Google’s Liz Reid on AI search, query shifts, and AI slop

Google’s vice president of Search, Liz Reid, is pushing back against the narrative that artificial intelligence is killing the search experience. In a recent interview on the Bloomberg podcast Odd Lots, Reid argued that AI is actually reshaping how people search and driving them to do it more often. Rather than replacing traditional web results, AI Overviews are helping users bypass low-value clicks while increasing overall search volume, according to Reid.

When asked about the impact of AI on click-through rates, Reid explained that the technology primarily reduces what she called “bounce clicks.” These are instances where a user lands on a page, grabs a quick fact, and immediately leaves. “If all you were going to do was go to the webpage, see the fact, and immediately click back, you’re going to spend like a half a second on the page,” she said. However, for users who intend to read a full article for five minutes, AI Overviews can actually guide them to the right page, reducing frustration and improving satisfaction.

Reid dismissed the idea that users must choose between AI and the web. Instead, she described a complementary relationship. “I think there’s this sort of myth that people want AI or the web,” she said. “I actually think what we see is that people want AI on the web together.” While AI excels at providing quick answers, users still turn to the web for deeper exploration, including opinions and unique human perspectives. “There’s an opportunity to help you get started and then make it easy for you to dig in,” Reid added.

Not every query triggers an AI Overview. Google uses a variety of signals to determine when AI adds value. “An important premise of this is that we shouldn’t give you AI for the sake of giving AI,” Reid said. “We don’t want to put an AI Overview if we think it’s not going to be high quality.” As models improve, Google can cover more cases, but the focus remains on delivering the best response for each question.

The shift to AI is also changing how people formulate queries. Reid noted that users are now writing longer, more natural language queries instead of relying on keywords. “People stop talking just in keywordese as much, and they start expressing more of what they want,” she said. This allows Google to better understand the user’s real problem rather than forcing them to translate their need into computer-friendly terms. “That becomes much easier for us to give an answer,” Reid explained, calling the trend “really exciting.”

On the advertising front, Reid emphasized that Google can still monetize search even with AI answers. She pointed out that ads currently appear on less than a quarter of all queries, and many AI Overview queries were never commercial in nature. However, when users want to buy something, clicks still matter. “The answer doesn’t buy the pair of shoes; you actually have to buy the shoes,” she said. AI could also improve ad targeting by making queries more detailed, and as search expands, so do commercial opportunities.

One key metric Google watches is whether users return to Search more often. “It’s another thing to get you to decide you’re going to bother to unlock your phone,” Reid said. The goal is not just more searches but more frequent engagement.

Google is also keeping its tools distinct. Search and AI Mode are primarily used for informational queries, while Gemini handles creative and writing tasks. AI Mode, in particular, is designed for longer, more complex, and conversational questions. “There’s plenty of people who co-use across them,” Reid noted.

Finally, Reid addressed concerns about “AI slop” flooding the web. She argued that low-quality content has always existed. “Before AI slop, there was slop. There was human-generated slop. Now there’s AI-generated slop,” she said. The real challenge, according to Reid, is not eliminating slop entirely but ensuring Google can surface great content while keeping spam at a “very low rate.” She acknowledged the financial incentives behind spam but stressed that user trust depends on Google’s ability to deliver high-quality information.