Breaking
UTM Parameters Turn $338B in Social Ad Spend Into ProofP&N Bank flips the customer-owned banking scriptHow to Get Products Found in Google and ChatGPT AI ShoppingWhy PageSpeed Scores Might Jump After Lighthouse 13.6TikTok Search Ads Adds Impression Share and Keyword PacksUTM Parameters Turn $338B in Social Ad Spend Into ProofP&N Bank flips the customer-owned banking scriptHow to Get Products Found in Google and ChatGPT AI ShoppingWhy PageSpeed Scores Might Jump After Lighthouse 13.6TikTok Search Ads Adds Impression Share and Keyword Packs
SEO

Google Search Console splits visual search out of Web

Search Console now separates Lens and Circle to Search traffic from text-based web results. No query data, new measurement for visual SEO.

Google splits visual search from text in Search Console

What Google just turned on

Google is rolling out a new split in Search Console performance reporting globally. The Search type filter now divides Web results into two buckets: Text-based and Multimodal. Multimodal captures web pages surfaced by searches that use an image, photo or screenshot rather than typed words. Google says these figures were not previously counted anywhere in the report.

The filter covers four entry points: Google Lens, Circle to Search on Android, image uploads to Google Search, and Chrome’s right-click ‘Search this image’ option.

What you can and cannot see

The new view reports impressions, clicks, pages, position, countries and devices. What it does not report is the query. Google states that ‘specific text query data isn’t available for this traffic.’ John Mueller put it more simply: site owners can see ‘which of your pages showed up & how often,’ but the photo itself stays hidden.

That is not the usual anonymised-query gap. In text-based reporting Search Console withholds rare queries and caps table rows. In the multimodal view, the queries dimension is removed entirely. Google’s language also leaves room for mixed image-and-text searches: the tooltip says an image was used ‘as part of the search’ and multimodal searches ‘mostly’ use images, so a photo plus typed refinement appears to fall into the multimodal bucket, while the typed part is still not reported.

The same split appears in the Generative AI features report. That report already lacks clicks and queries, so its multimodal slice reduces to impressions by page, country, device and date. It can show a camera search surfaced a page inside an AI response, but not what the camera saw or whether anyone clicked. UK regulators have asked for click-through data on generative AI features; this filter adds segmentation but not the missing click numbers.

Why this matters

Visual search is no longer an edge case. Google has reported nearly 20 billion monthly Lens searches back in October 2024, around 20 percent shopping-related, and a 65 percent year-over-year increase in visual searching by July 2025. Circle to Search passed 300 million Android devices. For retailers, travel, home and garden, and any catalogue built around objects people photograph, this was organic exposure that simply did not appear in Search Console.

That changes measurement. The new bucket cuts across the existing report rather than sitting alongside it. Search Console already has a record of redefining what ‘Web’ contains: AI Mode clicks merged into Web totals in June 2025, and logging errors changed impression baselines later. Until Google clarifies whether top-level Web now adds text-based and multimodal, a line labelled Web in a dashboard may not mean the same thing it meant last month.

A practical checklist

For SEO and agency teams, the update is less about chasing a new metric and more about avoiding false comparisons.

  • Open the Search type filter and check whether your properties already have multimodal rows.
  • Annotate the date the data starts, and segment Web totals before any year-on-year comparison.
  • Review pages with camera-driven impressions separately, especially product and category pages.
  • Export the new slice now; the API and bulk data export are not yet mentioned as carrying it.

None of this tells you what the image showed. But it finally shows which pages are winning camera-driven demand, and that is a gap every visual commerce team should want closed.

Source: PPC Land