01.01
Company-led setup
Define the client's product, website, audience, category, and market question before selecting competitors.
Separate each market, keep the source records behind every finding, and move from competitor discovery to a client-ready research brief without inventing performance data.
Agencies can use ChatGPT ad intelligence to define a client's market, review candidate competitors, compare captured creatives and prompts, request fresh country-specific evidence, and retain the dated records behind recommendations. Client performance still has to be measured in the client's own ad account and analytics.
A useful agency workflow begins with the client's product and buyer, not one shared industry list.
01.01
Define the client's product, website, audience, category, and market question before selecting competitors.
01.02
Show why each candidate appears and pin only the companies the client team agrees are relevant.
01.03
Use the same research question across selected countries when regional differences matter to the brief.
A client should be able to see the observation behind a recommendation.
02.01
Group observed headlines, descriptions, images, offers, and destinations around the buyer situations they address.
02.02
Retain the tested questions observed with ads so the research does not collapse into an unsupported keyword list.
02.03
Include countries, observation dates, source records, and inference labels alongside conclusions.
The dashboard and plugin share the same evidence boundary while supporting different working styles.
03.01
Use Your Market for the persistent client view, Pro Library for broad research, and Data Requests for missing fresh evidence.
03.02
Bring selected evidence into a supported assistant to draft hypotheses, compare options, and evaluate context-hint candidates.
03.03
Judge delivery and business performance using the client's Ads Manager and conversion data after launch.
Agency research is based on sampled public ad observations, detected landing-page signals, and labelled analytical inference. ContextHint does not provide competitors' private campaign settings or performance and should not be presented as doing so in client reports.