The Financial Times has a detailed interview with the newish Head of Google Search, Elizabeth Reid. The interview can be gated, so here are some of the quotes that stood out most to me from this interview.
Reid joined Google in 2003 as one of the first of 10 or so engineers at the Google NYC office, there were 500 to 1,000 employees in total at Google by then. She first working on local related search topics and then moved to Google Search a few years later.
Related to AI, she said, "We see some of the strongest growth in [Google] Search and people issuing more queries." "Besides seeing people ask more questions, they ask longer questions," she added.
It appeals a lot to younger searchers, she said. "We see it resonate in particular with younger users. They are often the first to push expectations about what should be possible and to adapt to new technology. More and longer questions. They start asking more nuanced questions."
They are also seeing a lot more multimodal searching, "We see a lot of growth in multi-modality: people asking these text-plus-image questions."
AI Search is more designed around "how can you further your journey without repeating it the same way you might to a human — rather than designing it in the sense of: do you have a friend to chat with and ask them their views? It’s much more about organising information."
On AI making mistakes, "It is the case with generative AI that the technology sometimes makes mistakes. We saw, with eating rocks, that it was an extremely small-use case. Despite our extensive work and testing, it was not the type of query we had seen previously." But Google is working on improving and said, "we have weighted factuality and put extensive work into that. We have continued to raise the bar on that for the past several months." "Our models are trained not just to try and be highly accurate, but to try and base their answers on information on the web," she added.
Here is an interesting quote around AI Answers not being designed to be the end of your journey. "AI Overviews aren’t designed to be a standalone product. They are designed to get you started and then help you dive deeper. And so, when it’s important, the idea is that you get some context on where to check and then you can choose to double-check more on some of them." But I am not sure about this...
Then we get this typical line from Google; "We see the clicks are of higher quality, because they’re not clicking on a webpage, realising it wasn’t what they want and immediately bailing. So, they spend more time on those sites. We see that it shows a greater diversity of websites that come up."
FT then continued to push on about traffic issues from Google to publishers. Reid responded:
We do believe, in [Google] Search, that people continuing to hear from other people is essential and at the heart of our product. That’s important, not just for a healthy ecosystem, but for users. Lots of times you want a quick answer, but often you want to hear from other people.I often use a fashion example: most of the people I know who want to delegate their choices to a bot for fashion are the set of people who weren’t trying to spend any time on fashion before.
The people who are following influencers and creators and others, they’re not ready to go there. They want to hear from the people they trust. So, we spend a lot of time thinking about, how do we elevate the right content? How do we present it? We run different experiments. We design it to not just show links, but think about where it could add additional links within the response. Not just at the end, but maybe we can say, “according to the Financial Times” and put a link to the Financial Times.
What you see with something like AI Overviews, when you bring the friction down for users, is people search more and that opens up new opportunities for websites, for creators, for publishers to access. And they get higher-quality clicks.
Then on Ads:
There are a lot of opportunities for ads. We show them both above and below in AI Overviews, but also within. Ads are relevant whenever users are going to make a choice that has some commercial aspect.When a query is predominantly commercial intent — like we think you want to buy something — then we might often show ads. But sometimes we think you probably don’t want to [see] ads, and so we don’t want to give everyone ads. But some people might want to buy something. If [you search] “how to clean a stain out of the couch” and the first thing we show is a bunch of ads, you’re like, “Whoa, I just wanted some advice.”
But if we’re giving you ideas and then we say, “if you’re having trouble you might want to consider a stain-remover product”, and then we give you some ads for stain-remover products, it feels natural and in context. And so, there are new opportunities.
How will Google Search change in the future:
- More multimodal searching
- It will get more personalised over time, not just in the results, but in how you learn well.
- Agents? "This question is about how you make use of tools. People use the word “agents” to mean different things. But the sense of “you can use tools to ask hard questions” will continue. [Google] Search will remain an information product at heart, but sometimes information is hard and there’s a lot of work."
Forum discussion at X.
