One year after its U.S. launch, Google's AI Mode has crossed one billion monthly active users globally, with query volumes more than doubling every quarter since launch and search behavior shifting in ways that have direct implications for advertisers, publishers, and anyone trying to understand where digital audiences are going.

Google today published a detailed data report examining how people in the United States have used AI Mode in the twelve months since it launched. The data, released on May 19, 2026, by Shivani Mohan, Vice President of Data Science and UXR at Google Search, covers behavioral shifts across five functional areas: exploration, decision-making, learning, task completion, and content creation. The findings draw on internal Google Search data and Google Trends data collected between AI Mode's U.S. launch in May 2025 and April 2026.

The core numbers are striking. According to Google, AI Mode queries have more than doubled every quarter since launch. The average AI Mode search query is now triple the length of a traditional search query. More than one in six AI Mode searches in the U.S. are multimodal - meaning voice, image, or video rather than text alone. Image-based queries in AI Mode are growing by more than 40% month-over-month since launch.

Query length and conversational structure

The structural shift in how people phrase searches is perhaps the most consequential change for the marketing ecosystem. Traditional search was built around keywords. According to the AI Mode U.S. Insights report, people are no longer worrying about the "right way" to formulate their questions - they are just asking. The conversational design of AI Mode permits multi-turn exchanges, and the data shows that follow-up queries in AI Mode have increased by more than 40% on average per month in the U.S.

What does that look like in practice? The report identifies the top five first words in AI Mode queries: "what," "how," "I," "is," and "can." These are not the keywords of a search engine optimized for short, discrete lookups. They are the opening words of a conversation. The top five keywords embedded within AI Mode searches - "find," "information," "identify," "explain," and "summarize" - point in the same direction. Users are offloading cognitive tasks to the search interface itself, rather than using it as a pointer to pages where they will do the reading themselves.

The query fan-out architecture that powers AI Mode is central to this shift. As documented by PPC Land, Nick Fox, Google's SVP of Knowledge and Information, described how the system breaks a single user request into potentially hundreds or thousands of sub-queries that are issued simultaneously, then synthesized into a single response with links. A user asking for a restaurant recommendation, for example, receives an answer that has already cross-referenced multiple booking platforms - a task that would previously have required 20 minutes of manual effort. That structural change in how responses are assembled is also what makes the three-times-longer query length possible: users are no longer writing machine-readable keyword strings, they are speaking naturally.

Multimodal growth and non-text inputs

The shift away from text as the default input medium is accelerating faster than most search trend analyses have captured. According to Google, more than one in six searches in the U.S. now use voice or images, and image search in AI Mode is growing over 40% month-over-month. PPC Land reported a 65% year-over-year surge in visual searches in July 2025, which predated much of this AI Mode growth. The current data suggests that momentum has continued.

The modalities available in AI Mode now include voice, images, video, and what Google calls "Live" search - a free-flowing real-time camera and voice session. Each modality creates a different data signal. Image searches in particular have implications for product discovery: users photograph objects they want to identify, compare, or purchase, and the conversational interface allows them to refine those searches through dialogue. Google introduced visual search capabilities to AI Mode in September 2025, specifically addressing scenarios where verbal descriptions are inadequate - clothing styles, interior design aesthetics, physical products seen in the street.

For advertisers, the multimodal question is not just about creative formats. It is about where in the decision journey a search is occurring. A photograph taken in a store is a very different signal from a typed brand query; the intent is immediate and localized. That distinction matters for audience segmentation and bidding strategy.

The five areas of use: Explore, Decide, Learn, Do, Create

Google's report organizes behavioral data into five functional categories that together describe a search product operating at a different level of task complexity than traditional search.

Explore covers open-ended discovery. According to Google's Trends data, brainstorming queries in AI Mode have grown 30% faster than AI Mode queries overall since launch. Searches starting with "where to," "where should I," and "ideas for" are all growing. The top ten things people ask about when their queries start with "where to" or "where should I" include car repair, general shopping, streaming shows and movies, streaming live sports, finding product manuals, filing forms or applications, buying groceries, buying concert tickets, and vacation planning. The breadth of that list illustrates how AI Mode is absorbing query categories that were previously handled by dedicated vertical apps or websites.

Decide covers purchase and comparison behavior. According to the report, searches beginning with "which" have increased 40% faster than AI Mode queries overall in the past six months. The fastest-growing specific query forms in this category were "which of" and "which one" - suggesting comparative evaluation is a primary use case. Among shopping topics, follow-up questions within AI Mode most commonly involve electronics, books, movies and music, apparel, health and beauty, automotive, home and garden, grocery, home improvement, toys and games, and sports and outdoors. When users search for stores or shops in AI Mode follow-up conversations, "near me" tops the list of subjects, followed by replacement parts and car dealerships with financing. The top retail attributes people search for when shopping via AI Mode are price, location, color, brand, availability, size, material, style, type, and quality - in that order. Price ranking first will not surprise marketers, but location ranking second is notable, suggesting proximity remains a strong filter even in AI-mediated commerce.

Learn reflects a broadening of the educational use case. People are using AI Mode to build study guides, create quizzes, and explore professional credentials. The top subjects for quiz and study guide creation are math, Spanish, history, English, biology, chemistry, vocabulary, algebra, geometry, and nursing. For in-depth explanations, users are asking about verb conjugation, physics and space, JavaScript and frontend development, household repair topics, tabletop games, historical events, and appliance repair. The credential searches are also instructive: Security+, Black Belt, Network+, bar exam prep, real estate license, CPA, CDL, and Scrum master certification all appear in the top ten. These are not student searches. They are mid-career and professional development searches, a segment traditionally served by specialized platforms.

Do covers task completion. According to Google's Trends data, AI Mode queries related to planning have grown 80% faster than AI Mode queries overall in the past six months. That encompasses schedule creation, to-do lists, exercise plans, and restaurant discovery. Canvas, a persistent planning workspace within AI Mode, allows users to organize projects over time. The top Canvas use cases for schedules and itineraries include beach and island resort vacation plans, museum and historical tours, local scavenger hunts, national park and hiking itineraries, quick day trips, housewarming and dinner party plans, honeymoon and couples getaways, toddler and kid-friendly vacations, bachelorette and bachelor party itineraries, and theme park strategies. For fitness planning, the top Canvas subjects are core and ab routines, lower body workouts, marathon training schedules, daily walking and step goals, and running loops. Financial planning queries in Canvas include dividend and investment strategies, retirement and 401k planning, expense tracking, and standard household budgeting.

Create covers content generation. According to Google, image creation queries on AI Mode have more than tripled since the start of 2026. The top things users ask AI Mode to help create are photos, quizzes and tests, logos, stories and poems, code and programs, messages, lists, documents, notes and summaries, and cartoons. For editing, the list includes photos, documents, videos, messages, code, PDFs, sentences, audio, notes, and essays. The breadth of that list means AI Mode is competing with - or supplementing - a wide range of software tools, from image editors to word processors.

Advertising and marketing implications

The query behavior described in the report has direct consequences for how digital advertising works. PPC Land has tracked the commercial infrastructure being built around AI Mode since Google formally announced shopping ads for the surface in February 2026, when the feature already had more than 75 million daily active users. By the time of this report's publication, monthly users have surpassed one billion globally.

Longer queries carry more contextual information, which in principle allows ad matching to be more precise. But existing keyword-based campaign structures were not built to handle conversational queries that average three times the length of traditional searches. Google has since extended AI Max to standard Shopping campaigns, specifically addressing the mismatch between conventional product feed optimization and the conversational query patterns now dominant in AI Mode.

The top retail attributes data - price, location, color, brand, availability, size, material, style, type, quality - effectively describes what product data needs to be present and accurate in a Merchant Center feed for a product to be surfaced in conversational shopping responses. Conversational attributes, a new Merchant Center data type referenced in PPC Land's coverage of Google's UCP expansion, allow retailers to supply richer contextual descriptions optimized for the longer, more natural-language queries that now characterize AI Mode shopping.

The planning query growth - 80% faster than AI Mode overall - is particularly significant for local and service businesses. Schedules, restaurant recommendations, and logistics queries are the categories where agentic features have the deepest foothold. Google added AI Mode to Circle to Search in July 2025 and tested seamless transitions from AI Overviews into AI Mode in December 2025, each step reducing the friction between a user's initial query and a full AI-mediated session.

The geographic dimension also matters for global advertisers. Google expanded AI Mode to over 40 countries and territories in October 2025, then extended it to nearly 200 countries following the Google I/O 2026 announcements on May 19. The behavioral data in this report covers U.S. users specifically, and the patterns may differ significantly in other markets. India, for example, showed distinct multimodal preferences when AI Mode launched there in June 2025, driven in part by the country's already high Google Lens usage.

The one-year data snapshot released today does not resolve the question that most concerns publishers and content producers: what happens to referral traffic as more queries receive complete answers inside AI Mode rather than routing users to web pages. That question sits outside the scope of this particular report. What the data does show is that query volume is up, not down - according to Google, search is at an all-time high, with AI Mode's new features cited as the leading reason. Whether that volume translates into comparable referral traffic, or whether it increasingly represents queries resolved without a click, is a separate measurement problem that the marketing industry is actively working to answer.

Timeline

Summary

Who: Google, reporting on U.S. user behavior in AI Mode. The data was published by Shivani Mohan, Vice President of Data Science and UXR at Google Search, with the Insights report issued alongside statements from Nick Fox, SVP of Knowledge and Information, on LinkedIn.

What: A one-year behavioral data report showing that AI Mode has surpassed one billion monthly active users globally. Key metrics include queries more than doubling every quarter since launch, average query length tripling compared to traditional search, follow-up queries growing more than 40% per month on average, image searches growing over 40% month-over-month, planning queries growing 80% faster than AI Mode overall, and image creation queries more than tripling since the start of 2026.

When: The report was published on May 19, 2026, covering data from AI Mode's U.S. launch in May 2025 through April 2026. Keyword and query-specific figures are derived from Google Trends data, which tracks the percentage of searches for a topic as a proportion of all searches at a given time and location rather than absolute volume counts.

Where: The behavioral data covers AI Mode usage in the United States specifically. The one billion monthly active user figure is global. The feature is now available in nearly 200 countries and territories following the Google I/O 2026 announcements.

Why: The report documents a structural shift in how people use search - from short keyword lookups to multi-turn conversational sessions that include task completion, creation, planning, and decision support. For the marketing community, the data reframes how audiences interact with search-adjacent advertising surfaces, changes what product data needs to look like to be retrieved by AI-mediated queries, and signals that query volume growth is real and continuing - even as the question of how that volume converts to referral traffic for publishers and advertisers remains open.

Share this article
The link has been copied!