Google AI Search: Why Full SERPs Win for Browsy Queries

When Google’s Liz Reid recently sat down to discuss the mechanics of AI Search, she introduced a concept that should reshape how SEOs think about user intent: Browsy Queries. Her insights, drawn from real behavioral data, reveal that search is far from a one-size-fits-all experience. Instead, users fragment their needs across multiple surfaces, and understanding these patterns is now critical for anyone hoping to win visibility in AI-driven results.

Search behavior is varied, not monolithic. Reid, Google’s head of Search, clarified that the ecosystem is more nuanced than a simple choice between classic search and AI. She groups traditional search and AI Mode together under the umbrella of “Search,” while positioning Gemini as a fundamentally different tool. “There’s sort of your main search page. There’s AI Mode. That’s part of search. And then there’s the Gemini app,” she explained. “There’s a lot of users, so their behavior varies across all of them.” This means the SEO community must stop treating Google as a single channel and instead recognize a complex search ecosystem where users co-use platforms based on task.

The patterns Reid described are telling. For informational queries, users gravitate toward Search or AI Mode. For creative or productivity tasks,like rewriting text to sound more formal,Gemini dominates. But the most revealing distinction lies between AI Mode and the classic search engine results page (SERP). Reid noted that users who go directly to AI Mode tend to have complex, long-tail questions where they expect to ask follow-ups. In contrast, for Browsy Queries, users “might choose to prefer all of the SERP.” This is the key insight: when someone is in a discovery or browsing mindset, they want the full richness of a traditional results page, not a single synthesized answer.

So what exactly are Browsy Queries? The phrase appears in Google’s internal vocabulary and job descriptions, and it consistently points to a discovery-level intent stage. A former DeepMind engineer described building a model to identify “browse intention” queries, which improved click-through rates by 5% for searches like “best places to visit in Orlando.” A Google commerce job posting mentions “browsy queries” in the context of shopping, where users haven’t yet narrowed down their needs. And a Google support page for video ads targets these queries as “lower intent, more ‘browsy’ Search placements earlier in their shopping journey.” Across all examples, the common thread is that the user is exploring, not deciding. For SEOs and merchants, this means optimizing for broad keyword phrases and creating content that guides users from inspiration to specificity, much like a pyramid where deeper pages offer more detail.

Reid also addressed how AI Search changes keyword fragmentation. Users have always had complex needs,like finding a vegan-friendly restaurant for five people in New York,but were forced to simplify them into keyword-ese like “restaurants New York.” With AI Overviews and AI Mode, people can now express their real need in natural language. “They don’t take their need and translate it to what the computer understands,” Reid said. “They try to give the computer their actual need and expect us to do the translation.” This shift has profound implications. A single complex query may not be solved by one web page, and many such queries are one-offs, reducing the ROI of optimizing for them directly. Instead, Google breaks down these long-tail phrases into smaller, highly specific queries,a process called query fan-out,and fires them to classic search. The AI then picks from the top results to synthesize an answer. This means SEOs should still focus on specific, well-optimized pages rather than chasing rare long-tail phrases.

The quality challenge is real. Reid explained that when queries become more diverse, Google can’t cache results as efficiently. “You have to take this question, there’s many parts, and you have to figure out how you break it apart,” she said. “If everyone uses the same keyword and it’s not personalized, then you can cache it all. If all of a sudden the queries get much more diverse, it has consequences there.” For SEOs, this underscores the importance of addressing real user needs rather than just targeting keywords. When auditing a page, ask: “What need is this filling?” and “How is this different and better?” Technical fixes are often less important than content that genuinely helps users solve their problems.

Ultimately, Reid’s message is clear: full SERPs still win for browsy queries, and the path to AI visibility lies in understanding the varied ways users search. By embracing the complexity of co-use, optimizing for discovery-level intent, and focusing on genuine needs, SEOs can navigate this new landscape effectively.