VIFLY
← Back to the blog
SEO for Creators 17/11/2026 13-minute read 2 views

Google Search Console Opens New API Filters

Google opens to API Search Analytics brand filters, multimodal search and multi-country segments. Here's how to exploit them without overinterpreting the data.

Receive future VIFLY analyses more easily:
Creator observing a photograph and a graphic in a luminous workshop, with the title "Does your SEO see everything?".
Contents
  1. What Google officially added on 16 November 2026
  2. Brand or off-brand: measuring notoriety and discovery without fragile regex
  3. Multimodal research: see the discoveries that start with an image
  4. Several countries and devices in a single request
  5. Three concrete examples for creators and independents
  6. VIFLY: A better report does not replace a strategy
  7. What should you do now? A seven-step action plan
Want to turn this reading into practical action?

Keep this in mind: your content attracts attention, but your ecosystem turns that attention into clients, bookings and opportunities.

Google now automates three analyses so far available mainly in the Search Console interface: separating brand and non-brand searches, measuring traffic from visual searches and grouping several countries or devices into one API query. For a creator, consultant or small business, the change does not promise more traffic. It allows you to understand more quickly whether visibility is based on an audience that already knows the brand, new discoveries or visual uses like Google Lens.

Google announced this development on 16 November 2026. It concerns the Search Analytics API, i.e. the technical interface used to extract performance data from Google Search Console to a dashboard, script or reporting tool. The classic Search Console report remains usable without development. The API becomes interesting when you want to follow the same indicators regularly, cross several segments or avoid repeated manual exports.

Information verified on 17 November 2026. The functions are available worldwide in the API, but some data remain reserved for eligible properties. Seegoogle Search Central official announcement and the search Analytics query method documentation before adapting a connector or dashboard.

What Google officially added on 16 November 2026

The update does not create a new classification report and does not change the engine criteria. Google translates into the Search Analytics API three functions already introduced in the Search Console Performance report: brandedClassification, webMultimodal type and filter Lists field.

  • Brand and non-brand applications: the API can group or filter performance depending on whether Google classifies the search as brand-related or not.
  • Multimodal Web Search: the webMultimodal type brings together performances from Google Lens, Circle to Search on Android and images sent to Google Search.
  • Multiple filters:filterLists allows you to select multiple countries or devices in a single query, instead of multiplying separate calls.

These capabilities can be combined to answer more specific questions: Are brand searches progressing mainly on mobile in France and Belgium? Which pages are discovered by visual research? Do non-brand performance vary between several markets? Google explicitly presents this evolution as a means of automating reporting and workflows, not as a method of directly improving ranking.

This distinction is important. An API is a structured gateway to data or actions. Here, it allows authorized software to request Search Console statistics. It does not give access to internal ranking systems, does not reveal all Google's signals and does not guarantee the presence of each data line. The documentation recalls that the API mainly returns the most important results and remains subject to the internal limits of Search Console.

Brand or off-brand: measuring notoriety and discovery without fragile regex

A trademark request contains or designates the name of a company, site or product that is strongly associated with that mark. "VIFLY", a spelling variant of the name or name of a single service can thus be classified as a brand search. Instead, an out-of-brand request expresses a general need, such as "creating a bio link" or "consultant reservation tool".

So far, many dashboards separated these groups with regular expressions, also called regex: a list of words, variants and spelling errors that the system had to include or exclude. This method remains useful, but it requires constant maintenance and easily forgets a product, variant or contextual association. The new filter uses Google's classification and displays two values: BRANDED and NON BRANDED.

However, there are limits to this classification. She is assisted by AI and may be wrong. It is only available for first-level properties with a sufficient volume of queries and impressions. A property limited to a subfolder or subdomain may not be eligible. If the property does not meet the conditions, the API does not return lines for this dimension. In addition, brandedClassification cannot be combined with a filter on the query dimension in the same query.

The proper use is therefore not to blindly replace any internal segmentation. It consists of comparing the Google classification with the reality of the activity. If a generic training name is classified as a brand or a homonym blurs the results, this limit should be documented. The filter helps to observe a trend; it does not constitute an audit of full reputation or an absolute commercial truth.

Multimodal research: see the discoveries that start with an image

Multimodal research combines several forms of entry. Instead of entering text only, the user can photograph an object with Google Lens, surround an item displayed on his phone with Circle to Search or import an image into Google Search. The API now allows you to request the performance of these uses with type: webMultimodal.

This report is particularly interesting for visual activities: artists, photographers, decorators, artisans, fashion designers, cooking, tourism, real estate or creative equipment. A product photo, schema, place or detail of realization can become a door to a page even when the user does not know the right keyword or the name of the professional.

However, the data should be interpreted with caution. Multimodal searches often rely on images rather than textual expression; The size of requests is therefore not available in this type of report. You can follow pages, countries, devices, impressions and clicks, but not always know which exact phrase the user had in mind. For properties with an active group export to BigQuery, Google indicates that multimodal data can also join this environment.

Reporting does not exempt fundamentals. Relevant, original, accessible images, properly integrated into a useful page and accompanied by a clear context, remain more important than a sophisticated dashboard. The VIFLY guide on the SEO and the Google AI for creators recalls that visibility is first born of an understandable, credible and linked resource to a next step.

Several countries and devices in a single request

The filterLists field is used to collect several exact values of the same dimension. A dashboard can request data from France, Belgium and Switzerland, or from mobiles and computers in a single call. Countries use three-letter codes and devices accept DESKTOP, MOBILE and TABLET values.

The main interest is operational. An agency, a software publisher or a company operating in several markets was previously required to initiate multiple queries, collect responses and manage date or filter deviations. The new structure reduces this work and facilitates the creation of groups such as "Francophone markets" or "main screens".

There is, however, a technical rule: the same dimension cannot be filtered simultaneously with the classical filter field and with filterLists in the same query. The lists work by exact correspondence. They do not add an implicit geographical logic: if you want to form an Indian Ocean group, you must explicitly define the countries to be included and verify that the volume allows a relevant reading.

For small structures, the API does not automatically become necessary. If you visit a property once a month and your data is kept in a simple report, the Search Console interface may suffice. Automation brings value when monitoring is recurring, shared, multi-property or linked to other business indicators.

Three concrete examples for creators and independents

Example 1: a coach discovers that his traffic depends mainly on his name

A coach sees his clicks increase and concludes that his new articles become more visible. However, the filter reveals that most of the progression comes from searches containing its name after an intervention in a podcast. This is positive: its reputation is growing. But its content still attracts few people who seek a solution without knowing it.

It then separates two objectives. Presentation pages and testimonials respond to brand searches. Its guides need to better address off-brand issues encountered by its audience. It follows both curves without opposing notoriety and SEO. One VIFLY LinkHub then collect resources, offers and evidence when the person seeks his or her name, but it does not replace the editorial work necessary to be discovered on a general need.

Example 2: a visual designer appears thanks to Google Lens

A ceramicist publishes original photographs of her pieces and pages explaining techniques, dimensions, usages and deadlines. Its conventional web traffic seems stable, but the multimodal ratio shows a progression of impressions towards several visual records. She does not know the exact words used, because the search started with images.

She avoids renaming all her photos in an emergency. Instead, it checks the quality of the pages concerned: net photography, useful legend, descriptive alternative text, consistent product information, identifiable author and workshop, correct mobile loading. Article VIFLY on clickable sources in Google AI Mode develops the same logic: a content must have a reason to be open, even when the engine already provides part of the response.

Example 3: a consultant compares La Réunion, la métropole et les apparatuses

A consultant offers remote accompaniments and local appointments. It brings together France and some French markets, then compares mobile and computer. The Panel notes that impressions increase on mobile, but that service pages generate few reservations from small screens.

The useful decision is not to produce more reports. He tests the mobile route, clarifies his performances and reduces the steps between the page and the niche. One page VIFLY Booking here becomes a logical continuity: the measure identifies a friction, then the booking route must solve it. The new indicator has value only if it improves a real decision.

The viewpoint VIFLY : a better report does not replace a strategy

The VIFLY perspective is simple: sEO data becomes useful when it allows to distinguish visibility, recognition and action. Brand clicks indicate that people are already looking for you. Off-brand clicks show you're discovered on a problem. Multimodal research reveals another gateway. None of these indicators alone proves that the user understands the offer, trusts or acts.

The VIFLY channel therefore remains « visibility → understanding → confidence → action → opportunity ». Search Console mainly measures the start of the course. The site, content, evidence, offer and reservation determine the suite. Article VIFLY on google's IA responses and visibility already explained that a stable position can coexist with an evolution of clicks. The new filters make this reading finer, but they do not turn a rented audience into a direct relationship.

This update also happens during November 2026 Core Update. It does not automatically explain a variation caused by the ranking update. Article VIFLY on the action plan during the core update recommends waiting for stabilization and comparing consistent periods. Adding new segments during a volatile period can help document, not attribute a certain cause.

Good automation should not produce a more impressive report. It must trigger a question or action: which off-label page deserves improvement? Which visual resource attracts a new audience? Which mobile version blocks the reservation? Which market really justifies an editorial investment? If no one reads the report or no decisions are made, the API does not create any value.

Nuances, limitations and points to watch

  • This is not a ranking update. Google presents new possibilities for filtering and automation. Using them does not give any SEO bonuses.
  • Brand data is not available anywhere. The property must be first level and have sufficient volume. An empty answer may reflect ineligibility, not a total absence of brand searches.
  • Classification can be imperfect. The AI-assisted system may confuse a brand, generic term, product or homonym.
  • Multimodal search does not necessarily provide the text query. An image can be the main entrance; pages and trends should be analyzed rather than looking for a non-existent keyword.
  • The API does not guarantee all lines. Search Analytics focuses on key data and applies internal limitations. An export should not be presented as an exhaustive journal of each print.
  • The dates are interpreted in the Pacific Zone. The API documentation uses Pacific Time Search Console dates, which can create a lag with local monitoring in Reunion.
  • Automation also increases the risk of error. A bad country code, an incorrect property, a staggered date range or an incompatible filter can produce a clean but misleading table.

What should you do now? A seven-step action plan

  1. Check the need first. Determine what decision you want to make before developing a dashboard: notoriety, discovery, visual performance, country or devices.
  2. Control your Search Console property. For brand segmentation, use a first-level property and check that the filter appears or that the API returns data.
  3. Keep a reference report. Export a stable period before editing your scripts. Note the dates, property, search type and filters applied.
  4. Add one segment at a time. Start by brand against brand, then test separately webMultimodal and country or device lists. This method facilitates the detection of an error.
  5. Link each segment to a page. Identify content that earns or loses impressions, then check their intent, quality, mobile display and next step.
  6. Match SEO and business result. Follow requests, registrations, downloads, sales or reservations related to the strategic pages. An increase in non-action impressions can remain an untapped opportunity.
  7. Document the boundaries. Indicate in the dashboard the data not available, the non-eligible properties, the classifications likely to be wrong and the configuration changes.

The new Search Analytics API does not make SEO easier; it makes certain questions more visible. For creators and independents, the advantage will not come from the number of graphics, but from the ability to distinguish a person who already knows the brand, a person who discovers an expertise and a person who seeks from an image. Measuring these paths then makes it possible to build a clearer route, from the Google result to the offer or reservation.

Frequently asked questions

Do you need to know how to program to use these new filters?

Yes to directly query the API or modify a connector, but not to view functions already present in Search Console. A third-party tool may also incorporate new features, subject to its own update.

Is the brand filter available for all sites?

No. Google reserves it to first-level properties with a sufficient volume of queries and prints. Subproperties and small volumes may not receive any data.

Does Google explain how it classifies a query as a mark?

Google indicates using an IA-assisted system capable of recognizing names, variants, errors and certain associated products. Classification may include errors and may not change classification.

What does the multimodal web report contain?

It covers performances from Google Lens, Circle to Search on Android and images imported into Google Search. Text query size is not available for this type of search.

Can several countries be grouped into one API call?

Yes. The filterLists field accepts several exact country codes. It can also assemble several types of devices in the same query.

Can this update improve SEO?

Not directly. It improves measurement and automation. Decisions made from data can lead to better pages or a more efficient route, but the use of API is not a classification factor.

Do these data explain the core update of November?

Not by themselves. They can segment variations, but Google does not publish the signals modified by the core update. We must wait for stabilization and control the technical, seasonal and competitive causes.

Sources

Key takeaway

Good content should do more than get read. It should guide people towards a decision.

If this article helped clarify your strategy, the next step is to make your profile, links, offers and bookings easier to understand.