Google’s Liz Reid explained on the Bloomberg Odd Lots podcast that AI Mode and AI Overview enable detailed, demand-based query patterns that create new challenges for Google. This points to a resulting change in search behavior that directly affects how to approach SEO.
Keyword segmentation in AI Search
Liz Reid explained that users have been looking for longer natural language queries but were forced to narrow them down to keywords like “best restaurants in New York” even though what they were really looking for was more specific like a restaurant with vegan options and an opening for a party of five.
I’ve been in SEO for a long time, and with close to 30 years in the business, keyword research has become the foundation of digital marketing. You choose the keywords you want to rank for and create content optimized for that keyword. The problem with short keyword optimization is that there are hidden meanings within that keyword and always have been.
One way Google has tackled the issue of hidden meanings within keywords is to use things like clicks to better understand what users meant when they typed in vague keyword phrases like “restaurants in New York.” Some SEOs believe that clicks are used to rank websites but another use of clicks is to understand what people mean when they type in obscure phrases. What Google has done for a while now is put the most popular definition of a keyword first and regardless of how many links a page has received, if the content matches the less popular definition the page will not rank.
Liz Reid said that people using AI-based search are using long queries that explain what the problem or information is, making it easier for Google to retrieve the information they are looking for. That change gets to the heart of the problem with natural search being solved by search AI and the implications for SEO are profound.
Liz Reid begins:
“We’ve seen with AI questions that are reasonably short. We’re seeing a lot of natural language questions, but it’s also not as basic as that.
It may also look like you were looking for restaurants. We used to laugh like this before I worked in search, working in maps and location, one of the areas that intersects with search, and people were just like “New York restaurants.”
And he’s like, what do you want me to do with that question? Like, okay, the best restaurants in New York are going to take three months and 99.9% of the people won’t be able to go to them.
Ok, but do you like, do you choose random, etc.?
But like, part of why people would do that is because they were so complex- I want a restaurant in this place for five people. It won’t be too expensive. I have a vegan membership. I have children too. That was the question they had in their minds.
And in the old world of keyword-ese, that information can be spread across the web. And so you wouldn’t feel confident enough to just pop the question.
And now with AI Overviews and AI Mode, you can start doing, and you see people doing this, they tell you the real problem, right?
They don’t take their need and translate it into what the computer understands. They try to give the computer their real need and expect us to do the translation.”
Great ideas to take out there are:
- A typical complex query asked in AI Search may not be solved by a single web page.
- Complex questions can be one-off and rarely, if ever, repeated, which in many cases can reduce the value of developing those phrases, because the time spent formulating them could be better spent doing something else.
- Given that the site will likely share AI Overviews (AIO) space with another site, it increases the need to improve other features such as positive branding, use of appropriate images, and use of videos to claim as much AIO space as possible.
- However, perhaps the biggest takeaway is that it’s not all long tail because Google breaks down longtail phrases into smaller keyword phrases that reflect part of the need for information, fan-out queries, and banish those from classic searches. Google’s AI then selects the top three for each question and uses that to compile an answer.
So it’s not really that SEOs have to prepare long-tail queries because the query followers use General Search, which returns everything to the specific queries that web pages are important and optimized for.
Addressing Real Needs
Reid didn’t go into detail on this point but it’s interesting nonetheless because he said the process of breaking down a complex natural language query into smaller queries becomes a quality problem. One of the problems with AI search is that people don’t search for the same keywords which means Google can’t store the same queries in the same way as with natural search.
He explained:
“I think it means you have to do it, it’s quality work, isn’t it?
You have to take this question, there are many parts, and you have to figure out how to separate it. And you have to do the work to think about things like latency, because you just can’t, you know, if everybody’s using the same keyword and it’s not personalized, you can save everything. If suddenly the questions become more diverse, you know, there are consequences.
But I think we can just see that it empowers people, right? That it takes some work from the search.
A few years ago, they said, What else can you do with Google search? But if you really ask them, OK, when was the last time you spent 20 minutes searching when you’d rather spend 2? It’s actually not that hard for me. … So it’s been an interesting thing to make people’s lives easier by helping them face their real need.”
From the outside, the idea of handling the real needs of the user sounds like one of those slogans that don’t help “be good” or “content is king”. But it’s actually a method that every SEO should check web pages. Instead of limiting its scope to keywords, topics, technical issues, look at how it fills a specific type of need.
Someone today asked me to take a look at their website that was having trouble getting listed. They suspected it might be a technical problem. My answer is yes, everyone hopes it’s a technical issue but in many cases, especially the one I was looking at, the problem becomes clear when viewed through the lens of asking, “What is the need to fill this page?” and asking, “How is this not just different from the other page but different and better?“
Watch Liz Reid’s interview here:
Google’s Liz Reid on Who Will Own Search in an AI World
Featured image by Shutterstock/TierneyMJ
