Research methodology
Know what the data means before you use it.
ChatAdsWatch is designed as an evidence-aware research tool. Source records, inferred labels and modeled estimates remain visibly separate.
Observed source records
An observation records an advertisement found beside a captured prompt in the imported source sample. Upstream appearance counts are preserved when supplied. They are not paid impressions, clicks, reach or spend.
Classification
Industry describes the advertised product or service—not necessarily the advertiser’s entire business. Topic and buyer intent may be inferred from the creative and associated prompt. Low-confidence or conflicting records remain Other or Unknown.
Coverage and freshness
The catalog is a dated snapshot, not a live census of every ChatGPT ad. Missing geography, targeting and performance information stays unavailable. Pages show their latest source date so readers can judge freshness.
Corrections
Classification can be wrong. To report a data or classification issue, include the advertiser, creative title and a short explanation in an email to admin@conversaic.io. Corrections should preserve the original record and record the review decision separately.