Google Search Console adds Multimodal filter for Lens and visual search traffic

A new Multimodal filter arrives in Search Console

Google has started rolling out a Multimodal filter inside Search Console's Performance reports, giving site owners a dedicated way to separate traffic that reaches them through visual search from ordinary typed queries. According to a report on dev.to, which cites Google's announcement, the filter sits under the Web Search type and covers four entry points: Google Lens, Circle to Search on Android, images uploaded directly to Google Search, and Chrome's right-click option to search for an image. The global rollout reportedly began on September 24, 2026, and, as with other Search Console updates, individual properties may gain access gradually.

The addition matters because Search Console has historically centred on text-led web search. Publishers of product imagery, how-to visuals, location photography or video thumbnails have had no clean, centralized view of whether those assets actually drive organic discovery.

Where the data appears and what it covers

The dev.to write-up notes that the new data lands in two places: the standard Search performance report and the Generative AI features report. Applying the filter narrows the numbers to the supported visual-search surfaces rather than leaving them blended into the wider web-search dataset, and Google states that the filtered results can be exported for analysis in external tools.

That separation is the practical value. A property might perform well in conventional text search while drawing very little measurable traffic from Lens or Circle to Search, or the opposite may hold for pages built around strong imagery. Someone using Lens to identify an object, circling something on an Android screen, or uploading a photo is operating in a different context from a user typing a query. With the filtered and unfiltered views available side by side, teams can compare the two patterns instead of treating every organic visit as though it originated from the same kind of interaction.

How SEO teams can use it

The sensible starting point, per the dev.to report, is measurement rather than assumptions. Once the filter shows up in a property, a focused review can include:

• Comparing multimodal performance against the broader Web Search view for the same period.
• Identifying which pages and content formats show up in the filtered data.
• Checking whether the images on those pages genuinely represent the page and its subject.
• Reviewing image alt text and relevant image structured data as part of routine technical SEO.
• Exporting the filtered report to combine with internal content, conversion or editorial analysis.

The patterns are likely to differ by site type. A retailer may find product pages in the dataset, a service business may see visual guides or portfolio pages, and a publisher may notice articles where original photography helps readers investigate a topic. The report supplies evidence of activity for a given site; a content review then determines whether the successful pages share qualities worth applying elsewhere.

Rollout and caveats

Availability should be checked property by property, since Google's rollouts tend to appear in stages rather than all at once. It is also worth tempering expectations: the filter makes visual-discovery traffic visible in one place, but it does not turn image optimization into a channel with guaranteed returns. A site with no measurable multimodal activity may simply have little to gain there, while a site with visible traffic now has a concrete basis for prioritizing media work.

Why it matters

The update acknowledges that a search session can begin with an image rather than a text box, and that such journeys are becoming a routine way for people to evaluate products, services and information. Until now, site owners had to treat traffic from Lens, Circle to Search and image uploads as effectively invisible within Search Console's reporting. One filter covering several Google surfaces replaces the need to assess each interaction separately, and exportable data lets teams fold visual discovery into existing analytics workflows. For anyone deciding where image optimization deserves time and budget, the change turns a guessing exercise into an evidence-led one, without replacing broader SEO reporting.