the "why" out of Ads Manager.

How it works

Three steps, no complicated forms

1

The tool learns your live campaign

A deep read of every ad: true cost per conversion — cross-checked against your sales system, not just Meta's reporting — fatigue curves, who buys, on which placement, and which copy themes correlate with cheap conversions.

2

You send a new creative

A video or image plus the planned copy. The tool measures everything itself: format, aspect ratio, length, hook brightness, copy themes.

3

You get a reasoned 0–100 score

Every attribute is compared against what's proven on your specific audience. Every point is backed by a number from your data — not generic internet "best practices."

Real output

What an answer looks like

The run above is real — and the "fair" score it gave matches how that creative actually performed in the campaign.

Why "Mensch"

Not just a judge. A coach

Mensch (Yiddish): a decent, straight-talking person who tells you the truth and helps you improve. Alongside the score, every check returns three kinds of lines — real examples from a live campaign's video analysis:

Keep

What already works

"The satellite dive and the venue renders — the location theme converts at ₪183 on average and has no active ad. The footage is already shot."

Fix

The gap, plus the fix

"The first 3 seconds are dark with no on-screen text — hook retention is only 7.7% (your best opener holds 27%). Open on the bright shot and put the deadline in the first frame."

Test

Where data is thin

"The lineup section skews young-club, but your actual buyers are women 45–54. Worth an A/B: an experience-and-venue cut vs the lineup cut."

What's different

Three things Ads Manager won't give you

1

Your data, not someone else's benchmark. Every comparison is against what worked on your specific audience — warm audiences behave the opposite of cold ones, and the tool knows the difference.

2

Real conversions, not just the pixel. The tool cross-checks Meta's reporting against your sales or ticketing system — scores rest on money that actually came in, not an algorithm's attribution.

3

The tool sees the creative. Not just metadata: it measures the opener, length against real drop-off, and runs a frame-by-frame visual analysis that explains why an ad holds or loses viewers.

Honestly, like a real mensch:

This is a decision-support tool, not a prophecy. A gap of under 10 points between two options is a statistical tie, and accuracy grows with every campaign the tool learns. What it does promise: your next decision will rest on what actually happened in your account — instead of finding out two weeks and thousands of shekels later.

Want to know what your data says?

Joining takes two minutes: fill in your details, approve a view-only connection request — and within one business day you'll have your first campaign analysis.

Get started