(7 min. read)
Key takeaways
- multimodal search uses an image as part of the query;
- a new search type, Web: multimodal, has appeared in Search Console;
- the data covers Google Lens, Circle to Search, uploaded images and image search in Chrome;
- multimodal search should be distinguished from traffic from Google Images;
- the “Queries” dimension is not available for multimodal searches;
- you can primarily analyse pages, devices, countries and search performance.
What is multimodal search?
Multimodal search allows users to search using more than one type of input. In Google Search, one of the clearest examples is combining an image with text.
Instead of typing “white women's trainers” into Google, a user can take a photo of a product or upload a screenshot and then add a question about what appears in the image.
For example, a user might:
- Take a photo of a lamp with Google Lens and look for a similar model.
- Select a pair of shoes in a photo using Circle to Search.
- Upload a photo of a product they found on social media to Google.
- Choose “Search this image” in Chrome to find out more about a picture.
The image becomes part of the query, and Google uses its contents to match relevant search results.
Image search, Google Images and multimodal search – what's the difference?
In SEO, it's easy to confuse multimodal search traffic with ordinary traffic from Google Images. However, they are two different types of data.
In Google Search Console, the Image search type refers to traffic from image search results. A user searches for something on Google and visits the Images tab, where one of the results is an image linking to a website.
With Web: multimodal, the situation is reversed: the image forms part of the user's search. Google can then display standard web results, product pages or other content related to what it identifies in the photo. The official Search Console documentation now distinguishes between Web: text-based, Web: multimodal, Image, Video and News searches (Google).
Put simply:
|
Search type |
Example |
|
Web: text-based |
A user types “red women's trainers”. |
| Web: multimodal |
A user uploads a photo of trainers or uses Lens. |
| Image | A user browses results in Google Images. |
This distinction matters when analysing a website's visibility. Strong performance in Google Images does not necessarily mean equally strong visibility in multimodal search.
What has changed in Google Search Console?
Google announced the update on 24 September 2026. Multimodal search data was added to the Search results performance report and the generative AI features report.
To find it, go to the Performance report and use the search type filter. If a website receives traffic from multimodal searches, you can select Web: multimodal. The rollout began globally, so the new data appears in Search Console for websites that actually gain visibility through this type of search (Google for Developers).
This is not an entirely separate tool or a standalone SEO report. Instead, it's a new way to report on and filter multimodal search traffic within existing Search Console reports.
How can you find multimodal search data?
- Open Google Search Console.
- Select the relevant property.
- Go to Performance /
Search results. - Open the search type filter.
- Check whether Web: multimodal is available.
- Apply the filter and analyse the data for pages, devices and countries.
In the standard report, Search Console lets you analyse clicks, impressions, CTR and average position. You can also export the data to Google Sheets, Microsoft Excel or a CSV file (Google).
Why aren't queries shown in the report?
This is one of the biggest differences between traditional SEO analysis and multimodal search analysis.
In traditional search, analysis often starts with keywords. We look at user queries, impressions, clicks, CTR and the pages shown for particular search terms.
In multimodal search, however, the Queries dimension is unavailable. Google explains that these searches are primarily based on images, so specific text query data cannot be displayed in this view (Google).
Instead, start your analysis with questions such as:
- Which pages generate the most impressions from multimodal search?
- Which URLs receive the most clicks?
- Does most of the traffic come from mobile devices?
- Which markets generate the greatest visibility?
- Do certain content types or product pages perform better than others?
It's a shift in perspective: instead of starting with a keyword, you start with the page and its visual content.
Multimodal search vs. AI Mode
Multimodal search and AI Mode are related, but they are not the same thing.
Multimodality primarily describes how a search begins – the user includes an image as part of their query.
AI Mode, on the other hand, is a Google Search feature that uses generative AI to present answers and results. Google provides a separate performance report for generative AI features, including AI Mode and AI Overviews. In this report, you can also filter data by the Web: multimodal search type.
So you shouldn't assume that every multimodal search takes place in AI Mode.
What does the new report mean for SEO?
The most important change is the ability to measure the visibility of visual content more precisely.
Until now, it was difficult to determine how much searches performed using photos or Google Lens contributed to a website's visibility. The new reporting makes it possible to isolate this area and track its performance separately.
This may be particularly useful for websites where images play an important role, such as:
- e-commerce websites,
- fashion and beauty websites,
- interior design websites,
- automotive websites,
- property websites,
- travel websites,
- product and catalogue websites,
- websites showcasing projects or portfolios.
For example, a user spots an interesting chair in a restaurant, takes a photo of it with Google Lens and finds a shop selling a similar model. For an e-commerce business, this is a very different route to a product than a traditional text query.
How can you analyse multimodal search performance?
To begin with, treat the new data as an additional traffic segment and monitor how it develops over time.
Focus on the following:
1. Pages with the most impressions
This helps you identify which types of content appear most often for image-based searches.
2. Clicks and CTR
A high number of impressions with few clicks may indicate an area worth investigating further.
3. Devices
Google Lens and Circle to Search are closely associated with smartphones, so it's worth checking the share of mobile traffic.
4. Countries
For international websites, you can compare visibility across different markets.
5. Page types
Check whether product pages, articles, galleries, category pages or other content formats dominate the results.
You can also export Search Console data and combine it with other sources in an analytics tool or dashboard.
How can you optimise pages for multimodal search?
Google has not published a separate set of guidelines for “multimodal SEO”. Multimodal search still relies on content accessible to Google Search, so standard SEO and image optimisation practices remain the starting point.
Google recommends practices such as:
- using high-quality images relevant to the page's content;
- placing images close to the text they relate to;
- using descriptive file names;
- writing useful alt text;
- using the standard HTML <img> element rather than placing important images exclusively in CSS backgrounds;
- making sure images can be discovered and indexed;
- using appropriate schema markup where relevant.
For example, a file name such as black-running-shoes.jpg gives the search engine more context than IMG_1847.jpg. The image itself, its quality, the surrounding content and accurate alt text are also important.
What doesn't the new report show?
The new data also has limitations worth bearing in mind.
First, Search Console does not show the exact image that a user submitted as a query. There is no equivalent of a traditional keyword list.
You also cannot break down the traffic into separate categories such as “Google Lens”, “Circle to Search” or “Search this image”. Google identifies these features as sources within Web: multimodal, but the report is designed to analyse multimodal search traffic as a whole.
There is no need to install a special script or add any extra code to enable the reporting. If your website generates visibility from this type of search, the data can appear in Search Console.
Does multimodal search change SEO?
The new report reveals a part of user behaviour that was previously much harder to isolate in the data.
Google is developing Search towards more visual and multimodal ways of finding information. The new Search Console data provides additional context: photos and graphics are worth analysing not only as part of Google Images, but also as part of the broader discovery process in Google Search.
For SEO specialists, this is an opportunity to understand more precisely how visual content supports page visibility. It can be a particularly valuable area of analysis for online shops, product websites and brands that publish a lot of images.