Search Console multimodal data: camera search is new traffic
Google added a multimodal search filter to Search Console on September 24, covering Lens and Circle to Search. It is new data, not a reshuffle — and there are no queries.
Google started rolling out multimodal search reporting in Search Console on September 24, and it answers a question every retailer has been guessing at: how many people found your product by pointing a camera at something. The Search Central announcement adds a multimodal search type filter to the Performance report. The catch — and the useful part — is that there are no queries behind it.
What actually happened
The new filter covers web search results where an image was part of the search: Google Lens, Circle to Search on Android, image uploads to Google Search, and Chrome's right-click "Search this image." You get the normal metrics — clicks, impressions, CTR, average position — sliced by page, country, device and date. It is rolling out globally, and it shows up only if your site actually gets this traffic.
The Queries dimension is switched off. Google's explanation, quoted by Search Engine Roundtable: "Because multimodal searches mostly use images rather than text, specific text query data isn't available for this traffic."
The detail that will confuse people next week came from Google's John Mueller, per the same report: "The data wasn't previously in the counts, the report has new data." This is not Google carving an existing slice out of your Web totals. It is traffic that was never counted. If your Web numbers step up this week, that is the reporting change, not your campaign.
Why multimodal search matters for your business
Your product photos are now an index, and you cannot optimize them with keywords. Every SEO lever you know operates on text. Multimodal gives you a channel with measurable clicks and zero query data, which means the ranking input is the image itself and the page around it. That shifts the work to things most small retailers skip: clean product shots on plain backgrounds, multiple angles, real resolution, a descriptive filename and alt text, and Product schema on the page the image lives on. The model has to figure out what the object is before it can decide your page answers it.
Check the pages, because the pages are the query. With no keyword dimension, the page-level breakdown is your only signal about intent. If your camera traffic clusters on three SKUs, those are the products people photograph in the wild — on a shelf, in a friend's kitchen, on a job site. That is a shopper with the physical object in front of them and no brand name to type. Fix those listings first.
Re-baseline before you report September. Any dashboard that trends Search Console clicks month over month just took an artificial bump. Note the date, exclude multimodal from the comparison, or expect to explain a number you did not earn. We keep saying the same thing about product data that machines read: the measurement layer is moving faster than most reporting stacks.
Export it now and keep it. Search Console's retention is finite and this data starts today. Pull the multimodal slice into your warehouse or a sheet on a schedule, so in six months you have a baseline instead of a rolling sixteen-month window.
Key takeaways
- Google began rolling out a multimodal search type filter in Search Console globally on September 24, 2026
- It covers Lens, Circle to Search on Android, Search image uploads and Chrome's "Search this image"
- Clicks, impressions, CTR and position are available by page, country, device and date
- No query data — Google says image-based searches do not produce text queries to report
- Google's John Mueller says this is new data that was not previously in the counts, so Web totals may rise
- With no keywords to target, the levers are image quality, alt text, filenames and Product schema on the hosting page
- Start exporting the multimodal slice now; the history only exists going forward
A search channel with no keywords is won on structured data and image hygiene. We audit product imagery, alt text and schema markup across a catalog, then wire the export so this new traffic lands in your reporting instead of quietly inflating it. See how we build storefronts machines can read, or ask for a product image and schema audit.
Sources: Google Search Central, Search Engine Roundtable.
- #search-console
- #google-lens
- #visual-search
- #seo
- #product-images
Tommy Rush — Founder, Rush Commerce
Operator turned builder. 15+ years running operations — now shipping the systems businesses run on. More
Get The Rush Report weekly — one email, zero fluff.
Keep reading
51% of US shoppers now use AI tools to shop, NIQ says
NIQ's Agentic Commerce Tracker puts AI shopping adoption at 51% of US consumers in the past month. Recommendations lead at 20%. Read the sample size before you act.
Read it24% of shoppers now compare prices with AI, not Google
Akeneo's September survey: 24% of US consumers use ChatGPT or Gemini to compare prices, 56% trust the answer, and only 32% trust retailers to price fairly.
Read it