How does the Competitor Ad Intel tool work?

Enter a competitor's website URL. The tool fetches their active ads from the Meta Ad Library API (Meta's public transparency database, which returns commercial ads for the EU, UK and Australia), captures screenshots of up to 8 of them, and runs vision analysis on each to extract creative style, hook type, funnel stage, production polish, and stealable elements. You see real ad data plus descriptive vision labels.

What is real data vs vision-derived labels?

Real data: ad count (the first page of up to 50 results, matched to the brand by page name), ad copy, ad images, run cadence, format mix, target locations and ages. All from the Meta Ad Library API. Vision-derived labels: creative style, hook type, hook technique, emotional tone, target audience, funnel stage, production polish, creative notes, stealable elements. Vision labels describe what is in the ad; they do NOT claim the ad performs well or works.

Does this tool estimate competitor ad spend?

No. The Meta Ad Library does not expose impression counts, click counts, conversions, or spend. The tool returns ad count, age distribution, format mix, and creative-pattern frequencies. It does not estimate, forecast, or guess spend, revenue, ROAS, or any performance figure.

How is this different from Meta Ad Library directly?

Meta Ad Library shows raw ads with no analysis. The tool adds vision-based creative analysis on a sample of the active ad set: hook type, creative style, funnel stage, stealable elements, and a structured playbook. Vision is the labeling layer; Meta is the data layer.

Can the tool analyze any brand?

Any brand whose Meta (Facebook/Instagram) ads reach the EU, UK or Australia, because those are the markets the Ad Library API returns commercial ads for. A brand advertising only elsewhere, such as the US, returns zero, and that is not evidence it does not advertise. Currently Meta only; Google, TikTok, and other platforms are out of scope.

What are stealable elements?

Stealable elements are descriptive observations of creative components visible in the analyzed ads: hook structures, overlay styles, CTA placements, format choices. They are suggestions for what another brand could test in their own A/B slate. They are not guaranteed performance predictions.

How fresh is the data?

Each scan is a real-time pull from Meta Ad Library; no caching across scans. Vision analysis runs at scan time on up to 8 fresh screenshots.

Your catalogue, open to AI shoppers.

See how

Decode a DTC competitor's Meta ad strategy.

Pulls a competitor's active ads from Meta's Ad Library (EU, UK and Australia) and breaks down each ad's hook, angle, call to action and funnel stage. Add screenshots or videos, or send an ad with the Cresva extension, to analyse the visuals too.

ParkerAttribution