Short answer
Google has added a new filter to the Search Console performance report that separates image-based searches. Traffic from users who reach your site by searching with a camera, uploading an image or searching a screenshot now appears apart from classic text searches. There is no query data for this traffic; only the pages that appeared can be tracked.
Highlights
- The new 'multimodal' filter covers traffic from Lens, Circle to Search, image uploads and Chrome's 'Search image' feature.
- The filter has started rolling out globally; data appears only on sites that receive traffic from these kinds of searches.
- There is no query data for visual searches; only which page appeared can be tracked.
2 min read
According to an announcement on its Search Central blog, Google has added a new search type filter to the Search Console performance report: multimodal search. The filter shows traffic from searches in which users use an image instead of typing text.
What traffic does the filter cover?
According to Google, it covers searches made with Google Lens, Circle to Search on Android, uploading an image directly to Google Search and Chrome's right-click 'Search image' feature. As reported by Search Engine Journal, the filter sits under the 'Web' search type and splits text-based and image-based traffic into separate columns.
Why is query data missing?
Citing Google's dimensions page, Search Engine Journal wrote that there is no text query data for visual searches because the search is largely based on the image. The Queries tab is not shown when the multimodal filter is selected; site owners can see only which page received this traffic, not which image the user searched for.
What changes for websites?
The distinction matters for e-commerce sites that rely heavily on product photos and for publishers producing visual content such as recipes, home decor and fashion: until now, visits from visual search appeared in the same column as text search and went unnoticed.
The lack of query data makes the 'which keyword did it come from' reporting that agencies are used to meaningless for visual search. Instead, the focus needs to shift to page-level performance and the quality of product images.
- In monthly reporting, read 'Web' traffic as separate text and image figures rather than a single line; if the image share is large, set aside a separate budget for product photo quality.
- Because there is no query data, use Export to download which pages get a share of visual search traffic and review them manually.
- If visual search traffic is growing, prioritise clean backgrounds, accurate file names and alt text in product images.
Frequently asked
- On which sites does the multimodal search filter appear?
- It appears only in the Search Console dashboard of sites that actually receive traffic from these kinds of searches; if there is no such traffic, the filter option may be empty.
Sources
Follow UNIT Journal
What's new in search, AI and technology, in your feed every day.
