Published on 2 October 2026Nicolas Sacotte

Google Search Console: New Multimodal Filter Reveals Visual Search Data

Google Search is becoming increasingly visual. Users have long since moved beyond text-only queries, now searching with photos, screenshots, or directly via their smartphone camera. Google has continued to drive this shift this year, integrating images, videos, and Chrome tabs into its AI-powered search experience. The problem until now: how much reach and traffic is generated through such multimodal searches was barely measurable in Search Console as a separate category. That is about to change. Google is introducing a new reporting feature for multimodal web search and has announced the global rollout on LinkedIn.

Multimodal Filter: Which Searches Are Included in the New Reporting

The following search types are now included in the data:

Image uploads in Google Search
Google Lens
"Search this image" in Chrome
Circle to Search on Android

Whenever impressions or clicks are generated through these methods, they can now be evaluated using the new Multimodal Filter. For SEOs, this finally makes it measurable which content performs via visual searches and which pages are worth optimising further.

Where to Find the New Multimodal Filter

The filter is located within the Performance report in Search Console. Under Search type > Web > Multimodal, you can now filter specifically for multimodal searches.

The data is available both for standard Search Results and for the Generative AI Features report. For context: it was only at the beginning of September that Google made the AI Performance Reports available globally, allowing users to analyse the visibility of content in AI Overviews, AI Mode, and other generative search features.

Using "Export", the data can also be downloaded and analysed further in other tools.

New Data for Websites with Extensive Visual Content

The new reporting is particularly valuable for shops, publishers, and any website with a large amount of visual content. The filter shows which pages are generating impressions and clicks through multimodal searches. This makes it considerably easier to assess the contribution that visual search pathways make to organic visibility and traffic.

The Catch: No Search Queries

The new reporting does come with one limitation, however. For multimodal searches, Google does not provide a query dimension. This is understandable, since with Google Lens, screenshots, or image uploads, there is often no conventional text-based search term at all.

In practice, this means SEOs can see which pages receive impressions and clicks via visual search — but what image or visual search prompt triggered them remains unknown.

Initial Verdict

For image SEO, this is nonetheless a significant step forward. Visual content can now be more effectively linked to its actual search performance.

There is also this to consider: Google is making multimodal data available for Platform Properties such as YouTube, Instagram, and TikTok as well. Platform Properties were only rolled out globally in August. This already allows SEOs to understand how social media posts and videos perform in Google Search, Discover, and Google News. The multimodal data now adds a further layer of analysis for visual search pathways.

About the author

Nicolas Sacotte

Founder of Tasketeer

Nicolas Sacotte is founder of Tasketeer and has worked in search engine optimisation since 1998. He started during his business studies by building and optimising his own websites. Since 2012 he has run the SEO and content marketing agency ContentKing.de and advises companies in the DACH region, including TARGOBANK, comdirect, Interhyp, O2, SIXT, BURDA and sevDesk.

Tasketeer grew out of this agency practice: SEO rarely fails for lack of analysis, but because insights never become tasks with owners and deadlines.

Alongside client work, Nicolas builds and runs his own platforms, including visicheck.ai, womosuche.de, Zoom.Reviews and motorrad.news. There he implements what he writes about. His focus today is GEO (Generative Engine Optimization): how content becomes visible in ChatGPT, Perplexity and Google’s AI answers, how AI crawlers can be steered deliberately, and what structured data and grounding pages contribute.

He shares his knowledge in workshops, as a speaker and author at CAMPIXX, and in specialist articles on contentking.de. He also developed the SEO and GEO training in Tasketeer.