The era of the long-form narrative arc designed to keep users scrolling is colliding with a new search reality. Google AI Overviews and AI Mode now compress multi-source information into immediate factual summaries, and the data proves this is not a passing trend. AI Mode has surpassed 1 billion monthly active users globally, with queries more than doubling every quarter since launch. The average AI Mode query in the United States is now triple the length of a traditional search query, and more than one in six searches include an image, voice input, or a live conversational exchange. For SEO practitioners and content architects, the message is unambiguous: burying the primary answer beneath brand storytelling or poetic preamble makes it harder for generative systems to surface your page. The winning strategy is to lead with the answer.
AI Mode Queries Surge 3X: What the New Search Behavior Reveals
The volume numbers are striking, but the structural transformation of the query itself carries far greater strategic weight. According to data published by Shivani Mohan, Google’s vice president of Data Science and UXR, on the Google blog in May 2026, users are not simply searching more frequently; they are searching in a fundamentally different way. The one-shot keyword fragment that defined search engine optimization for two decades is giving way to multi-turn, conversational, and multimodal interactions.
Image-based queries within AI Mode are growing more than 40 percent month over month. Follow-up questions are climbing at the same rate. Users start a search, refine their intent, and continue the conversation in a loop that bears no resemblance to the static keyword model. Google’s internal analysis of the most common first words in AI Mode queries reveals a clear pattern: “What,” “how,” “I,” “is,” and “can” now dominate, replacing the short-head nouns that once ruled search boxes. People are asking questions the way they would ask another person, not typing fragments the way they would type into a legacy search bar.
Five Verbs That Replaced the Keyword
Mohan’s report organizes the new behavior into five distinct modes that correspond to different user intents. These are not keyword categories; they are task categories. Understanding them is essential for anyone who wants to remain visible in AI-driven search results.
Explore queries represent open-ended, brainstorming-style searches. They are growing 30 percent faster than AI Mode traffic overall. Decide queries, built around comparative language such as “which of” and “which one,” are growing 40 percent faster than the baseline. Learn queries focus on understanding new concepts or exploring professional development opportunities. Do queries, tied to planning activities like workout routines, travel itineraries, and household budgets, are growing 80 percent faster than overall AI Mode queries over the past six months. Create queries, which involve generating images or other media within AI Mode, have more than tripled since the start of the year.
None of these modes maps cleanly onto a single keyword. Each maps onto a task. That distinction marks a turning point. The industry is not navigating a routine update to ranking factors; it is navigating a behavioral transformation that demands a complete rethinking of content architecture.
What Are the Five New Query Modes in Google AI Mode?
The five modes are Explore, Decide, Learn, Create, and Do. Explore queries are open-ended and brainstorming-oriented. Decide queries use comparative language to choose between options. Learn queries help users understand new concepts. Do queries focus on planning and execution. Create queries involve generating images or media. Each mode requires a distinct content structure, and the fastest-growing categories are Decide and Do, which together indicate that users want immediate, actionable answers rather than general information.
The Telegraph Lesson: History Repeats Itself in Search
Journalists in the nineteenth century faced a similar disruption. The invention of the telegraph imposed mechanical and economic constraints on how news was written. The Associated Press paid operators by the word, which made verbose introductions an expensive luxury. Newspaper compositors setting stories in heavy metal type needed a reliable method for cutting content from the bottom up to fit physical page constraints without losing critical facts. The solution was the inverted pyramid: lead with the most important information, then layer supporting details in descending order of relevance. This structure became the industry standard between 1880 and 1890.
Generative search engines now operate under a similar set of computational and mathematical constraints. Large language models assemble responses by extracting the most relevant passages from source content. If a page buries its primary answer inside a meandering introduction or a brand voice exercise, the system either skips it entirely or produces a hallucinated summary that misrepresents the original. The parallel is exact: telegraph economics forced journalists to lead with the news; AI Mode economics force content creators to lead with the answer.
A personal experience from 1986 underscores the lesson. As the newly appointed director of corporate communications at Lotus, I placed a thick binder of press clippings on the desk of CEO Jim Manzi. He glanced at it and said he did not want to see my reports until I could measure the impact of public relations in cold, hard cash. That demand sent me down a forty-year path into measurement and accountability. Generative search now makes the same demand of every writer on the web: prove your value immediately, or be skipped. Protecting a vague brand voice by hiding core facts inside fluffy introductions is a fast track to search irrelevance.
Practical Steps for AI-Driven Search Optimization
The behavioral data from Google, combined with the historical precedent of the telegraph, points to a clear set of actionable strategies. These five steps apply directly to content created today.
1. Lead With Entity-Dense Opening Sentences
The first sentence of every major section should deliver the primary definition, key metric, or main conclusion. Anchor that sentence with specific brand names, geographic markers, exact dates, and verified numerical values. Vague generalizations reduce the mathematical readability of prose and increase the risk that AI systems will hallucinate or misattribute information. Specificity is the new currency of search visibility.
2. Implement Scannable Hybrid Layouts
Body copy should be organized into short paragraphs of two or three sentences. Major section headers should be followed immediately by concise summaries, ordered lists, or structured tables. This dual-purpose layout allows search crawlers to extract clean segments for generated overviews while giving human readers a frictionless scanning experience. The same structure serves both the algorithm and the person.
3. Write for the Follow-Up, Not Just the First Click
With follow-up questions and multi-turn conversations growing at more than 40 percent month over month, content must anticipate the second and third questions a reader will ask after the one the title promised. Each section of a long-form article should be able to stand on its own as the answer to a specific follow-up query. AI Mode is most likely to surface those standalone chunks, not the entire page.
4. Build Content Around the Five Verbs, Not Five Keywords
Before writing a piece, determine whether it is designed to help someone explore, decide, learn, create, or do something. A comparison page written for “which” queries requires a fundamentally different structure than a guide written for “how to get started” queries, even when the topic and the target keyword appear similar. The verb determines the architecture.
5. Treat Image and Multimodal Content as a Ranking Input
Image queries are growing 40 percent month over month, and AI Mode integration with tools like Nano Banana has driven a tripling of image-creation queries since January. Alt text, image context, and visual content quality are no longer secondary to the words on the page. They are part of what AI Mode reads when assembling a response. Every image must be treated as a ranking input, not an afterthought.
Structural Clarity Over Brand Storytelling
Four decades in measurement and content strategy have led to a single conclusion: success in the current search era is not about flooding the web with generic automated drafts. It is about mastering structural clarity so that both search algorithms and human readers receive immediate, verified utility. The inverted pyramid worked for journalists because it respected the constraints of the telegraph and the printing press. A similar structural discipline now works for SEO because it respects the constraints of generative search synthesis.
The temptation to protect brand voice by delaying the answer is strong, but the data is definitive. AI Mode queries are longer, more conversational, and more multimodal than any previous search format. Users are asking questions the way they would ask a knowledgeable colleague, and they expect an answer in the first sentence. Content that delivers that answer clearly, specifically, and immediately will be the content that survives the next wave of search evolution.
