The Creative Genome Teardown
Meta tells you which ad won. It can't tell you why.
I take 90 days of your Meta ad data, tag every creative across 9 strategy dimensions, and show you exactly which hooks, angles, and creators drive real first-party revenue. Read-only access. Report in 5 business days. No new dashboard, no SaaS seat.
Read-only access · 3 to 5 business days · If you learn nothing new, you don't pay
You keep paying to relearn the same lessons
You spend $20k or more a month on Meta. You churn through UGC creators and video batches. And when a creative wins, the honest answer to “why did it win?” is a shrug with a confidence interval. Ads Manager will tell you ad 0417 beat ad 0392. It will not tell you whether it won because of the creator, the hook in the first three seconds, the angle, or the caption-heavy edit.
So every new batch is a fresh guess, and when a winner fatigues you can't rebuild it, you can only reshoot and hope. Nobody budgets a line item called “relearning,” but on a $20k+ account it is routinely five figures a quarter. The “why” is not missing. It's sitting in your ad account right now, unlabeled, and you keep paying full CPM tuition to relearn it every quarter.
Every creative, decoded into 9 strategy dimensions
Every ad you run is a bundle of decisions, and any one of them can be the reason it won. Meta reports on none of them. The teardown tags every creative from your last 90 days across all nine, then rolls up spend, ROAS, CPA, CTR, 3-second hook rate, hold rate, and the full retention curve by each one.
That's the difference between “this ad won” and “founder-on-camera problem-callout hooks with a hard CTA are the best dollar in your account.”
- 01Creator on camera
- 02Hook type
- 03Opening line
- 04Copy angle
- 05Format
- 06Tone
- 07Product focus
- 08CTA style
- 09Visual style
Numbers your finance team can check
Attribution products die inside teams the day their numbers disagree with Ads Manager. This one is built to reconcile instead.
Reconciles with Ads Manager
The analysis uses your unified attribution setting, so Meta-reported figures match what your team already sees. No “our model says” hand-waving.
Meta ROAS and True ROAS, side by side
First-party tracking ties actual leads and sales back to the ad that drove them, including delayed conversions and phone-closed deals the pixel never sees.
Coverage on every view
Every single chart states what percentage of spend it is built on. No cherry-picked date ranges, no unexplained gaps, and unanalyzed spend is never ranked.
What lands in your inbox
A report, a Loom, and a dataset. Not “UGC works,” but “creator-on-camera problem-callout hooks ran 2.1x True ROAS on $38k of spend while studio product-demo openers ran 0.9x on $22k.”
- 01
The Teardown Report
Your top 5 winning creative patterns with the actual spend, ROAS, CPA, CTR, hook rate, and hold rate behind each.
- 02
The Loser List
The patterns quietly eating budget right now, named, with the dollar figure attached.
- 03
The Hook Teardown
Opening lines and 3-second hook rates ranked across the whole account, and where retention dies.
- 04
3 to 5 Creative Briefs
Written from your data, ready to hand to editors and creators for the next batch.
- 05
Loom Walkthrough
A 20 to 30 minute walkthrough of all of it, plus a live review call to argue with the data.
- 06
The Full Tagged Dataset
Every creative, all 9 dimension labels, all metrics. Yours to keep and slice however you want.
One call. Five days. Zero access risk.
Connect, read-only
A 20-minute kickoff call where you grant ads_read access. It physically cannot touch campaigns, budgets, or billing. If you close by phone or on delayed timelines, we also grab a CRM export so revenue lands on the right ads.
Analysis runs, you don't
Over 3 to 5 business days, 90 days of creatives and metrics get pulled, tagged across the 9 dimensions, and reviewed by me personally before any number is rolled up.
Review call, then revoke
We walk the findings, the loser list, and the briefs together. You finish with next-batch briefs your team can shoot next week, and you can revoke access the same day.
Total time cost on your side: about 45 minutes.
No testimonial wall. Here's what I'll show you instead.
This offer is new, so this page won't be padded with logos I don't have or quotes I could have written myself. You are not asked to trust me. You are asked to check my work.
Built by a media buyer
I run Meta ads on my own offers and built AdGenome because I was tired of relearning my own account every quarter. I do every teardown personally. You're buying the answer, not the software.
A sample teardown, before you pay
A full teardown from a real account (numbers intact, brand redacted) is available on the fit call, so you can judge the depth of the analysis before you spend a dollar.
Methodology in the open
The 9 dimensions are listed on this page. The numbers reconcile against your own Ads Manager. Every chart declares what percent of spend it is built on.
Who this is for (and who it isn't)
A fit
- Brands and agencies spending $20k+ a month on Meta
- Shipping new video or UGC creative at least monthly
- Roughly 30+ distinct creatives in the last 90 days
- Especially strong if you close by phone or on long cycles: that's exactly the revenue Meta's pixel misattributes
Not a fit
- Accounts under $10k a month
- Accounts running three evergreen ads
- Anyone shopping for a media buying agency: I analyze, your team executes
Not sure which side you're on? That's what the fit call is for.
Flat fee. Priced like a deliverable, not a subscription.
One flat fee, paid once. No subscription, no renewal call, nothing for your team to adopt.
Single Brand
$750one-time, flat
One brand, one ad account, under ~$20k/mo spend. The full teardown, scaled to a smaller dataset.
- All 9 dimensions tagged across 90 days
- Winning patterns + loser list with dollars attached
- Hook teardown (3-second and hold rates)
- 3 creative briefs for your next batch
- Full tagged dataset, yours to keep
- Loom walkthrough + review call
Standard Teardown
$1,500one-time, flat
One ad account, $20k to $150k/mo. The right pick for most brands reading this page.
- All 9 dimensions tagged across 90 days
- Top 5 winning patterns with spend + True ROAS behind each
- The loser list quietly eating budget, with dollars attached
- Hook teardown (3-second and hold rates, retention curves)
- 3 to 5 creative briefs for your next batch
- Full tagged dataset, yours to keep
- Loom walkthrough + kickoff and review calls
Agency Rollup
$2,500one-time, flat
Multiple brands or accounts in one comparative teardown. See which creative DNA travels across clients.
- Everything in Standard, per brand
- Cross-account pattern report
- Which hooks and creators travel between brands, and which don't
- Per-brand breakouts + briefs
- Full tagged datasets for every account
If the teardown doesn't surface something you'd act on, you don't pay. Same teardown, same guarantee, every tier.
For scale: at $20k/mo you spend about $667 a day on ads. The Standard teardown is roughly two days of spend, and one paused loser typically covers it within the month.
Intro pricing while the case-study bench gets built. No timer, no “3 spots left.” It just goes up later. Not sure which tier? Book the fit call.
The guarantee
Tell me something I don't know, or it's free
Read the report. If it does not surface at least one creative insight you did not already know and would genuinely act on, say so on the review call and you pay nothing. Not a discount, not a credit: nothing. You keep the report and the dataset either way.
Questions buyers ask before paying
Will the numbers match Ads Manager?
Yes. The analysis uses your unified attribution setting, so the Meta-reported figures in the teardown reconcile with what your team already sees. True ROAS is shown next to Meta ROAS, not instead of it, and every view states what percentage of spend is covered.
Is my ad account safe?
The connection uses ads_read, Meta's read-only scope. The tooling has no write permissions, so it cannot touch campaigns, budgets, or targeting even by accident. You grant access on the kickoff call and can revoke it the day the review call ends.
How accurate is the AI tagging?
Claude does the first pass, I review every label before anything is rolled up, and you receive the complete tagged dataset. Every label is auditable: if you find a bad one, you can see exactly which numbers it touched. And the guarantee backstops all of it.
We already run Triple Whale / Northbeam / Motion.
Keep them. Attribution tools tell you which ad drove revenue. Creative analytics tools show engagement patterns. Neither rolls true first-party revenue up by hook type, copy angle, creator, and format across your whole account. That join is the product, and it arrives as a finished report, not another seat.
We close by phone, weeks later. Pixel ROAS is fiction for us.
That's the exact case this was built for. True ROAS comes from your first-party tracked leads and sales matched back to the originating ad, including delayed and phone-closed deals. You see it next to Meta ROAS, so you know precisely where and by how much the pixel undercounts.
Couldn't my analyst do this in-house?
Do the math they would face: 150 to 400 creatives, nine labels each, plus metric rollups per label per dimension. That is a week of analyst time you almost certainly have not staffed, done here in 3 to 5 days with purpose-built tooling, and it arrives as briefs your creators can shoot from, not a raw spreadsheet.
What if it just confirms what we already know?
Then you don't pay. If the teardown doesn't surface at least one insight you didn't already know and would act on, say so on the review call and the fee is zero. You keep the report and dataset regardless.
Why is intro pricing this low? What's the catch?
I'm pricing for case studies, not margin. The tooling already exists because I built and run AdGenome on my own accounts, so my cost is analysis time. Early buyers trade a reference-worthy result for a fee that will look small later. That's the entire catch.
Your last 90 days already paid for these answers
Somewhere in the spend you have already burned is the exact profile of your next winner: the hook type, the angle, the creator, the format. You can keep re-buying that information one test batch at a time, or you can have it extracted, priced at about two days of spend, guaranteed useful or free.
Read-only access · 3 to 5 business days · You do not pay if it teaches you nothing