Image SEO used to be a checklist item: compress the file, add alt text, move on. That no longer covers it. Google describes Search as evolving to include more visual and multimodal experiences, and in those searches the query is a photo, not a string of words. Someone points their phone at a lamp in a hotel lobby, circles a jacket in a video or right-clicks a product shot in Chrome, and Google has to decide which page best matches the picture. With the right signals, that page can be yours.
Until now, that traffic had no filter of its own in Search Console. On September 24, 2026, Google announced web multimodal search reporting: a new multimodal search type filter in the performance report for Search results and in the report for Generative AI features. It covers searches with Lens, Circle to Search on Android, image uploads to Google Search and the Chrome right-click "Search this image" feature.
That turns visual search SEO into something you can measure. Below we cover what the filter shows, how to analyse the data, and how to optimise images and pages so they get found when someone searches with a camera. For webshops, this is where image SEO pays off most directly.
What the new multimodal filter in Search Console shows
Multimodal is now a search type in two places: the performance report for Search results and the Generative AI features report. According to Google, the data includes:
- Lens, for example when someone points a smartphone camera at an object
- Circle to Search on Android
- Image uploads to Google Search
- The right-click "Search this image" feature in Chrome
This update is designed to give you insights into how your content is surfaced when users search using images (such as with a smartphone camera).
Google Search Central Blog
One line deserves attention. Google says you will start seeing these metrics if your site is receiving traffic from these queries. An empty multimodal view is a result in itself: either the rollout, which started globally on September 24, has not reached your property yet, or Google is not surfacing your pages for image-based searches. For a fashion or furniture shop, the second explanation should worry you.
How to analyse and export multimodal search data
Treat the first weeks as a baseline. The rollout only began on September 24, so you are building a history from scratch.
Start with pages, not totals
Set the search type to multimodal and look at which pages collect impressions. Those URLs show which products and topics Google already connects to visual searches. Then compare them with the regular web search type. A page that barely appears for text queries but does show up for image searches has photos doing work its copy is not. A strong page that never appears in multimodal results probably has generic or hard-to-crawl images.
Export the data and join it with your own
Google's advice is to click Export to download the data and analyse it in your other tools. Joined with your product catalogue, the export answers questions the interface cannot:
- Which categories earn visual impressions, and which visual categories earn none?
- Which pages get impressions from image searches but few clicks?
- Do pages with original photography outperform pages with supplier images?
Add multimodal as a separate line in your SEO reporting, so visual search does not vanish inside your organic totals.
Check the Generative AI features report too
The same filter works in the Generative AI features report. Comparing both shows whether your image-based visibility comes through regular results, generative AI features or both.
Why visual search works differently from text search
With a typed query, the searcher tells Google what they want. With a camera, they show it. The intent is usually simple: what is this, where can I buy it, does it come in another colour. Google has to recognise what is in the image and then pick the page that best fits that intent. So showing up depends on two things: can Google connect your image to the object in the photo, and is your page clearly about that object? Image SEO handles the first part, page relevance the second. You need both.
Image SEO basics that decide whether Google Lens finds you
Use original photography
If the same manufacturer photo appears on many webshops, Google has no reason to pick yours. Original photography gives it one. Shoot several angles, close-ups of materials and details, and at least one shot of the product in use. Those in-context images also look more like the photos people take themselves, which is exactly where a Lens search starts.
Get image quality and context right
A sharp, well-lit image where the product fills the frame is easier to match than a dark, cluttered one. Use enough resolution to show detail and compress the file so the page stays fast. Place images close to the text that describes them. Use real image elements in the HTML rather than CSS backgrounds, which Google does not pick up as images, and if you lazy-load, confirm that Google can still render them. That check belongs in every technical SEO audit.
Write descriptive file names and alt text
File names and alt text are among the few text signals attached directly to an image. A file named header-image-final.jpg says nothing, while walnut-round-side-table.jpg describes the product before Google looks at a single pixel. Alt text does the same in a short sentence: describe what is in the image, specifically enough to tell it apart from your other images, without keyword stuffing. It also helps visitors who use screen readers.
Make product and content pages relevant to image searches
A strong image on a weak page still loses. The page has to answer the intent behind the photo.
Add product structured data
Product structured data tells Google exactly what a page sells: name, images, brand, price, availability and, where they are genuine, reviews. Keep the markup consistent with the visible page and with your Merchant Center feed if you use one. Mismatches can lead to products being shown with the wrong details, or not at all.
Match the page to what people photograph
Give every variant that looks different its own crawlable image. Write product copy that names what a camera can see: material, colour, shape, pattern and style. On content pages, an image should illustrate the topic of the page, ideally next to the heading it belongs to, rather than decorate it. A plant photo on the care guide for that plant is a credible match; a generic stock image is not.
Where ecommerce teams should start
You do not need to reshoot the whole catalogue at once. Work in order of impact:
- 1Export a first multimodal baseline and note which pages already get visual impressions.
- 2Pick the categories people are most likely to photograph, such as clothing, furniture, lighting and plants.
- 3Replace supplier images on the best sellers in those categories with original photography.
- 4Fix file names and alt text in your product templates, so new uploads are right by default.
- 5Validate your Product structured data against your feed and the visible page.
- 6Compare the multimodal data month over month and move on to the next categories.
If you sell on Shopify, check what your theme does with alt text and Product markup first. A Shopify SEO review of the theme fixes those templates for the whole catalogue at once.
Frequently asked questions
What is multimodal search in Search Console?
It is a search type filter in the performance report for Search results and the Generative AI features report, announced by Google on September 24, 2026. It shows how your content is surfaced when people search with images via Lens, Circle to Search on Android, image uploads to Google Search and the Chrome right-click "Search this image" feature.
Why do I not see any multimodal data?
Google says the metrics appear if your site is receiving traffic from these queries, and the rollout started globally on September 24, 2026. No data means either the rollout has not reached your property yet, or your pages are not surfacing for image-based searches. If it is the second, start with your images and product pages.
How do I optimise images for Google Lens?
Use original, sharp photos that show the product clearly and in context, with descriptive file names, specific alt text and relevant text nearby. Make sure Google can crawl and render them, and make the page clearly about what is in the image.
Is image SEO still just alt text and compression?
Those still matter, but image SEO for visual search also depends on original photography and on a page that answers the searcher's intent. The multimodal filter now lets you check whether that work shows up in your data.
Get found by people who search with a camera
Visual search rewards shops that treat images as content rather than decoration, and Search Console can now show whether it works. Want to know where your product images and pages stand, and which fixes matter most for Lens and image-based searches? Talk to LASEO about ecommerce SEO. We will review your multimodal data, photography and product markup together and tell you where to start.