Google has made a major part of visual search measurable in Search Console. Starting September 24, 2026, publishers can use a new multimodal search type filter to isolate performance generated when users begin a search with an image rather than a conventional text query.
The reporting covers searches initiated through Google Lens, Circle to Search on Android, image uploads to Google Search and Chrome's “Search this image” feature. The rollout is global.
Visual search traffic is finally separated in Search Console
Until now, publishers could see aggregate Search performance but had limited first-party visibility into traffic specifically originating from Google's increasingly important image-led search experiences.
The new filter creates a dedicated way to analyze how web content performs when the search journey starts from a photograph, screenshot, camera view or another image.
The filter is available in two Search Console reports
Google says multimodal reporting is available both in the standard Search results Performance report and in the newer Generative AI features report.
That placement is notable because multimodal discovery can intersect both traditional web results and generative search experiences.
Clicks, impressions and landing pages become measurable
Publishers can isolate performance from multimodal searches and analyze the resulting clicks and impressions, as well as the pages users ultimately reach. The filtered data can also be exported for analysis in external tools.
This creates a practical way to identify which pages and visual assets are attracting discovery from image-led search behavior.
Lens and Circle to Search are part of the dataset
The dataset includes Google Lens and Circle to Search, two experiences that increasingly blur the boundary between image recognition and conventional search.
It also includes searches initiated by uploading an image directly to Google Search and Chrome's right-click “Search this image” workflow.
The missing piece: textual queries
There is an important limitation. Search Console does not provide conventional text-query reporting for these multimodal searches when the interaction is primarily image-based.
That means publishers can measure exposure, clicks and landing-page performance without necessarily seeing a textual phrase that represents what the user searched for. The optimization workflow therefore becomes more page- and visual-centric than classic keyword analysis.
Why this matters for Visual SEO
The new reporting creates a measurable feedback loop for image optimization. Publishers can identify pages receiving multimodal exposure, compare impressions with clicks and investigate the images and surrounding content associated with successful visual discovery.
For ecommerce, travel, local search, products, fashion, food, landmarks and other visually driven verticals, the new filter can expose a layer of search demand that conventional query reports do not fully describe.
The NetContentSEO view
This is effectively the arrival of a first-party Visual SEO performance dataset inside Search Console.
The key shift is that search intent no longer needs to begin as words. A camera image, screenshot or object selected on screen can initiate the retrieval process, while Search Console now gives publishers a way to measure the resulting web visibility.
For SEO teams, that suggests a broader optimization model: pages should not only answer textual queries but also contain clear, useful and contextually meaningful visual information that Google's multimodal systems can connect to real-world objects and user needs.
Google is rolling the integration out globally, and properties will begin seeing the metrics when they receive traffic from eligible multimodal searches.