Ads tell you who is spending to grow. We collect them where we can do it cleanly, model spend as a range, and tell you exactly how much of the picture we cover. This module ships as Beta at launch. The label stays until the coverage earns GA.
What's in the Beta
How Spend Bands Work
Spend is modeled as reach × frequency × CPM. None of those three terms is known exactly, so the result is a band, not a figure. In the EU and Russia, reach comes from the official library. Frequency and CPM come from category and country assumptions that are part of the published methodology and its version number.
Outside those markets our collectors see the ads, their dates and the apps they point to, but not their reach. So there you get the ads, not a spend number. We'd rather leave the field empty than fill it with a guess dressed as data.
Coverage May Be Partial
Some sources have no official API. Our own collectors read only public pages, and those pages change, slow down or block automated visitors. When that happens we back off and switch the source off. We don't rotate identities or spread traffic across IP addresses to get around a limit. That means TikTok and Instagram coverage can be partial or uneven, and we say so on every result instead of hiding it.
Every collector follows the same rules:
- No login and no fake accounts.
- No CAPTCHA solving.
- An honest User-Agent that says who we are.
- A global request budget per source.
- A kill switch per source, plus a contract test that fails loudly when a page changes.
What We Show and What We Don't
You get derived figures: spend bands, scores and our own indexes, with a link or an official embed back to the original ad or video. We don't re-host video or ad creatives, raw library fields stay out of exports, and data from the YouTube Data API is refreshed or deleted within 30 days, as its terms require. Advertisers appear as companies. We don't collect personal contact details.
