Google began adding a “multimodal” search type to Search Console’s Performance report on 24 September, separating web results reached through an image from those reached through typed text. Google said the traffic was “not previously in the counts”, which means site owners have never seen it in the report.

The gap matters for B2B SaaS marketing teams because product screenshots, dashboards and workflow diagrams are the kind of visuals a searcher can capture and send to Google to identify a tool. Until now, any visit that began that way was invisible in the tool teams use to justify organic search investment. The catch is that the new view shows which pages surfaced, not why: Google supplies no query data for it.

What did Google add to Search Console on 24 September?

Google added a Web multimodal search type to Search Console on 24 September, announcing on its Search Central blog that it was “rolling out globally starting today”. The new category sits inside the existing Search type filter. Under Web, the report now separates Text-based from Multimodal, and Google’s help documentation defines the second as web results where an image was used as part of the search.

The multimodal bucket covers four entry points, according to Google’s announcement as reported by Search Engine Journal and Search Engine Roundtable: Google Lens, Circle to Search on Android, image uploads to Google Search, and Chrome’s right-click “Search this image” option. The same split also appears in the Performance report for generative AI features.

Visibility depends on traffic. Google said the metrics appear if a site receives visits from these searches, so a property with none may show an empty view at first. The reviewed sources do not say whether existing Web totals now include the multimodal traffic or whether it appears only when the filter is applied, so teams should check their own accounts before reading anything into a change in the headline numbers.

What does the multimodal view show, and what is missing?

The multimodal view in the Search results report shows impressions, clicks, click-through rate and average position, broken down by page, country and device, and it can be exported like other Search Console data. Queries are the missing piece. Google’s explanation: “Because multimodal searches mostly use images rather than text, specific text query data isn’t available for this traffic.”

That leaves a page-level picture. A team can see which URLs appeared after an image-led search and how often people clicked. It cannot see the image or screenshot that started the search, and it cannot separate Lens from Circle to Search, image uploads or Chrome’s image search, because all four sit in one bucket. A team therefore cannot say from Search Console alone that a particular screenshot drove a particular visit.

The generative AI features report has tighter limits still. Google’s help documentation describes impressions as its primary metric, and Search Engine Land noted that click and query data remain unavailable there. The Helm reported in August that the generative AI performance report carries no click or query data, which left teams short of evidence for judging what an AI opt-out might cost.

What can B2B SaaS teams do with page-level visual search data?

B2B SaaS teams can start by exporting the multimodal view for a fixed date range and recording impressions, clicks, click-through rate and average position by landing page, country and device. Google has said the traffic was absent from earlier reports, so that export is a new baseline rather than a continuation of the old one, and comparisons with pre-24 September totals could mix two different measures.

The pages that appear are the useful part. Product pages, documentation with interface screenshots and posts built around diagrams or charts are the likeliest candidates, and a page with many multimodal impressions but few clicks points to a mismatch between the image that prompted the search and what the landing page shows. Descriptive file names, captions and alt text, along with the same information in crawlable text, are standard image-search practice, and this data gives teams a reason to check them on the pages that already surface.

Scale is the unknown. Google reported more than 20 billion Lens searches a month in 2024, and Alphabet chief executive Sundar Pichai said in April 2025 that Circle to Search was available on more than 250 million devices, but those figures measure usage and availability, not visits sent to websites. The Helm found no published figure for the share of web clicks that Lens or Circle to Search generate, or for how many Lens searches end in a click to an external site.

Google has not said how much of the web’s traffic starts with an image, and the new filter gives site owners their first way to size it for themselves.