A glass loupe magnifies a boot photograph beside printed branded and generic search phrases and country stamps on a reporting proof.

Search Analytics API Adds New Filters: Define the Report Before Moving Budget

Google’s November 16 Search Analytics API update gives marketing teams more ways to separate search performance in automated reporting. It also creates a management question: when the definition of a report changes, which apparent movements describe demand and which describe the new measurement?

The announcement brings brand classification, visual-search reporting and multi-value country and device filters into the API. The immediate opportunity is to improve an existing decision process. A new slice of search data does not, on its own, justify moving acquisition budget or declaring a campaign successful.

Give every new slice a decision owner

Begin with the question a report must answer. A brand-demand view might inform a discussion about awareness; a regional view might help prioritize localization; a visual-search view might reveal pages that deserve closer image and product-content review. These are different decisions, even when the same analyst produces the reports.

Write a short definition beside each view: purpose, property, date range, search type, markets, devices and responsible owner. State what action the data can support and what additional evidence that action needs. For example, search interest can prompt an investigation of a market, while a spending decision also needs commercial results and delivery capacity.

Keep missing classification separate from zero demand

Google says branded-query data is available only for eligible properties. An empty response therefore needs an availability check before it becomes a business conclusion. Do not label an unverified blank as evidence that nobody searched for the brand. Ask the analyst to distinguish unavailable classification, an unsuccessful extraction and a valid result.

Keep the previous brand-query definition available while reviewing the new one. If the team previously maintained its own vocabulary of brand names and variants, that was a particular measurement method. A different classifier may produce a different boundary. Explain the change in the report rather than silently presenting the two series as identical.

  Beauty Search Needs a Social-to-Google Handoff Map

Preserve the boundaries of regional and visual reporting

The new visual-search type covers Google Lens, Circle to Search on Android and image uploads in Google Search. Multi-value filters let a request include several countries or devices. Neither convenience removes the need to document what the resulting numbers include.

Keep a saved country and device list for each named region. A label such as “Europe” is insufficient if one reporting period includes a different set of markets. Likewise, display visual-search results with their own definition before combining them with another total. Ask whether categories overlap and what aggregation was used; adding two visible numbers is not a reconciliation method.

Approve a reporting change before approving a budget change

Run the existing and proposed views over a comparable period, record the differences and explain which definitions changed. Agree who accepts the new view and when it becomes the reporting baseline. Where a movement cannot yet be explained, retain that uncertainty in the commentary instead of converting it into a performance claim.

Google’s API documentation also warns that results are bounded and do not guarantee every data row. Treat the output as a defined reporting extract. A useful handover contains the saved specification, an availability note, comparison results and an action owner. That small discipline helps a better reporting interface lead to a better decision rather than merely a different chart.

Sources

Related reading

Alice Butler

Brandformance editorial contributor covering marketing strategy, digital media, SEO, analytics, ecommerce, martech, and marketing operations. Articles are prepared from cited public sources using an AI-assisted multilingual workflow with source, language, duplication, image, and rendered-page quality checks.