Google's Liz Reid on Who Will Own Search in a World of AI

Episode Summary

Executive Summary: Google Search VP Liz Reed argues AI is expanding search rather than replacing it: AI Overviews and AI Mode help users express real needs in natural language, surface better results, and bring more people into search more often. She says Google’s challenge is balancing usefulness, ads, ecosystem health, and quality amid AI-generated slop and changing user behavior.

Main Topics: AI as an expansion of search, not a replacement (Priority: 5/5): Reed says users want AI and the web together. AI Overviews can answer quick questions, while deeper or more subjective needs still require web results, expertise, and follow-up exploration. How Google decides when to show AI Overviews (Priority: 5/5): Google does not show AI just because a query has a question mark. It uses user signals and query intent to determine when AI adds value and when traditional search results are better. Different products for different intents (Priority: 4/5): Reed distinguishes between Google Search, AI Mode, and the Gemini app: Search is more informational and browsy, AI Mode handles longer conversational questions, and Gemini is more for creativity and productivity tasks. Monetization and the ad business (Priority: 5/5): The discussion explores whether AI reduces clicks and ad revenue. Reed argues many queries never had ads anyway, shopping queries still create commercial opportunities, and richer prompts may even improve ad targeting. Search quality, slop, and ranking (Priority: 4/5): Reed says AI slop is not new—human-generated slop already existed—and Google’s core ranking and spam-fighting systems remain central to surfacing trusted content over low-quality pages. Changing user behavior and query complexity (Priority: 4/5): As users become more comfortable with AI, they ask longer, more specific, natural-language questions rather than short keyword queries, which helps Google infer intent and deliver better answers. Future interfaces and form factors (Priority: 3/5): Reed is skeptical that one interface will replace all others. She thinks search, chat, apps, voice, desktop, mobile, and maybe glasses will coexist depending on task and context.

Key Arguments: AI Overviews are shown only when they add value; Google avoids forcing AI into every query. Users increasingly express their real intent in full natural language, which makes search more useful. Google Search and AI tools are complementary because quick facts, deep research, shopping, and human perspectives are different needs. The presence of AI can expand the total number of questions people ask, increasing usage rather than just cannibalizing clicks. Ad monetization is still viable because many AI-assisted queries are noncommercial, while shopping and decision-making queries remain monetizable. Google measures success by whether users return more often and whether the experience is genuinely helpful, not just by raw token-like usage or clicks. AI slop is a ranking and trust problem Google has long handled; the company already filters spam and low-quality pages at scale. The future likely involves multiple entry points and form factors rather than a single universal search box.

Data Points: Google Search ad coverage: less than a quarter of queries - Reed notes that only a subset of search queries ever show ads, so many informational queries are noncommercial anyway. Google tenure of Liz Reed: over 20 years - Reed describes her long career at Google and current leadership of Search. Search team scope: product, engineering, user design, and data science - Reed explains her role as VP of Search. AI product comparison: 3 surfaces - Reed distinguishes main Search, AI Mode, and the Gemini app. Ranking scale: many, many more pages crawled than indexed - Reed uses this to explain spam/slop filtering and quality control. Query behavior trend: meaningfully longer queries - Reed says AI Overviews are associated with longer, more natural-language queries. Podcasting analogy: people may spend a couple of seconds or a whole hour - Used to illustrate how some questions need quick answers while others require depth.

Pivotal Quotes: "I think what we see is that people want AI and the web together." — Liz Reed: Reed summarizes Google’s thesis on the relationship between AI features and traditional search. "We shouldn't give you AI for the sake of giving AI." — Liz Reed: Explaining when Google chooses to surface AI Overviews. "What we really mean is there's a whole bunch of questions people have... And AI lowers that barrier." — Liz Reed: On why AI is expansionary for search and can increase the amount of curiosity users pursue.

Implications: For users, search may become more conversational and useful; for Google, success depends on balancing convenience, trust, and ads. For the industry, AI likely reshapes search into a multi-surface ecosystem rather than a single box.