What I Do

Search Console Multimodal Search: See Google Lens Traffic Separately

7 min read

Google added a multimodal search filter to Search Console. How to separate traffic from Google Lens, Circle to Search, and image uploads from text search.

Search Console Multimodal Search: See Google Lens Traffic Separately

On September 24, 2026, Google released Search Console multimodal search data. A search type filter called "Web: Multimodal" was added to the Performance report and the generative AI performance report, so you can now see how often people who searched with images, through Google Lens or Circle to Search, saw and clicked your site, separately from text search. In this post, I'll cover exactly what kind of search multimodal search refers to, where and how to check it, and what to review once you see these numbers.

You've probably seen a bag you liked while walking down the street, opened your phone camera, and looked for similar products at least once. As more people search with photos instead of typing words into a search box, marketers naturally start wondering, "Are people who search with photos landing on our pages too?" Until now, Search Console had no filter to pick out just these searches, but with this update you can check for yourself.

What was added to the Search Console multimodal search report?

According to the Google Search Central blog announcement, two places changed.

  • Performance report (Google search results): In the search type filter, web search can be split into "Web: Text-based" and "Web: Multimodal."
  • Generative AI performance report: Impressions from AI Overviews and AI Mode can also be split into text and multimodal the same way.

Think of it like a shop with two entrances. Until now, customers were only counted as a total; now you can count customers who came through the front door (words typed into the search box) separately from those who came through the side door (a camera or image). Google has also reflected this data in the Performance report for search results help article.

So what exactly counts as the customers coming through the side door, that is, multimodal search?

Multimodal means using several forms of input together, such as images, not just text. Google's announcement and help article describe "Web: Multimodal" as traffic from searches made with images, such as with a smartphone camera, and list the following four paths as included.

  • Google Lens: Searching with your camera, images, or screenshots (About Google Lens)
  • Circle to Search: A feature on some Android phones where you circle or tap an image or text on screen to search (Circle to Search help article)
  • Image upload: Uploading an image to Google Search to search
  • Chrome image search: Right-clicking an image in Chrome to search it (Search this image)

The part that's easy to confuse here is the existing "Image" search type. According to the help article, "Image" is traffic from results shown in the Google Images tab, while "Web: Multimodal" is traffic from web search results where an image was used as part of the search. In other words, someone who typed a query and browsed the Images tab and someone who took a photo and searched with it are counted in different buckets.

How do you check multimodal data in Search Console?

It's simple to check.

  1. In Search Console, open the Performance report (Google search results).
  2. Click the Search type filter at the top of the report and choose "Web: Multimodal."
  3. Switch between the Queries, Pages, Countries, and Devices tabs to see which pages got impressions and clicks from image search.

If you want to compare it side by side with text search, use the Compare feature. According to the advanced filtering help article, the search type filter can't be removed, you need to use the compare feature to view multiple search types together, and comparisons are limited to two items. If you compare "Web: Text-based" and "Web: Multimodal," you can see on one screen how the same page performed through each path. For more detailed analysis, you can also download the data with the Export button and work with it in a spreadsheet.

You can use the same filter in the generative AI performance report as well. This report was released to some sites in June 2026, and according to the help article, it became available to all websites worldwide on August 31. It shows how many times links to your pages appeared in AI Overviews and AI Mode as impressions, so you can separately see when your pages were shown as links in AI answers to people who searched with images.

Note: What should you know when looking at multimodal data?

  • Gradual rollout: Google said it's rolling out gradually worldwide starting September 24. It's not unusual if you don't see the menu yet.
  • You only see it if there's traffic: The announcement says the metrics start appearing when your site has traffic from these searches. Sites far removed from image search may see almost no numbers.
  • The generative AI report centers on impressions: The default metric in this report is impressions, and according to the help article, it doesn't include data from Search Labs experiments.
  • Wording in the Korean help article: The Korean help article describes "Web: Multimodal" only as "web search results where an image was used as part of the search," and the list of paths such as Lens and Circle to Search is in the English help article and the announcement.
  • Google Search only: Search Console covers Google Search data. Traffic from image searches on other search engines, such as Naver (Korea's largest search engine), isn't included.

What should you review after looking at multimodal data?

In the announcement, Google said it hopes this data helps you understand your site's performance and optimize your visual content. However, the announcement doesn't describe any separate optimization method just for multimodal search, so I recommend reviewing against Google's image SEO best practices.

  • How images are embedded: Google explains that it doesn't index CSS background images and finds images in the src attribute of HTML <img> elements.
  • Alt text: Google says it uses alt text along with computer vision and page content to understand what an image is about. It's better to write a sentence describing the image than to list keywords.
  • Text around images: It recommends placing images near relevant text, on pages that match the image's topic.
  • File names and quality: It advises short, descriptive file names and sharp images rather than blurry ones, while also keeping page speed in mind.

For pages about things people might look up with a camera, like product photos, interiors, or fashion, a natural order would be to first see which pages are getting impressions in the multimodal filter, then review those pages' images against the items above. In an earlier post, I noted that traffic from AI features was hard to separate out, but now you can break out both AI feature impressions and image-based searches. Once the multimodal filter opens up in this blog's Search Console, I'll look at which posts show up and write it up again.

FAQ

How are the 'Web: Multimodal' and 'Image' search types in Search Console different?
"Image" is traffic from results shown in the Google Images tab. "Web: Multimodal" is traffic from web search results when people search with images, such as through Google Lens, Android's Circle to Search, image upload, or Chrome image search.
Where can I see Google Lens search traffic in Search Console?
In the Performance report (Google search results), change the search type filter to "Web: Multimodal." Impressions from AI Overviews and AI Mode can be seen with the same filter in the generative AI performance report. Google said it's rolling out gradually worldwide from September 24, 2026, and the metrics only appear if there's traffic from these searches.
Can I compare text search and multimodal search on one screen?
Yes. With the compare feature in the search type filter, you can view "Web: Text-based" and "Web: Multimodal" side by side. Comparisons in Search Console support only two items.

Related posts

Search ConsoleAI SearchSEO