We built the benchmark before the product. Post an ad, I'll score it here.

Hi, Oussama here, founder of SaliencyLab. We launch on 10 October.

The short version: upload an ad, get a scored verdict in about ninety seconds, benchmarked against 10,000+ real ads with their public outcomes attached. Five KPIs, one verdict (Scale, Sharpen or Rebuild), a predicted attention heatmap frame by frame, and then BuyerLens, which interviews synthetic buyer profiles to explain where the resistance sits.

What it is not: a survey panel, a ROAS predictor, or a black box. Scores are model predictions validated against public engagement and click-intent outcomes, and every report shows the sample size behind its percentile.

Between now and launch I'd rather show it than describe it. Drop a link to any public ad (TikTok, YouTube or Meta Ad Library) in this thread and I'll run it and reply with the verdict, the weakest KPI and what BuyerLens says about it. Your own ad or a competitor's, either works. First twenty, then I'll keep going as time allows.

One thing I'd like from you in return: what would you want in the benchmark that isn't there yet?