The MMM Data Platform gives you access to a variety of data, including data for Google Query Volume (GQV) and YouTube reach and frequency. Although Meridian does not require GQV data, including GQV as a control variable can reduce bias in your model.
Note the following:
YouTube reach and frequency data: Reach and frequency data is available only for YouTube campaigns. It is pulled for Google Ads and de-duped across Display & Video 360 (DV360) YouTube data.
GQV data: The GQV report contains both brand and generic keywords. The brand terms included in the GQV report are based on the Knowledge Graph (KG) IDs associated with the brand names submitted in the data request. The generic terms include all terms from Google's product and services categories that are associated with the submitted brand names, but exclude the submitted brand names. Generic terms are generated automatically without customer input.
Query volume data is indexed. Raw query counts are scaled by a common randomly generated index value. (The same index value is used for all geographic regions within a country.) From a Meridian modeling perspective, it makes no difference that indexed values are used in place of raw values. Query volume is a control variable, and scaling a control variable has no influence on model fit.
The GQV data feed delivers indexed query volume relevant to the brands requested from MMM Data Platform, which is not specific to product types or ad types. AI overviews are generated for searches just like any other ad or search result, so the query volume does not change with regard to AI overviews. Query volume measures the number of searches done by users, not the number of overviews or ads that are generated for those searches.