Enter the team password
loading ad plates…
loading playbook…
loading competitor intel…
The score is a craft floor, not a prediction. It catches broken work (no clear product, unusable crop, an unreadable offer) before money is spent. Sale-event statics are judged on offer clarity — not punished for being sale creatives. Validated against ₹1.3L of linked Meta spend, it does NOT predict ROAS — the money signals on each card come from real linked ad results instead, and sharpen as the team records ad names. Bands: 75+ strong craft · 55–74 clears floor · 40–54 fix first · <40 reject. Scoring the same image twice shifts the number by about ±2 — 66 and 68 are the same score; exact re-uploads now return the existing verdict instead of a new number. Lifestyle, urgency, FOMO and problem-solving concepts are judged within their own lane — not as failed product shots.
Right now we score creatives here, run them on Meta, and the two never meet. That means nobody can tell you which scores actually made money. Two clicks fixes it.
Maniac_AN_ST_TrackPant_Static_Ad2_29/07/26.The name has to be exact. Copy it from Ads Manager rather than typing it — file names cannot be used instead, because three different creatives here already share one file name.
Catalog / Advantage+ shopping ads cannot be linked at all — their image comes from the Shopify feed, not from a file we scored. Skip those; image and video ads are the ones that matter here.
loading history…
Pre-flight predicts kill-risk, not ROAS — live gates remain the truth. Scores feed the monthly rubric recalibration.