Multimodal Search in Google Search Console: A Wix Guide

Google has quietly added a new filter to the Search Console Performance report, and it started showing up for site owners in September 2026. It splits your web search data into two types, Text-based and Multimodal. If you have logged into Search Console lately and seen a new multimodal line, you are not imagining it, and you are probably wondering whether it matters. This is the full answer. I will explain what multimodal search actually is, what the new filter does and does not show you, what it means for a Wix site, how to do the image SEO that helps, and why a second new report in Search Console deserves far more of your attention than this one. I work only on Wix SEO, I am the UK's only Wix Marketplace Legend Partner working solely in SEO, and I have checked all of this against my own Search Console and Google's own documentation, so everything here is based on fact, not guesswork.

What is multimodal search?

Multimodal search is a search that starts with an image instead of, or alongside, typed words. Rather than typing a question, someone points their phone camera at an object, uploads a screenshot, or circles something on their screen, and Google returns results based on what it sees. The word multimodal simply means more than one mode of input, in this case a picture plus, sometimes, a few words to refine it.

In practice, Google counts these as multimodal searches:

• Google Lens, where you search using your camera or an existing photo
• Circle to Search on Android, where you circle part of what is on your screen to search it
• Uploading an image directly to Google Search
• The "Search this image" option when you right-click an image in Chrome

This is not a niche behaviour. Google has said Lens alone is used for around 20 billion visual searches a month, and that number has grown quickly as phone cameras and on-device AI have improved. People use it to identify a plant, find where to buy a piece of furniture, translate a menu, check a landmark, or shop for something they have seen but cannot name. For the person searching, it feels natural. For a site owner, it is a channel that has existed quietly for years and is only now getting its own line in Search Console.

The important thing to understand is that a multimodal search is image first. There may be no typed keyword at all, or only a short refining phrase. That single fact shapes everything about how the new report works and what you can do with it.

Why Google added a multimodal filter to Search Console

Search Console has always let you filter your Performance data by search type. For years those types were web, image, video and news. As visual and AI-assisted search have grown, the plain web category has started to hide very different behaviours inside it. A traditional typed query and an image-led Lens search can both end on a normal web result, but they are not the same kind of visit, and lumping them together makes your data harder to read.

The new multimodal filter is Google's way of separating those two behaviours so you can see them apart. It sits inside the web search type and splits it into Text-based and Multimodal. It is part of a wider pattern in 2026 of Google giving site owners more visibility into where their traffic really comes from, as search stops being a single box you type into and becomes a set of very different surfaces. The multimodal filter is one of those windows. The generative AI report, which I will come to, is another and a more important one.

What the new multimodal filter actually shows

When you open the Performance report in Search Console, you can now filter your web results by Text-based or Multimodal. For the multimodal slice you get the same core metrics you are used to: impressions, clicks and average position. You can see how often your pages appeared in image-led searches, how often people clicked, and roughly where you ranked.

There is one important gap, and it is the thing to understand before you read too much into the report. You do not get the queries. Because the search started from an image, there is often no typed keyword to report, so the queries list is empty when you select the multimodal filter. You can see that visual search found you, and how often, but you cannot see what the image was or what the person was looking for. That makes the multimodal filter a measurement tool rather than an optimisation tool. It tells you the size of the channel, not the words to target, because there are no words to target.

You will still find it useful for a few things. You can see whether visual search is a meaningful part of your visibility or a rounding error. You can watch whether it grows over time. And you can compare it against your text-based numbers to keep it in proportion. What you cannot do is build a keyword strategy from it, because the raw material of a keyword strategy, the query, is not there.

How to read the multimodal data honestly

Here is the part most write-ups will skip, because it is less exciting than the headline. For most Wix sites, multimodal search is a small slice of traffic today. On my own site, image and visual search together make up around one percent of impressions. Unless you are in a strongly visual niche, you will probably see something similar.

That does not mean ignore it. It means keep it in proportion and know when it matters more. Multimodal and visual search are far more valuable to you if your business is inherently visual, for example:

• Products and ecommerce, where people photograph an item to find it or something like it
• Interiors, furniture and homeware, where people search from a picture of a room or a piece
• Food, recipes and hospitality, where a dish or a menu is photographed
• Art, design, crafts and anything bought by how it looks
• Local and travel, where people point a camera at a place

If that is you, visual search deserves real attention and the image SEO below is worth doing thoroughly. If you sell a service, or your value is in words and expertise rather than pictures, visual search will stay a minor channel and you should spend your effort elsewhere. The multimodal filter is what lets you make that judgement with your own numbers instead of a guess.

The image SEO that makes you findable in visual search

Whether visual search is one percent of your traffic or twenty, the way you improve it is the same, and it is ordinary good image SEO that you should be doing anyway. There is no separate trick for multimodal search. Google understands your images through the same signals it has always used, so make those signals clear.

Alt text. Every meaningful image should have real, descriptive alt text that says what the image actually shows. Alt text exists first for people using screen readers, and it also tells Google and visual search what the picture is. Write it as a plain, honest description, not a list of keywords. "Grey fabric three seat sofa in a living room" is useful. "sofa sofa cheap sofa buy sofa" is not, and it can do harm. Decorative images that carry no meaning can have empty alt text so they are skipped.

Descriptive file names. Name the file before you upload it. "grey-fabric-three-seat-sofa.jpg" tells Google something. "IMG_2931.jpg" tells it nothing. This is a small habit that quietly helps every image you publish.

<a href="/wix-schema-markup">Image structured data</a>. Where it fits, mark your images up so Google understands their role. Product images benefit from Product structured data, which can make them eligible for richer results. Standalone images can use ImageObject. If you publish images others might license, Google supports image licensing metadata that can add a licensable badge. Structured data does not force a result, but it removes doubt about what the image is.

Make images crawlable. Google can only understand images it can reach. Do not block your image files or your image folders in robots rules, and make sure images load properly rather than being hidden behind scripts that a crawler cannot run.

Context around the image. Google reads the words near an image to understand it, so a relevant caption, heading and surrounding paragraph all help. An image of a product sitting under a clear heading and a real description is far easier for Google to place than the same image dropped on a page with no supporting text.

File format and size. Use modern, efficient formats and compress your images so they load fast. This helps visual search indirectly by helping your Core Web Vitals and mobile speed, which affect your rankings everywhere, not just in images. A heavy, slow page is a weaker page in every search type.

Original images. Original photographs of your own products, work or premises are easier for Google to understand and more likely to stand out than generic stock images that appear on thousands of other sites. If you are serious about visual search, your own images are an asset worth investing in.

Doing image SEO on Wix

The good news for Wix users is that a lot of the technical side is handled for you, and the parts that are not are simple to control. Here is how the points above map onto Wix specifically.

Alt text on Wix. Click any image, open its settings, and fill in the field that asks what is in the image. That is your alt text. Do it for every meaningful image, in plain descriptive language. It takes seconds per image and most Wix sites I audit have it half done or left blank, so this alone is often an easy win.

File names on Wix. Rename your image files to something descriptive before you upload them through the Media Manager. Once an image is in Wix it is harder to change the underlying name, so the habit is to name it well on your computer first.

Image delivery on Wix. Wix serves your images through its own image system, which produces responsive, appropriately sized versions and modern formats for browsers that support them, and it lazy loads images below the fold. This means much of the file format and sizing work is done for you. Your job is to upload a good quality original and let Wix handle the delivery, rather than uploading something enormous or something tiny and blurry.

Structured data on Wix. For supported content types you can use the Advanced SEO Structured Data Markup field in the page SEO settings. For anything custom, including ImageObject or Product markup with images, you paste your JSON-LD, wrapped in a script tag, into the Custom Code area of the dashboard set to load in the head. This is the same dual-layer approach I use on every site, standard schema for Google plus an entity layer for AI answer engines.

Crawlability on Wix. Modern Wix sites output standard, crawlable HTML and generate a sitemap for you, so your images are discoverable by default. The thing to check is that you have not accidentally hidden important images or blocked anything in a custom robots rule.

Do those and your Wix site is set up properly for visual search. There is nothing exotic to add.

The report that actually matters more, the generative AI performance report

If you only take one thing from this article, take this. There is a second new report in Search Console that is far more important for most sites than the multimodal filter, and the two are easy to confuse because both are new and both are about how search is changing.

In June 2026 Google introduced a <a href="/wix-generative-engine-optimisation-services">generative AI performance report</a> in Search Console. It shows how your pages perform inside Google's AI answers. According to Google's own documentation, it currently covers two surfaces, AI Overviews and AI Mode, and Google says it expects that list to grow over time. It excludes Search Labs experiments.

Here is what it shows and how it works, straight from Google:

• It reports impressions, clicks and position for your pages inside those AI features.
• An impression is counted when a link to your site appears in a generative AI feature. If two results from your site appear in the same AI feature, they count as a single impression.
• You can group the data by Pages, Countries, Devices and Dates. There is no queries dimension, so like the multimodal filter it does not give you the exact search terms.
• It draws on the web search type and shares the same limits as the main Performance report, such as the row limit and time period.
• It is rolling out over time, so not every property has it yet. Sites that have opted out of AI features will not appear, and very recent data may show as dotted lines because it is still preliminary.

Why does this matter more than multimodal search? Because this is where search is genuinely shifting for the majority of sites. AI Overviews and AI Mode now sit at the top of a large and growing share of results, they answer more of the question before the user clicks, and they change how visible you are even when your ranking has not moved. For a service business, an expert, or any site whose value is in words and knowledge, being surfaced and cited inside AI answers is a far bigger opportunity, and a far bigger risk if you are absent, than image search will ever be. This report is the closest Google has come to letting you measure that. If you are going to spend time in Search Console this year, spend it here.

Multimodal search and AI search are not the same thing

It is worth being clear, because the two new features arrived close together and both get called the future of search. They are different.

Multimodal search is about the input. The search starts from an image rather than typed words. The multimodal filter measures that.

Generative AI search is about the output. Google answers with an AI generated summary that may cite and link to sources. The generative AI report measures that.

A search can be one, the other, both or neither. Keeping them straight stops you chasing the wrong thing. For most Wix sites, the input change, images, is minor, and the output change, AI answers, is major. Put your effort where the change is major.

What to do right now

Here is the honest, short version of what to actually do.

• Open the Performance report and note your Text-based versus Multimodal split, so you have a baseline you can compare against later.
• Tidy your image alt text and file names, and add image structured data where it fits, as normal good practice that helps visual search and accessibility together.
• Check whether the generative AI performance report is available on your property yet, and if it is, note your baseline there too, because that is the number worth watching.
• Put the real effort into being genuinely useful and first-hand, with clear, original content, because that is what earns you a place in both classic search and AI answers.

The multimodal filter is a useful new view and it is worth knowing about. For a small number of visual businesses it points at a real opportunity, and for the rest of us it is a glance, not a project. The bigger story in Search Console this year is the arrival of proper AI reporting, and that is where I would put your time. If you get your images tidy and your content genuinely helpful, you are set up for both, which is exactly where you want to be.