Google just gave website owners a window into a type of traffic they’ve never been able to measure directly: searches that start with an image instead of a typed query. On September 24, 2026, Google Search Central announced a new “Multimodal” search type filter in the Search Console Performance report, splitting your traffic into “Text-based” and “Multimodal” for the first time.
If you’ve ever wondered how much of your traffic comes from someone pointing their phone camera at a product, circling something on their screen, or uploading a screenshot to Google, you’re about to find out. Here’s what the update actually does, where to find it, and what to do with the numbers once you see them.
What Google Just Rolled Out
The new filter lives inside the Performance report you already use every day. Under “Search type,” alongside the existing “Web,” “Image,” “Video,” and “News” filters, you’ll now find an option to break results down by “Text-based” versus “Multimodal” search.
According to Google’s announcement, a search counts as multimodal when it’s “triggered by a query that uses an image, photo, or screenshot” rather than typed text. That covers a wider range of behavior than most site owners realize, and it’s been growing quietly in the background for a couple of years while most SEO reporting stayed focused on typed keywords.
What Counts as “Multimodal Search” in Practice
Multimodal search is any search that starts with an image rather than words. Google’s rollout groups four specific triggers under this label:
- Google Lens, where someone points a phone camera at an object, a plant, a product tag, or a page of text and gets search results back.
- Circle to Search on Android, where a user circles or taps something on their screen inside any app, not just a browser, and Google searches for it without leaving that app.
- Image uploads directly into Google Search, where someone drops a photo or screenshot into the search bar instead of typing.
- “Search this image” in Chrome, the right-click option that’s been in browsers for years but was never broken out as its own traffic category.
None of this is brand new behavior from users. What’s new is that Google is finally letting you see it separately from typed search, instead of burying it inside your regular “Web” numbers.
Why Google Built This Report Now
Visual search has been growing for years, but SEO tooling hasn’t kept pace with it. Rank tracking, keyword research, and most Search Console reporting were all built around the assumption that a “search” starts with typed text. That assumption stopped being fully true a while ago, especially on mobile, where Circle to Search and Lens are now built into the operating system rather than tucked away in an app.
Google has also been pushing users toward AI Mode and other visual-first discovery tools, and internally that means more of its own traffic mix is shifting toward non-text input. Giving publishers visibility into that shift is a reasonable next step: it’s hard to optimize for a type of search you can’t measure, and until now, multimodal traffic was invisible in your reporting even if it was already showing up in your analytics as unexplained “Google / organic” sessions.
Where to Find It in Search Console
To see your own numbers:
- Open Google Search Console and go to Performance > Search results.
- Click Search type, the same filter you’d normally use to switch between Web, Image, and Video.
- Choose Multimodal (or compare it side-by-side with Text-based) to see clicks, impressions, average CTR, and average position for image-triggered traffic.
- Use the date comparison tool to check whether multimodal traffic is growing, shrinking, or flat for your site over the last few months.
If your site gets little or no image-based traffic, you may see very small numbers here, and that’s expected — this isn’t a report every site needs to obsess over daily. But if you run an e-commerce store, a recipe blog, a plant or animal identification site, or anything with visually distinctive content, it’s worth a look before you decide it doesn’t apply to you.
What You’ll See and What You Won’t
This is the part worth understanding before you get too excited about the new data. Google’s own documentation is explicit about a limitation here: you don’t get query-level data for multimodal search. Because these searches start with an image rather than text, there’s no keyword to report on the way there is for a typed query.
What you do get is aggregate performance: clicks, impressions, click-through rate, and average position, filtered to multimodal traffic, at the page level. You can see which of your pages are earning visual-search visits and how that traffic is trending, but you can’t see exactly what someone photographed or circled to land there. Think of it as a traffic-quality signal rather than a keyword research tool useful for spotting patterns, not for building a new content calendar around “multimodal keywords” that don’t technically exist.
Why This Matters for Your SEO Strategy
For most site owners, this report is less about a new tactic and more about validating (or correcting) assumptions about where your traffic actually comes from.
If You Run an E-Commerce or Recipe Site
Product pages, recipe pages, and anything with strong, recognizable imagery are the most likely candidates for meaningful multimodal traffic. If someone photographs a product in a store, or a dish they saw on a friend’s table, and lands on your page through Lens, that’s a visit your old reporting never separated out as anything unusual.
This is a good moment to double-check the basics that make pages eligible for this kind of discovery in the first place: high-resolution product and hero images, descriptive (not generic) file names, complete alt text, and Product or Recipe structured data with image fields properly filled in. If you haven’t audited your structured data recently, this is as good a reason as any to do it now. Incomplete or missing schema doesn’t just affect rich results; it affects whether Google can confidently match an uploaded photo to your page in the first place. None of this is new advice, but it now has a traffic number attached to it that you can actually track over time.
If You Publish Editorial or Blog Content
Text-heavy content explainers, opinion pieces, and news will generally show smaller multimodal numbers, and that’s fine. Where it’s worth paying attention is any post that includes charts, infographics, screenshots of tools or interfaces, or diagrams people might photograph off a screen or reference back to later. If you cover software, data, or how-to topics with visual steps, checking this report periodically can tell you whether those visual assets are pulling their weight.
How to Act on This Data This Week
You don’t need a new strategy overnight, but a few concrete steps are worth doing now that the data exists:
- Check your baseline. Filter Performance by Multimodal search type and note where you stand today; even a rough sense of “not much” versus “surprisingly a lot” is useful context you didn’t have last week.
- Cross-reference with your top pages. See which URLs are earning multimodal clicks and whether they’re the pages you’d expect (product pages, recipes, visual guides) or something unexpected.
- Audit image fundamentals on your top performers. Confirm images are crawlable (not blocked by robots.txt), compressed but not degraded, served at reasonable resolution, and paired with accurate
alttext and file names. - Verify structured data on visual content.
Product,Recipe, andImageObjectschema help Google understand what an image actually shows, which matters more, not less, for search that starts with a picture instead of a keyword. - Set a recurring check-in. This is a new, evolving report. Rather than a one-time audit, add a monthly glance at the Multimodal filter alongside your usual Search Console review.
The Bigger Picture: Google Is Investing in Multimodal Discovery
This report doesn’t exist in isolation. It arrives alongside Google’s continued expansion of AI Mode and its broader push to let people search however feels natural by typing, talking, or pointing a camera at something. Treat this less as a one-off Search Console feature and more as an early data point in a longer trend: search is steadily becoming less dependent on the keyword box, and reporting tools are slowly catching up to that reality.
For SEO professionals, the practical implication is that “image SEO,” a category that’s often treated as secondary to keyword-driven content strategy, deserves a second look. It’s no longer just about ranking in Google Images; it’s about being discoverable and correctly represented when someone’s query is a photo rather than a phrase.
That’s the same underlying shift covered in our guide to ranking in AI search. The common thread across AI Overviews, AI Mode, and now multimodal search is that Google is getting better at understanding intent from inputs that aren’t a clean, typed keyword, and sites that make their content easy to parse visually and structurally tend to benefit across all three.
Frequently Asked Questions
Does this replace the existing Image search filter in Search Console?
No. “Image” search type still refers to results shown on the dedicated Google Images tab. “Multimodal” is a separate category covering web search results that were triggered by an image-based query — the two can overlap, but they’re tracking different things.
Do I need to do anything to start seeing multimodal data?
No setup is required. If your site already appears in Google Search Console’s Performance report, the Multimodal filter will show data automatically once you have qualifying traffic, with no new verification or configuration needed.
Will this replace keyword research for image-heavy pages?
Not directly. Because Google doesn’t expose query-level detail for multimodal search, you can’t use this report to find new keyword opportunities the way you would with a typed-search query list. It’s better used as a performance and trend indicator layered on top of the image SEO fundamentals you’re likely already tracking.
Is multimodal traffic counted in my regular Search totals?
Yes. Multimodal clicks and impressions were already included in your overall Web totals before this update; the new filter simply lets you isolate and view that slice separately. Your total numbers shouldn’t change; only your ability to segment them does.
Limitations and What to Watch
A few caveats are worth keeping in mind so you don’t over-read this data:
- No query data means you’re working with directional signals, not specific search terms to target.
- This is a new, evolving report. Google has changed Search Console reporting boundaries before, and multimodal categorization could be refined as usage grows.
- Small sites may see negligible numbers, and that alone doesn’t mean anything is wrong with your images it may simply reflect the nature of your content.
- Correlation isn’t causation. A rise in multimodal traffic doesn’t necessarily mean a specific change you made caused it; treat trends over a few months as more meaningful than week-to-week swings.
Google has quietly been routing a meaningful and growing slice of search traffic through images instead of text for a while now, and for the first time, you can actually see your share of it. The data won’t hand you new keywords to target, but it will tell you whether your visual content is pulling weight and for e-commerce, recipe, and visually driven sites especially, that’s a report worth checking regularly rather than filing away and forgetting.



Leave a response