3
New API capabilities in the November announcement
Google Search Central
2026-02-21
Start date for branded classification data
Google Search Console API documentation
Google added branded-query segmentation, multimodal search reporting, and multi-country/device filtering to the Search Analytics API on November 16, 2026. The capabilities are available for automated reporting, but there is an important catch: branded-query requests return no rows for ineligible properties. Before an empty result becomes a zero in a stakeholder dashboard, marketers need to check what the property can actually report.
What the Search Analytics API update delivers
According to Google’s announcement, the update introduces three capabilities. The brandedClassification field lets reporting teams group or filter branded versus non-branded performance. The webMultimodal search type retrieves multimodal performance. And filterLists accepts multiple countries or devices.
For in-house marketers, brand segmentation offers a way to distinguish brand-related demand from discovery. Agencies can use the filtering capabilities to standardize reporting across selected markets. Freelancers can bring these views into automated dashboards rather than relying on manual reporting work. These are reporting opportunities, not promises of better search performance.
The release status matters. Google describes these API capabilities as added and available—not a test or a future announcement. However, the same announcement describes multimodal data in BigQuery exports as forthcoming. Do not treat API availability as confirmation that the same data is available in those exports.
Multimodal reporting itself is not new with this announcement. Its rollout in the Search Console interface began globally on September 24, 2026. November’s change brings that reporting into the API.
Empty brand results do not establish zero traffic
The eligibility limitation deserves attention before dashboard design. Google says branded-query reporting requires a top-level property with sufficient query and impression volume. URL-path and subdomain properties are excluded, according to its branded queries filter guidance.
For an ineligible property, a branded-query request returns no rows. That response is not evidence that the business received no branded traffic. It means the requested branded reporting is unavailable for that property.
An empty brand report is not the same thing as an audience that stopped searching for you.
Keep that distinction visible in the dashboard. If eligibility prevents reporting, label the view as unavailable rather than filling it with zeros. Otherwise, stakeholders could read a reporting limitation as a change in demand.
For agencies and freelancers managing several properties, check each property separately before standardizing the view. Do not assume a branded report that works for one client will work for another. A consistent dashboard should communicate differences in data availability, not conceal them behind identical-looking charts.
Google’s brand labels may differ from yours
Google uses an AI-assisted system to classify branded queries, rather than simply applying a keyword regular expression. It acknowledges possible misclassification and says the filter does not affect rankings.
That makes the classification method part of the reporting context. If your current brand report follows a manually maintained set of keyword rules, do not assume Google’s split will match it. A difference between the reports should prompt investigation before anyone presents it as a change in audience behavior.
Add a short methodology note explaining which classification the report uses. If you retain your existing brand keyword report, label it separately so stakeholders understand that the two views use different methods. The goal is not to declare one view interchangeable with the other, but to make the comparison understandable.
There is also a historical boundary: branded classification data starts on February 21, 2026, according to the API documentation. Mark that start date when planning historical comparisons. Do not present periods before that date as though the same classification data were available throughout.
Request rules need a reporting check
The new fields come with constraints that your developer or reporting provider should test before putting the results into regular reports.
First, filtering or grouping by brandedClassification cannot be combined with a query filter in the same request. Keep query-filtered requests separate from requests using the branded classification. This is particularly relevant when adapting a report built around existing brand keyword rules: adding the new field is not a substitute for reviewing the request.
Second, a request cannot filter the same dimension through both filters and filterLists. When using the new multi-country or multi-device capability, check whether that dimension already appears in the existing filters. Review the complete request rather than only the newly added field.
For multimodal performance, set type to webMultimodal. This retrieves performance from Google Lens, Circle to Search on Android, and image uploads in Google Search. Where relevant, give this data a distinct reporting view and compare it with the existing Search Console view before sharing it with stakeholders.
What to do this week
Start with a reporting review, not a broad dashboard rollout. Choose a property, confirm its eligibility, and ask your developer or reporting provider to test the new fields. Use that review to document both the available data and the request constraints.
A practical checklist for the handoff:
- Check the property: Confirm whether it meets the branded-reporting requirements before promising a branded/non-branded dashboard.
- Review the requests: Separate query filters from branded classification and check country/device dimensions for conflicting filters.
- Document the method: Explain Google’s AI-assisted classification and investigate differences from your existing brand rules.
- Validate the presentation: Label unavailable data explicitly and compare the multimodal view with Search Console before stakeholder use.
OnlyaClick tip: Make the unavailable-data label part of your dashboard template before adding the branded chart, so an eligibility limitation cannot silently become a zero.
Give the person reviewing the report enough context to interpret it without opening the API documentation. That means keeping eligibility, classification method, and the historical start date close to the relevant view. Clear labels are part of the reporting work, not an optional explanation after someone questions the numbers.
The bottom line
The Search Analytics API update gives marketers more flexibility in automated search reporting. Brand segmentation, multimodal performance, and broader filters can support clearer measurement—but only when the implementation respects eligibility and request rules.
This week, prioritize validation over presentation: check the property, test the fields, document the classification, and distinguish unavailable data from zero traffic. Use the new reporting capabilities as measurement tools, not evidence of performance gains, and keep forthcoming BigQuery functionality out of availability claims.
Frequently asked questions
Why does branded-query reporting return no rows?
Google says branded-query requests return no rows for ineligible properties. Eligibility requires a top-level property and sufficient query and impression volume; URL-path and subdomain properties are excluded. Check eligibility before interpreting an empty result, and label unavailable reporting explicitly rather than presenting it as zero branded traffic.
Can I combine brandedClassification with a query filter?
No. The API documentation says filtering or grouping by brandedClassification cannot be combined with a query filter in the same request. Keep those requests separate. If your existing brand report uses keyword rules, review its request structure and document that Google’s AI-assisted classification may produce a different split.
What does webMultimodal reporting include?
Setting type to webMultimodal retrieves performance from Google Lens, Circle to Search on Android, and image uploads in Google Search. The November announcement makes this reporting available through the API. Multimodal data in BigQuery exports is described as forthcoming, so API access does not establish export availability.
Does Google’s branded-query classification affect rankings?
No. Google says the branded queries filter does not affect rankings. Its classification uses an AI-assisted system, and Google acknowledges possible misclassification. Treat the labels as a reporting method, document them in dashboards, and investigate differences from your existing brand keyword rules before describing those differences as audience changes.
Sources
- New filtering and reporting capabilities in the Search Analytics API, Google Search Central (Nov 16, 2026)
- Search Analytics: query, Google Search Console API documentation (Nov 16, 2026)
- Introducing the branded queries filter in Search Console, Google Search Central (Mar 11, 2026)
- Announcing web multimodal Search performance reporting in Search Console, Google Search Central (Sep 24, 2026)
This article explains public news and data for general information. It isn’t legal, tax or financial advice.
