How people use AI Mode in the United States

Remember the good old days when SEO was all about cramming the perfect combination of keywords into a 700-word blog post? You’d sit there trying to think exactly like a robot to make sure a robot could find your content. Well, grab your tissues, because that era is officially clearing its desk.

According to a recent internal report from Google detailing the first full year of AI Mode in the United States, generative artificial intelligence has officially flipped the search landscape upside down. With over one billion monthly active users globally, Google’s Gemini-powered AI Mode isn’t just an experimental tab anymore, it is the new normal.

For eCommerce brands, digital marketers, and SEO specialists, the data reveals a massive shift in consumer behavior. People are no longer treating Google like a static directory; they are treating it like a personal assistant.

1. Queries are longer and louder

For decades, the golden rule of search was brevity. If you wanted to go to a romantic restaurant in Boston, you typed “romantic restaurant Boston.” Today, users are typing out full paragraphs without a single care about “formatting” their thoughts for an algorithm. The data highlights a profound shift in user formulation:

• The average AI Mode search query is now three times longer than a traditional keyword-based search.
• Users are treating the search bar as a back-and-forth dialogue, letting their thoughts flow naturally.
• The most frequent starting words for queries in AI Mode are “What,” “How,” and “Can”.
• When analyzing what users actually want to achieve, the top keywords driving these AI searches are “Find,” “Information,” “Identify,” “Explain,” and “Summarize”. 

If your SEO strategy relies entirely on exact-match short-tail keywords, you are optimizing for a version of the internet that is fading fast. Content must now answer complex, highly specific situational questions.

2. Search gets complicated

Consumers are no longer satisfied with a single “search and click” action. They are digging deeper, treating Google as a collaborative partner to refine their intent.

Monthly follow-up queries within AI Mode have increased by an average of over 40% in the US alone. Users are starting a search, getting an answer, and then asking Google to filter, modify, or expand on that exact result.

More than 1 in 6 searches in AI Mode are now multimodal (non-text inputs). In fact, image-based queries are one of the fastest-growing sectors, skyrocketing by more than 40% month-over-month since the feature rolled out. If a user can take a photo of a broken kitchen appliance part, upload it, and ask, “Where can I find a replacement for this specific piece near me?”—and get an immediate, accurate local eCommerce recommendation, the traditional search funnel is effectively bypassed.

3. Deciding and shopping in the AI funnel

For retail and eComemrce players, the “decide” phase is where the money is. Google’s data indicates that users frequently start a shopping journey via traditional search and then jump into AI Mode to execute deep-dive comparison work. Interestingly, searches starting with “which” grew 40% faster than overall AI Mode queries over a six-month period, suggesting that people are increasingly using AI Mode to choose between options.

The top categories where users are actively asking follow-up questions to make a final purchase decision include:

1. Electronics
2. Books, movies or music
3. Apparel (being clothing the top most searched-for category)
4. Health / beauty
5. Automotive

When navigating these product categories, consumers are asking the AI to evaluate very specific operational attributes. The top metrics driving these conversational product evaluations are price, location, color, brand, and availability. If your structured product data (Schema markup) isn’t pristine, the AI won’t be able to pull your inventory into these hyper-customized consumer comparisons.

4. Planning, doing, and the rise of Canvas

Google is heavily leaning into “agentic” tools, features that don’t just display information but actively help remove the friction of executing a task. Queries related to planning grew 80% faster than the platform average. Through features like Canvas, which gives users a workspace to organize long-term projects, consumers are generating highly personalized routines:

• Fitness: Building marathon training schedules, core routines, and step goals.
• Finance: Creating 401k strategies, dividend trackers, and household budgeting spreadsheets.
• Travel: Mapping out intricate vacation plans (with Hawaii, Tokyo, and Italy topping the custom itinerary lists).

Furthermore, when users use AI Mode to find brick-and-mortar businesses or local services, they are looking for specific experiential attributes. For restaurants, the top follow-up queries focus heavily on whether a venue is Kid/Family friendly, has a specific view, features a bar, or accommodates vegan/vegetarian diets.

Changing the definition of “searchable”

As Google notes in its report, AI hasn’t just changed the mechanism of search; it has expanded the definition of what can be searched in the first place.

Users are using the tool for massive creative generation (image creation queries have tripled since the start of the year) and using it to build everything from AP Biology study quizzes to highly specific Python programming guides.

The internet is shifting away from indexing words and moving toward understanding context, intent, and conversational flow. Brands that survive this transition will be those that stop writing for search engines and start optimization for real, messy, complex human conversations.

Image: ChatGPT Images 2.0