Google Data Shows AI Search Users Moved Over Keywords, Your Content Didn’t

0 0 3 minutes read

Google Data Shows AI Search Users Moved Over Keywords, Your Content Didn’t

In May, I wrote that the new search user was a trend and that Google’s product announcements were a distraction. Google has now provided us with a yearly amount of data that confirms the change in behavior and put some numbers on it. In the report “How People Are Using AI Mode in the US” (published on May 19, 2026, on The Keyword blog), Shivani Mohan, VP of Data Science & UXR, Google Search, describes the human search engine that many SEO teams have built their 2025 strategies around.

According to the report, a typical AI Mode query is now three times the length of a normal search query.

That single figure makes up a significant part of what many SEO teams prepared last summer. A year ago, the working assumption of most keyword strategies was that users would type in three to four words, then scan the results. Google’s data says that inference now explains a fraction of what AI Mode users actually do.

User Has Left, Content Has Not

The report covers the period from the introduction of AI Mode in the US in May 2025 to April 2026. A few other numbers make up the picture. Follow-up questions in AI mode grew by more than 40% on average per month, which means users don’t just drop one answer and walk away; they stay in the conversation and go deep. Multimodal interactions now include more than one in six AI-mode searches, meaning voice, image, or video input rather than typed text, and image input searches have increased over 40% in the month since launch.

The top five keywords in AI mode search are 1. Information, 2. Identify, 3. Find, 4. Explain, 5. Summarize. The top five opening words are “what,” “how,” “I,” “I,” and “knowing.” If we look at the third entry for “Me,” People are reporting personal context in the search bar. Not “running flat feet shoes.” Something along the lines of “I have flat feet and my knees hurt, can you help me find a running shoe that won’t make it worse?” An example of Health and Wellness in the report is less specific: “I hate cardio. Give me a routine that avoids it but still works.”

That is not a keyword. That’s someone talking to someone who might actually help them.

What the Content Gap Looks Like

The report organizes AI mode behavior into five categories: Explore, Decide, Learn, Create, and Act. Brainstorming-related questions grew 30% faster than the rest of the AI Mode questions. Programming inquiries increased rapidly by 80%. Inquiries starting with “where” have grown rapidly by 40% in the past six months, suggesting that AI Mode has become a real tool to support everyday shopping, not just a discovery layer.

This is a gap that many content strategies have yet to address. Content is designed for the typing user [best running shoes 2025] and getting to the listicle doesn’t help the user who asks, “I’m training for my first 5K and I’ve never bought running shoes before, what pair should I start with and how do I know they’re the right fit?” Both questions reveal the intention to buy shoes. Only one of them explains what the AI Mode user actually does.

The problem is that many groups still write short questions. They develop page titles, meta descriptions, and H2 structures for three to four keywords that represent a shrinking portion of how people actually find answers.

3 Things You Should Do Differently Now

Check your top 10 pages against how someone would ask for that information in a conversation. Take the main keyword of each page and rewrite it as natural language information the way a Google AI Mode user would type it. If your content doesn’t answer the long form of that question, there’s a gap your competitor will eventually fill.

Treat follow-up questions as a content signal, not a statistical footnote. A 40% monthly increase in follow-up questions tells you that users are not satisfied with one answer. If you know what the most common entry questions are for your site, the follow-up question is now more important as an entry point. For many sites, that follow-up questionnaire is not available yet.

Start preparing your visual assets for the multimodal index. One in six AI Mode queries is already non-text, and image input search is the fastest growing query type in the system. Different text written for accessibility and different text written to help a user who has captured a product and is asking AI Mode what it is and where to buy it are different things. The context of the image surrounding your product and knowledge assets needs to come where the questions are already there.

Google now has over 1 billion active users of AI Mode worldwide, and the platform’s query volume has doubled every quarter since its launch. The behavior change I wrote about in May is no longer predictable. It is a data set. The question for employees is not whether they can answer it but how quickly they can close the gap between the content they published last summer and the user searching right now.