Pay attention to shifts in search intent – with Mark Williams-Cook

I'm Mark Williams-Cook, and this is SEO in 2025.

Mark, what's your number one SEO tip for 2025?

My SEO tip is that SEOs should be monitoring search intent shift. So being aware of when what users want, what they mean by a search term changes over time and adapting accordingly.

Okay. Monitor search intent shift.

Love it.

Okay. So adapting to what users actually want, how do you find out what users want?

Yeah, right. This is, that's the million-dollar question. So I think it's a fascinating topic because this happens in several different ways when we're talking about intent shift. And some, I think people are already aware of. And others are kind of a bit harder to realize that it's happening in the background. So to take a really, I'd say, quite obvious example would be looking at seasonal intent shift. Firstly, I think it's helpful to have in your mind, I like to think of it like a search term is like a spotlight. And then we've got this universe of kind of different questions, which is what people want to find out. And That one search term kind of moves slowly, moves around that as the intent shifts. So take an example of Halloween. If you just do some searches around Halloween at most times in the year, you actually get a lot of informational sites giving you information about the origin of Halloween and how it, you know, and the background and what it's about. When you start coming up to October, and obviously you can see searches for Halloween spike, What we tend to find is that the type of results we get around Halloween searches are a lot more commercial in nature. So they're showing e-commerce sites that sell Halloween costumes, candies, that kind of thing for Halloween. And that's because Google has monitored, as we very much know, now know after the DOJ trial, how users are interacting with the SERP. And they do so in a predictable fashion and that intent of what they actually want. On average changes. So that's one kind of clear example that we can expect. The other way that I monitor search intent shift is with the, and you've heard me talk about this before, is the People Also Ask data. So People Also Ask data gives you what I call an intent proximity, which is when someone searches for a phrase or a question or just a, even a single keyword. The people also ask questions give us insight into what Google thinks is going to be helpful as a follow-up question. Okay. And the reason that's so important is we know one of the signals that Google uses on the SERP to monitor the satisfaction of the user is things like if they go back and redefine their query. So if someone generally does a search and then goes back and actually redefines their query. Google can get a hint of maybe we weren't showing quite the right thing the first time around. And the People Also Ask data that Google shows you gives you a hint to this. A really good example that I've seen over the last few years is searches actually around COVID. So as part of our agency, we monitor how the People Also Ask questions change over time. So we're scraping them and storing them. and seeing what Google's suggesting. And if you look at searches or the questions that people also ask, questions that are generated around COVID 2 years ago, there are questions that are essentially these kinds of panics, like, what do I do if I have COVID? And what are the symptoms? And what is the law? Like, am I allowed outside? Fast forward 2 years, the questions now have very much come around to People understand what COVID is, and it's now more like, what's the current law? Can I go outside? So the intent has shifted very much from, please tell me what this is and how do I know if I've got it and what on earth am I allowed to do, to how do I manage my symptoms and basically get on with my life? So that, to me, is a big signal across any industry you're working in. If you see these changes, that you need to update your content to match the current Intent.

I'm still thinking about your spotlight phrase. I like that. I'm imagining this chap dancing across the stage and the background changing and the background dancers or other participants changing, but spotlights on the keyword phrase, but the relevance, the relatedness with other things that are going on changes over time.

Yeah, I probably shouldn't have given people kind of an insight into the The weird theatre that's going on in my mind. But yeah, so it's kind of like that spotlight is going over maybe like the different people also ask questions, you know, over time. So, with COVID it's kind of just shifting in one direction, right? Because our collective knowledge about that thing is changing. So, that spotlight's like creeping this way. Whereas with Halloween, it's more of kind of, it's oscillating between, okay, I'm here and then I go here and then I'm back. And I'm here and I'm back. There are other ways as well this changes. So this would, to me, was absolutely fascinating. In the DOJ trial, Google talked specifically around a system they called Instant Glue, which is apparently a version of NavBoost. Yes, exactly, which is looking at how and remembering how users are clicking and adjusting SERPs. And They gave a really brutal example in the DOJ trial, which was, if someone is doing a search for nice pictures for, say, their PowerPoint, you know, Google's like, okay, yeah, here are some nice pictures. And then for, bizarrely, they gave the example, say there was a terrorist attack in Nice. And then people search for Nice pictures, which is obviously spelled exactly the same, they would realize that how people are interacting with the SERP has completely changed. And they may be clicking on news websites, or they're clicking on pictures of the place Nice rather than nice pictures. And that's how they readjust their, their SERP. Okay. Another obvious example was the assassination attempt on Donald Trump, completely changed the search intent around, you know, Donald Trump and the type of content people were showing because obviously people were interested. In that story. So while there's not necessarily anything, well, there are a few things, but there's not lots of things lots of people can do to prepare for that, because you can't tell when that shift is going to happen. It's useful knowing from an SEO point of view, those mechanisms exist. And Google is thinking about that, you know, news, newspapers will write stories about celebrities dying ahead of time. And that's all, you know, rather morbidly ready to go. Because they understand that search intent is just going to immediately change. Again, it's why I focused a lot of my time and research on the People Also Ask data, because that is the data that updates the fastest that I've seen that we get from Google. When something happens in the news, the PAA questions will change within an hour or two. It also tells us that We know SERP features are generated post-ranking. So Google has their big churning kind of system where they're deciding where web pages rank. And then based off of that, and maybe metrics they pump out after that, the SERP features are generated. And that's how they can change so quickly. Because there's no way to go into, you know, the guts of Google and the ranking algorithm and have things change. And I think that itself is a really important thing to understand. Because for years, we've, we've had Googlers tell us about, you know, we don't use clicks directly in ranking. And I think that's what they mean in terms of, in my opinion, Google does not use clicks as a quality indicator for individual websites. They're doing, in my opinion, 2 things with them, which is one, They're building click models. So I think Google has, when they have a SERP, they have a predicted click-through score for those SERPs that is based on aggregated click-through data that they have. So it's not used individually on any one particular site. It doesn't matter if your site gets clicked on or not. The other end is they use click data post-ranking. But the important thing I think for SEOs to understand, and it fits in with this intent shift thing, is That's like a query-based thing. That's Google adjusting SERPs based on intent shift around the query. It's not to do with an individual site, how good it is and where it ranks. It's got to do with, from Google's point of view, yes, your site is a very attractive target for this keyword. And then maybe if Google gets data from the user, they that they're after something else. It's not that your site's any better or worse. It's that the user wants to aim at a different target. They're not interested in what you're saying anymore. And I think that's missed by many SEOs because we get this blinkered view that a key phrase is a static target. And why don't I rank for it? I did this content, you know, and now I don't. And it's actually more to do with Google's understanding of the query and the user has shifted. And you haven't shifted along with it.

Oscillates is another nice word that you used there. How should an SEO strategy oscillate in order to actually stay on top of the optimum focus in terms of keyword targeting, content targeting?

Yeah. So the other details we got from like systems like Instant Glue, were, uh, so to remind everyone, Instant Glue is the system that Google was using to, uh, look at how people were clicking on results and changing them, was they specifically said one of kind of the issues is that old sites, old content gets more attention because it's kind of there, it's embedded, it's grown, people know about it, people have linked to it. So part of the function of Instant Glue, to paraphrase again what they said in the DOJ trial, which Which was to kind of artificially give more attention to new things. And I think that's critically important because, and this has long been a theory with different SEOs, which is that there's almost like a trial period for sometimes for new content that you publish where you publish something and it kind of seems to rank for a little bit. And then sometimes it drops out completely and other times it will stick, but maybe a bit lower. And the sort of collective theory over the years was this is Google giving you your kind of trial period for your content and seeing how users like it. So a couple of things here, as we've discussed before, so we run a tool called Also Asked, we released a timeline feature to do exactly this, which essentially you, you can see the history of any key phrase over time and the volatility and how much the people also ask questions have changed around a topic. Of course, if you want to scrape it yourself as well, that's absolutely possible. But that's one of the main ways I personally track intent shift, which is I look at how often those people also ask questions change. I think that's such a good kind of canary in the mine, if you like, indicator. To give you some detail on that, you usually get the 4 people also ask questions attached to a query. These are usually quite static. Branched off from these, there's, for each one of those 4, there's another 4 or 5 questions. So there's sort of around 100 or so questions, give or take, in any full tree. The second level ones tend to change fairly regularly, because Google is experimenting around, they're kind of A/B testing themselves, seeing that they're like, using this as an induction loop to try and understand the queries better. Again, that's something that came quite clear from the DOJ trial, which is that a lot of Google's magic is actually just seeing how users react to search terms and trying to iterate and improve on that rather than, okay, we, you know, we've used BERT and we understand the nuance of this query. It's actually kind of, oh, we'll just kind of keep chucking progressively slightly better stuff at the wall and seeing what sticks and then chucking new stuff. So that's a really useful tool. Personally, how we score that is if the, one of the first primary root questions changes, we have our own like volatility calculation. And then if any of those first 4 change, that's like a fairly big deal. If the second tier changes, we, that kind of gives us a little bit extra of volatility because some of these are quite static even all the way down. But again, to use examples like The COVID one, it completely changed the whole tree from the first root questions changed over 12, 24 months as the collective knowledge of people progressed. So there's ways you can do this programmatically as well, which are really exciting. So again, I think we previously spoke about monitoring content and if we have answered all the questions using PAAs. You can do that now with Screaming Frog. So brilliant tool that loads of SEOs use. They have in their version 20, this custom JS functionality. And it's absolutely fantastic what you can achieve from this. So you can do crawls and you can query APIs as you go. So one thing I've been doing is actually checking if these PAAs are changing against the content as we crawl it. So we're essentially asking. We're fetching the PAA data as we crawl, and we're using ChatGPT to ask, have we answered these questions on the page? And just straight fresh out of a crawl, we can get a list of these are the, this is where the intent shift has changed. And these are the new questions you need to answer. And it's so good because if you've got a site with thousands of bits of content, it can become really unwieldy to manage and try and stay on top of what's changing. So this is like a An automated thing, you get on with, with some other work, and then you can get this report on, well, actually, these 10 topics, the questions have changed significantly, and we have kind of gaps in our content. By updating that, you're lighting up that instant glue thing of you have new content. So Google's going to give it a bit more attention. You've matched it up with what people are searching for. So you've reduced the chances of people having to redefine queries because their questions have changed. And that's something we've been having a lot of success with. You see a lot of SEOs talk about, you know, giving a strategy of just kind of updating content a lot and putting out fresh content. And I think that's got merit, but this is like a really specific targeted way to do that where you get the most bang kind of for your buck.

And it's a specific targeted way to look at present-day data as well, because many tools are able to give great historical data, but that's not necessarily a good representation of what's going to happen in the future. And I appreciate that some things you can't forecast, you can't forecast COVID happening and assassination attempt on a president. But surely some future projections can be relevant and appropriate to make. And ideally, I'd like a tool that actually told me what is likely going to be the search volume in 3 months' time in the future. Does such a tool exist at the moment?

The crystal ball tool?

Why not?

So I think that's possible, but that's human intelligence, I think, from people knowing their, their sector and their industry. So like, if you asked me about, for instance, that question for SEO, you know, we can, we can make probably some quite good predictions about maybe SearchGPT and generative AI search, and the kind of things we're seeing at the minute that we know are going to be big. I'm not sure if there's any tool that encompasses all of that yet that can, you know, predict those trends. But if there is one, I would love access to it. Yeah, right.

Me first, without telling anyone else.

Yeah.

If an SEO is struggling for time, what should they stop doing right now so they can spend more time doing what you suggest in 2025?

Some people are going to hate this one, which is, I think a lot of SEOs spend too much time on kind of like metadata type tasks. Now, things like title tags, super important for SEO, right? But I think spending a lot of time on things like meta descriptions and such isn't that helpful. And the reason is that. Again, this is a, this is more a query-based thing. And what I mean by that is I think the stats are somewhere between 50 to 80% of meta descriptions, depending whose data you look at, get rewritten by Google. And that's done on a per query basis. So again, across that spectrum of Google thinking, okay, you've typed this in, this is the appropriate page, and here's what I think you want. that's on that page, they then pluck that out and put it in a meta description, because it's in Google's interest to again, have you happy with the SERP. My kind of point to people is, if you are, say, writing a meta description, you can only provide one static description of everything that's on the page. And there might be 10, 20, 50, 100 search terms for that page. So even in the best case scenario where you manage to write something that's better, it's still only probably going to be shown 20-30% of the time. So by, you know, you're immediately losing a huge chunk of value. Kind of combined with that, Google's just recently announced they are going to be able to use the Open Graph titles they find in HTML as the page title link in the SERP if they like. And we've seen this happen regularly, which is Google rewriting titles in the SERP. And they commonly use normally what looks like the H1 of the page, like the big text, they can use internal links. And this is because I think SEOs are again, breaking away from maybe what's best for the user in terms of people are writing title tags for search engines. And then you land on the page and the H1 is kind of what's for the user. And it feels to me Google's kind of coaxing them back being like, yeah, this isn't kind of what we're after. This isn't what we want you to do. So we're just going to ignore the title you've written and we're going to use what we think is appropriate in the H1. Again, Google gets it wrong sometimes, but on average, I think that's, that's helping. So again, title tags, everything like that are important, but I think because of the way Google's rounding off those corners for us, if you like, itself. I think you can spend more time on, you know, to use the phrase that's been used to death, you know, providing value around your actual content and spend more time on writing, having good meat and making sure we understand what users are searching for. And we're making them happy rather than polishing a meta description.

Mark Williams-Cook is director at Candor and founder of Also Asked, and you can find them over at withcandor.co.uk. Mark, thanks so much for being part of SEO in 2025.

Thank you for having me.

I've been your host, David Bain. Get your copy of SEO in 2025, the book, over at seoin2025.com.