POPULATION INTELLIGENCE · AD-PERFORMANCE PREDICTION

Pick your winning creative before you spend.

Upload 2–4 ad variants, set Meta-style targeting, and the same synthetic panel — drawn from a calibrated population model — makes a forced choice between them. You get a ship / tied / rework verdict with paired statistics, cohort flips, and the panel's own words on why. Catch the dud before the auction does.

Test an ad →Ads Labo via API

8B-PERSON POPULATION MODEL·CTR/CVR CALIBRATED TO WORDSTREAM 2024·93.3% PEW PARITY ON OPINION ITEMS

WHAT YOU GET BACK

A verdict you can act on, preference shares with honest error bars, cohort flips, and clustered creative direction — plus the verbatim reasons each respondent gave you.

THE VERDICT

Ship it, tied, or rework.

Run 2–4 variants and every panelist sees all of them (in randomized order) and picks the one they'd actually tap — or neither. The report leads with a one-line call backed by a paired sign test: because the same person judged every variant, persona noise cancels and ~100 panelists give you real decision power.

PREFERENCE + PULL

Shares with honest error bars.

Preference share per variant with 95% Wilson confidence intervals, plus a 0–100 Pull Index from each panelist's 0–10 tap-likelihood scores. These are indexes by design — the panel measures which creative wins and by how much, and never dresses that up as a real-world CTR.

COHORT FLIPS

Where the winner changes.

Preference broken out by age, gender, income, and interest — with the flips called out explicitly: 'B wins overall, but women 25–34 pick A at 61%.' If a flipped segment is your actual buying audience, that row matters more than the headline.

CREATIVE DIRECTION

What to fix, in their words.

Every panelist names the strongest thing the winning ad did and what sinks the weakest — clustered into 3 strengths and 3 weaknesses per variant, with the raw verbatim quotes underneath. When the verdict is 'rework', the why-neither reasons are the brief.

HOW IT WORKS

STEP 1

Upload your variants

2–4 images or videos, up to 200 MB each. A vision pass reads each creative once — category, tone, appeal, hook strength, brand timing — and every panelist reacts to all of them in randomized order.

STEP 2

Configure targeting

Geography, age range, gender, income tier, IAB Tech Lab interest taxonomy. Live debounced reach estimate as you build the audience.

STEP 3

Pick the platform

Meta Feed, Reels, Stories — the panel judges your creative in placement context. TikTok placement coming soon.

STEP 4

Read the verdict

Ship / tied / rework with a paired sign test, preference shares with 95% Wilson CIs, cohort flip callouts, clustered strengths and weaknesses, verbatim quotes.

WHAT TO USE ADS LABO FOR

Creative pre-flight

Test 2–4 ad variants on the same panel before committing media spend. Get a defensible winner with paired statistics and the reasons why.

Hook iteration

Iterate on first 3 seconds of video, opening lines, scroll-stopper imagery. Pit every cut against the incumbent and keep the one the panel picks.

Audience targeting validation

Compare preference across cohorts before launching the campaign. Catch mis-targeted creative before it hits the auction.

Geographic / language splits

Test the same variants across 23 market-level audiences. Catch when an ad reads great in the US but lands flat in Singapore.

Platform fit

Reels vs Stories vs Feed. Each placement carries different attention norms; the panel judges your creative in placement context.

Kill the dud early

When the panel picks 'neither', that's the cheapest bad news you'll ever buy — a rework signal before the auction charges you for it.

HONEST FRAMING

Built for the question a panel can actually answer.

A synthetic panel is genuinely good at one thing: comparing. Which variant wins, with whom, and why — that’s an ordinal judgment the same panelist can make across all your creatives, and it’s what Ads Labo reports, with paired statistics and confidence intervals attached. What it deliberately does notdo is impersonate a media plan: we surface real-world platform benchmark rates as separate reading context, never as our prediction, and we don’t output CPM or CPI — those are auction-driven and not predictable from creative alone. Use Ads Labo to walk into the auction with your best creative, then let a small live budget confirm the absolute numbers.

PANEL SIZES

Larger panels tighten confidence intervals and let cohort cells reach statistical significance.

100 RESPONDENTS · $0.015 EACH