The Power of Google Discover to BOOST audience for Publishers with Reach's Nicola Agius

Hi everyone, I'm Steve and I'm Barry, and this is Beers and SEO with Barry and Steve.

And today we are really excited because we have a very special guest on our show, which is Nicola Aegis, who is the Director of SEO and Discover at Reach. Welcome, Nicola.

Wow, how did you learn how to pronounce my surname? No one can normally pronounce that.

Do you know what? I actually had to search it up because I, because I've known you for a while, but I've never really had to pronounce your surname. And then when it came to today, I thought, actually, do you know, I need to know, I need to research that. But do you know, we were talking a little bit before we started recording, but I think you're the only SEO expert in the world that I know, which is, which is director of SEO and specifically Discover.

Yeah. Yeah. Yeah, it's interesting. I suppose I've been at Reach for about 2 years now, and when I first started, I was just SEO director. And then I would say a few weeks in, or maybe a month, I was having a conversation with my boss at the time, and it was like, oh, you're director of SEO and Discover. And I was like, oh, okay. Because we have quite a big focus. I suppose for a lot of publishers, Discover is the big traffic driver. Yeah. And so we had a big focus on that. And so I suppose that tweak to the job title made sense. And do you know what's really interesting now? That— what's the statistic? That Google sends twice as much traffic via Discover than it does search. Is that correct?

Yeah, yeah, yeah, yeah, yeah. It's been quoted recently. Yeah, yeah, it's a huge driver, particularly to news publishers, right?

And I remember we were at the, um, I think it was the NEST London meetup a couple of years ago, and I was on a panel, and I remember everyone, a lot of the other publishers were saying, you know, oh, we just treat it as bonus traffic. And I just remember thinking, really, just bonus?

Yeah.

And I wonder now if 2 years on, when there's just been such a shift between search and discover, if people are still just treating it as bonus. Like, I do feel like—

They're not, but they should. But we can dig into that.

Yeah. So we're gonna have a slightly different format for this session. We're gonna cover, um, 3 or 4, um, stories that, that we've definitely been talking about as SEOs over the last month, some more recent than that. And then we're going to do a deep dive, which is why we've called you in, Nicola, in for Discover. And because obviously we've had the specific Discover Core update, which is, I believe, is still rolling out. So there's lots of insights into that that we could, we could share. But before we go on to that, Let's go into one of the stories that we said that we'd talk about, and it's really around a piece that was published in the Reuters Institute of Journalism at the beginning of the year looking at technology trends. And they worked in conjunction with Chartbeat, and they produced a particular chart which showed an apparent decrease in in overall traffic for both Discover and organic search. And this chart has been quoted. I think we've all been to conferences over the last 3 or 4 weeks where, when it's looking at the future of search or the future of digital publishing in general, that chart is being quoted a lot as the truth. And it's given the impression that there is no future. future in search.

Mm-hmm.

So, and I think the Press Gazette followed that up in the UK asking a number of senior publishing leaders, you know, their thoughts on this chart. And I think the feedback or the worrying feedback on it was that they may consider not investing in SEO across the year. So we want to respond to that. And I don't know, Barry, if you wanna sort of like take away first, because we have some opinions on that.

I have very strong opinions on that. The Chartbeat data, I think, is deeply flawed and doesn't reflect the reality of the actual traffic trends that we see in the UK and Europe. I think it's maybe very heavily US-focused in its data and maybe skewed by several very large publishers as part of the dataset who have seen decreases over the years, especially last year. And those decreases may not necessarily have been Because of the obvious reasons, but also because of things like site reputation abuse, algorithm updates, those sorts of things. Then that same piece on Press Gazette has an interesting few counterpoints being made by several publishers, including Carly Stephens from The Mail, Stuart Forrest from Bauer, who say that actually, no, we're not really seeing those sorts of declines from publishers at all for traffic. We've seen some decreases, but fairly stable traffic across the board. And certain tactics that work well in Discover and News have still continued to work. AI isn't really coming to steal all our traffic, especially for news type content. And there was also a competing study released a few weeks later by Graphite with data from SimilarWeb, from the top 100,000 websites, I think, that, yeah, show a decrease of Google traffic to the wider web, but that decrease is 2.5%.

Mm-hmm.

So I think people who just quote that Chartbeat piece as gospel saying, oh yeah, Google traffic is down. I think they're tapping into a panic mindset at publishers that doesn't actually reflect reality and makes publishers pivot away from optimizing for search. And I've been very vocal about this on LinkedIn and at conferences as well, that I think that is an absolutely catastrophic mistake. I think too many publishers have been coasting on Google's own growth for too many years to fuel their growth because Google has kept growing in terms of the number of queries that have been performed and especially the number of clicks that they generate to the wider web. Now that trend has flatlined. Google isn't growing anymore. It is growing in queries. It's not growing in clicks. And that's partly because of AI and partly because of other search features as well. And it has become harder without a doubt for publishers to claim traffic and to claim growth from search, especially if you're an SEO mature publisher. You know, the challenges are very real and very, very obvious for publishers. But I think when the going gets tough, the worst thing you can do is throw in the towel and say, oh yeah, well, that's not going to optimize anymore. It's too hard. Every single data source we have shows that Google as a whole, Search, Discover, and all the other surfaces, is by far the largest driver of traffic to publishing websites and will continue to remain that for the foreseeable future. AI tools have not made any significant dent in Google as the dominant source of traffic. So if you as a publisher decide to just invest less in Google or abandon it entirely, You're committing suicide. You're just doing the worst possible thing for your own long-term survival. And I understand publishers have to pivot away to different kinds of audience engagement tactics, to look at multichannel, multi-format formats for their readers, for their audience growth. But if that comes at the expense of still focusing on Google, of still making sure that you do all that you can to maintain and even grow traffic there, you are optimizing for a smaller slice of the pie at the expense of the biggest slice of the pie. And your competitors will be very grateful when you do that because they will just eat your slice of the pie. So in a nutshell, I think publishers need to just look at the actual data and not at these panicked graphs that they are presented at conferences as gospel and see for themselves, right? Where is our traffic actually coming from? Who are we actually competing with to get this traffic? And how can we make sure we remain effective in optimizing for those channels?

Yeah. Nicola, do you have a take on that? I mean, we know as SEOs that, you know, the daily SEO work, it's messy, right? We know we, we, we have to surf the waves, uh, and the spikes. Uh, that's just the nature of it, of SEO, especially when a core update, uh, comes on, which we can talk about a bit later. So we're used to this, but you know, the narrative around that chart is that everything's in decline. So, um, what would you must, I mean, obviously you, you, you, what, what, what's your take on that over at, um, Reach?

Do you know what's interesting? Like I remember around December time, I was being asked by lots of people, what are your predictions for 2026? And we're all asked to like give our predict— and I hate SEO forecasting because I just think it's really out of our hands. Like who knows what Google's going to change or roll out next? I think, um, we, you don't know what's going to happen. So I think a good place to start with would be to look towards the tech giants. What are they doing? And I've seen some pretty big SEO jobs. going on. Like, was it recently they were advertising for an SEO lead? I mean, I think we should be paying attention to what's happening on LinkedIn. I, I haven't seen a decline in hiring or opportunities. The, the biggest companies are definitely investing in SEO. So, um, rather than taking that piece of data from Chartbeat as gospel— and by the way, Chartbeat has some accuracy issues with splitting between search and discover as well.

Yeah.

So I wouldn't always— like, there's so many pieces of, uh, research that comes out and they all tell such doom and gloom stories. But yeah, I would take a bit of a step back and look at what's happened, the bigger picture.

Right. Yeah, totally agree. It's just obviously annoying when you're at a conference and these charts are presented at the very beginning. And because the data, because it looks all official and you've got these, you've got these lines going down. A lot of people will take that away, um, and, uh, not dig into the, into the detail of that. So which, which we're trying to, um, to talk about on every podcast that we, we cover here. Um, okay. So sorry. Sorry, Nicola.

Do you know what though? If my competitors, competitors want to stop investing in SEO, then you know what? Maybe you should. Yeah, definitely take your eye off the ball. I don't mind you doing that.

Listen to what Nicola says. That's right.

Yeah.

Um, okay, so cracking on, it's going to be a swift, uh, movement into the next, um, uh, story. It's really around the Competition and Markets Authority, um, here in the, in the UK who are coming for Google, I've written down here. But, um, a little bit of context for that because it, um, has resulted in Google responding with that, with, uh, announcing that it's going to release some more controls for publishers to control their visibility of their content in AI overviews. And, um, so a little bit of context on that, then we'll take, um, take your thoughts. But, um, so the things started to cook off basically in October 2025. CMA designated Google, uh, with strategic market status, um, in search services, which basically then allowed it to introduce targeted rules or content requirements for Google's businesses. So basically, once it had done that, it can now put a number of recommendations of how it wants regulate or introduce some regulation to Google. So, uh, as I said before, part of that is that it wants Google to introduce some publisher controls to make sure content publishers get a fairer deal by giving them more choice and transparency over how their content is used in Google's AI overviews. Um, there's another element that the CMA want to introduce, uh, which is around fair ranking. So making sure Google's approach to ranking search results is fair and transparent for businesses with an effective process for raising and investigating issues. Good, good luck. But, uh, you know, that's, uh, we, you know, it's all, all, this is all good and things that we've been asking for for a number of years. There's another element that CMA want to introduce, which is around choice screens. So making it easier for people to switch the search services they use by making, uh, default, uh, default choice screens on Android mobiles a legal requirement. So it's basically the, the opportunity to opt out of Google when you're on these devices. And another point that they've got there around data portability, making it easier for people and businesses to make use of Google search data. So, yeah, so from that, Google responded with saying, yes, it will introduce— it responded very quickly, actually, to this by saying it would introduce a tool to allow publishers to opt out. of AI Overviews. Now, we don't— unless Barry or Nicola, maybe you know something, but the last time I checked, there's very little detail released about this. We initially thought it would be some kind of initial on/off switch, but we went to the News in a Digital Age last week, which was an event in the UK, which was delivered by FT Strategies in partnership with Google. We had Google spokesperson as a keynote speaker, and they again didn't reveal much, but they said it would be more than just an on-off switch. There'd be a fair bit of, um, uh, um, depending on what it is. So, you know, quite customization or specific, who knows, it could be localization, who knows, but it sounds like it's going to be a bit more complicated than a simple on-off switch. Um, I also noticed at that conference that, um, when, uh, they, they, they said, we know that that you want more data, i.e., in Google Search Console. And they said, we hear you. And that was the end of that. So I don't— I'm not going to hold out too much for additional data. But when it comes, we're always grateful, but we need more. So yeah, so what's your response on that, Nicola, with the ability to have more control over your content outreach in AI Overviews?

Well, they said it was quite complicated. As you were saying, it's not an on/off button, which makes me wonder if you opted out, what would the consequences be? Would that make you less visible in Discover and Search? Also, realistically, the way that people search is changing, especially young people, and they trust what They really, they like the AI-generated answers. So do we, do we really want less visibility in AI if that's the way it's changing? Maybe it's not that we want less or reduced visibility, but maybe we just want to be compensated. If you're going to rent our content effectively, could you pay us some rent?

Right. So yeah, you opt out. What do you get rewards? You get less, less visibility. What do you think about that, Barry?

I find it very interesting. There's 2 aspects to this that I want to highlight. The first is that Google had already thought about allowing very granular control to websites over how they're shown or shown at all in AI Overviews, AI Mode, Gemini, all those things, and decided not to implement that. There was a slide leaked from Google from, I think, 2 years ago, if not longer. That showed Google were outlining all the different options that they could give publishers on the web to control the visibility in AI-generated responses. And they decided to pick one of the worst options that gave the least amount of control to publishers. So it was a very deliberate choice by Google at the time. And then the CMA in the UK basically forced Google's hand and said, no, actually we want you to give us more control so that we can decide as publishers whether or not We are shown in an AI-generated response. And Google, of course, immediately said, of course we're going to do that. It's just a very difficult engineering challenge because it's integral into search, all of those things. So yes, Google just plays the victim straight away and says, yeah, okay, we're going to try our best. And it's going to be hard to separate these AI controls out from search in such a way that it doesn't negatively impact websites in the broader ecosystem. And I don't know how true that is, but you know, I'll put my skeptical hat on to a certain extent.

Okay.

The second part of it is, of course, that it, like Nicola already said, you're sort of skirting the issue a bit. It's not necessarily about being cited as a source or not. The underlying problem is that all the AI systems have basically been trained on the greatest scale of copyright infringement in the history of mankind. And websites aren't being compensated for that at all. And the traffic we get from citations in AI-generated answers are negligible compared to classic search results. Now, I know we cannot put the AI genie back in its bottle. It's out there. But the utter lack of responsibility that the tech giants, Google included, but they're not the only one, are adopting in this is frightening and shows how callous they are in their attitude towards an actual healthy web ecosystem. Google claims to care about the web ecosystem, but their actions don't necessarily reflect their words. And third, I found this really interesting as well, that very shortly after this was published, Microsoft launched an AI citations report in Bing Webmaster Tools. Master Tools, which looks great, by the way. And it basically gives web publishers the information that Google says they can't necessarily generate in Search Console, showing websites when they get cited in AI-generated answers on the Bing search engine for which kind of prompts. So the technology is obviously there. If Microsoft can do it, why can't Google do it? And all, all of us clamoring for this extra data in Search Console and Google says, oh yeah, we can't really do that. Of course Google can do it. And it's not the first time Google has deliberately withheld data that is very useful to us. For example, we don't have a top stories filter in the search report in Search Console, even though publishers have been screaming for that for as long as top stories have existed. Same with featured snippets. Same when you show up as a, as any sort of rich result on a classic search result, you only get separate reports in Search Console for the separate tabs, the separate verticals within Google Search, not for what happens on the main Google search results. But of course Google can do that, they just choose not to do it. And Microsoft basically fired a shot across the bow and said, actually you can do it and we are going to do it, here you go, users. So it'll be interesting to see if and how Google responds to that, or if they just ignore Microsoft's ramblings in the margins because their market share isn't significant to pose a threat to Google.

Okay, so watch this space. Okay, so we'll move to the 3rd story, which was something which was an article which Rand Fishkin published in January, which I thought was still relevant to conversations, particularly around AI tracking. So of course, we've all read the report because it was an interesting study into the reliability of prompt results. So a little bit of background. So basically, Rand at SparkToro had just a around 600 volunteers. They ran about 12 prompts each, so that's about almost under 3,000 kind of prompts. And they were testing what was returned, um, and under certain queries. And they found that that list that was returned, um, changed quite a lot. And I think looking at it, it was something like there's less than 1 in 100 chance of ChatGPT or Google AI producing the same list in any 2 responses responses. The ordering of it can, uh, you know, so, so best, best restaurants in New York or whatever, the ordering is going to change practically each time you do that. Um, and even the length of the list. So, so what is included, it's, it's, there's a lot of randomness on there. So the— I mean, um, it's a fantastic, um, article and piece of research and, um, But the, the kind of the takeaway is that it illuminates that if you're going to do not use these AI tracking tools in the same way, like search rankings, that, you know, looking at the data that you have in tools like SEMrush or Google Search Console, be very, very careful about what is returned in these tools. Now, Rand was looking at maybe there's another kind of metric to use such as visibility. So what is the percentage overall of your brand appearing in these 3,000 returns, these prompt returning results basically? So maybe percentage results rather than going, right, okay, you're not appearing at all or you're at number 1 each time, you know, just look, treat with caution and maybe there's other metrics? I mean, this is part of the whole conversation around AI tracking and visibility is what are the metrics that you need to agree on as a business and use the tools more as directional as opposed to sacrosanct, really. Nicola, are you using any kind of like tools over there at Reach? And what's your take of what you've seen so far?

I know Similarweb has some kind of tracking, I believe, but Um, to be honest with you, we're not going in hard on that just yet because what's the opportunity there for publishers? It's not just like at the minute, like getting cited in ChatGPT or Gemini, that's not going to drive traffic to our website, or it's very unlikely to drive traffic to our website. That doesn't mean to say there's not an opportunity for publishers. I think I think publishers maybe need to think a little bit outside the box when it comes to the opportunities, um, with AI visibility. So, um, I've— I try to go to as many talks on AEO and AI and as I can, and I've accidentally joined a few that were more aimed for brands rather than publishers, and they're obsessed with AI visibility. But when you go to those talks, what they the speakers always tell them, you know, the key to getting that visibility is publishers. So I think that's where the opportunity is, um, and that's talks that we're starting to have, but probably I can't say too much.

Yeah, so this would be, this would be referred to as almost like branded content, right?

Yeah, yeah.

Citations are the new links, right? Everybody wants to get cited on publishers. You don't even need to get a link because it helps you get cited as a, as an source in an AI-generated response. And I find it, I mean, where do I even begin with this utter shitshow that is AI tracking? First of all, I wonder how many queries that, for example, OpenAI reports people do in their tool are actually tool-based queries trying to establish some sort of visibility index for their clients. And how many are actual users who actually perform prompts? In OpenAI. Imagine the amount of energy wasted on these prompts because LLMs are very energy intensive. And the amount of energy being utterly wasted by these tools so that some C-level people in some companies have a vague glimmer of control over their brand perception in the wider world, which is mainly bullshit because of the sheer unpredictability of the AI answers, as the study that Van Fisken showed highlighted. The answers are pretty much always random, and there's very little consistency in how brands are being shown and how prominent they are cited in a source. And as per Nicolas' point, I think publishers as a whole really should stay the fuck away from trying to get themselves tracked in AI-generated answers because it doesn't I think as a publisher, if you are traffic dependent as a publisher, any effort you spend on optimizing for being cited in AI overview is absolutely wasted. It is a total colossal waste of your time. You should just focus on being the highest quality, most authoritative, authoritative publisher that you can be. And guess what? That means you will get cited in AI-generated responses. Period. You don't need separate AI optimization specialists, geo bullshit artists to help you with that. And I mean, I understand that people want measurement, they want KPI, they want that illusion of data. And I have previously gone on record saying any data is better than no data, but I have to amend that and saying bad data is worse than no data. And the AI trackers generally give you bad data. Bad data that gives you that illusion of certainty where there is absolutely none. And it's not really actionable either. You can't really do anything with that in any meaningful way. It's just another noise in, in the data stream. It's not a signal, it's pure noise. And, and as anybody who does any sort of data analysis knows, you want to filter out the noise and focus on the actual signal. No. So it feels like such a colossal waste of time and energy and money to focus on getting some vaguely coherent idea of a metric of how popular your brand is in AI-generated responses, especially when there is so much else that you should be spending your effort on and your money on rather than these tools, which actually does move the needle in a meaningful way. Exactly.

Yeah, but the, but the dashboards look so pretty though, don't they?

Oh, they look fantastic. They're utterly meaningless, but they present you with pretty graphs and charts, so everything's right in the world.

Can I say as well that towards the end of last year, I'm not going to name them, but there was a very big AI tracking tool that had this very incredible glitzy event in London. And oh my gosh, to borrow one of Barry's quotes, it was a, it was a shit show. And I think anyone in the audience that knew about SEO, just sitting there listening to the speakers speakers, um, not the headline speakers, the other speakers that went on trying to give like AO optimization tips. It was so cringe and so inaccurate. Um, yeah, I was concerned because, you know, before that, before I actually went to that event which was supposed to promote the tool, I was in two minds, like, should we invest in this? Let me go along and check it out. And then I went, I was like, okay, no, definitely not. I'm gonna walk away. That was awful.

Yeah.

Yeah. Well, yeah. And yeah, I've actually seen less of those events actually. It was definitely the last 3 months before Christmas, it was a lot of activity in that area. And I think there's definitely been more communication about the dangers of making those tools your main strategy. So that's why we're here. That's why we're here in this podcast to communicate that.

Yeah.

Um, right. So, um, Nicola, we have you here, um, apart from your brilliance in digital in general, but, uh, I thought we could also, uh, sort of deep dive into all things Google, uh, Discover. It's obviously a big area of focus for publishers. Um, so, uh, I wonder if you could talk about a little bit about how you tackle that outreach and kind of like your main considerations, um, uh, particularly for those that might be listening who, um, who, who might be considering how, how do they start off, uh, investing resource into Google Discover? How do they get into Google Discover? I know that there's certain limitations on, on what brands are visible in there, i.e., could a small startup, you know, continue with the content team and get in there in the first couple of weeks, or what, what's the— so there's a few questions there, but do you know what, in general, be good to know.

I almost smirked there when you said, how can a new site get in, get into the Discover feed? I smirked. I was like, oh no. And then I was like, wait, the amount of spam that I've been seeing recently and the spam websites that have been getting in. So I've been monitoring this quite closely. It's the same private blog network over and over again, and they're buying expired domains from local businesses. So it'd be like Paul's plumbing business and, you know, Sharon's hairdresser, and they just delete the content and then put a load of blog, the same blogs about the, the, um, speed on national speed limits on the motorways changing, about, um, the pension limit, um, age increasing. And it's the same content getting in there over and over again. So actually, yes, probably a new— as long as you've got a bit of authority, a new site could get— shouldn't theoretically. But actually, to be, to be honest with you, um, I did catch up with Google last week About this core update, the February core update, which is the first of its kind because it's a Discover core update. And this has been designed specifically to try and fix the spam issue that we're having. That's why they've rolled it out, um, is to try and fix that. So theoretically, new sites, I wouldn't say could get in there. That said, yeah, I've seen a lot more smaller niche sites get into the Discover feed at the minute. That's not to say they're new sites. Um, but yeah, an opportunity for definitely more niche websites right now.

So how does, um, how does it work, Nicola? How does, uh, what, uh, what is Google pulling, uh, from what it knows, uh, about the user and how does it display it in the results?

Supposedly it's supposed to show you content based on your search history, your locations, your interests, because Google has a lot of information about And it's supposed to be a feed that's personalized to you, and everyone's feed is different, so my feed won't look like your feed, etc. Um, but we do have some really good, um, tools available to us that shows us what's kind of in the, in the general feed. There's a few tools that we use. Um, so we use one called Marfeel, which has a good Discover monitoring feature. We've started using John Shahata's from NewsDash, uh, Discover Pulse, which is really good as well. And there's also another tool called Discover Snoop, which I haven't used, but I know a lot of publishers rave about it.

So can you, um, what kind of intelligence can you get from Google Search Console? Is it, is it limited?

It's not that. So obviously, like, Search Console is the best data, like, for looking at your, what's working for you. But I don't know about you guys, but whenever you're in the newsroom, they don't want to know what's just working for them. They're like, what's happening to your competitors? What are they seeing? Why is this person seeing? So those tools, those third-party tools, they give you the kind of broader overview of what's happening in the wider market. And if you're really going to go in on Discover, you need to have that knowledge. You need to be able to identify, okay, this has dropped out. Like, for example, a topic might work really well for us. Like, for one of our sites, like, film content worked really well for us, and then it dropped out. And it's important for you to know, is that because of something we've done, or is it because of something Google has done on their end? And then you can use these tools to see, oh wow, actually Google's completely, uh, decided to demote that category across the Discover feed. No one's getting much content in for that category right now. So, um, it's helpful to know, is it something you've done or is it something they've done on their end? And if Google has decided that it just wants to make film content drop out of the Discover feed, there's nothing you can do. So if that is a topic that you are targeting for Discover, Maybe pull back a little bit, watch what's happening, maybe pivot slightly and wait for film to come back. I'm not saying that you should stop writing film content, but if you're writing with Discover in mind, it's helpful to have that knowledge.

I think what you're saying there is that Google, uh, optimizing for Google Discover is extremely frustrating, right? Because it changes all the time. And, um, from my experience of working with, uh, newsrooms You know, it's spiky. So we, we know that it's unlike organic search in the way that you can start building up a bedrock of content that then accrues more and more organic visibility, and you can see clear paths. And, you know, no one likes to really put predictions in, but you can see a general upwards pattern, whereas Google Discover Um, you know, publishers may not receive anything, and then obviously they have to do some preliminary optimization for it to, to get into the, into the feeds. But yeah, it can drop off. So that's why I always say it's a tactic rather than a strategy. But a lot of publishers put a lot of eggs in the, in the Discover basket, right?

Yeah, I mean, we do very reg— I, my team do very regular reporting on it. For us, it definitely is a strategy. We go quite We try to target Discover because it's such a huge traffic driver for us. And when it comes to optimization for Discover, I suppose the key is monitoring what's working for you, monitoring what's working for other sites, and looking for patterns. Buzzwords, like if pension, like if money content is working really well for you, or food content is working really well for you. So recently, for example, like food content works really well for a lot of our sites, but we saw a lot of bit of a drop. So I checked to see what's happening across the wider market. There hasn't been a drop-off, but actually the kind of food content that Google is giving visibility to has changed. So rather than, you know, the secret ingredient to making your chicken soup perfect, now it's like news about, you know, the best celebrations have like got rid of a chocolate in the tub or something like that. And um, uh, it's promoting like supermarket content, that kind of content. So you really have to pay attention to like what's working and what's not, and we, we monitor that very, very closely.

Yeah, it's interesting. How do editorial teams respond to that? Because in the old days of 2022, um, you would normally have a strategy with regards to, um, focusing on particular news topics and then backing those up with evergreens, right? So You could create a situation where a team, you, uh, you, you, uh, say, right, we're going to target this topic, we're going to cover the breaking news around that topic, and we're going to have pillar content, evergreens all around that. And almost the, the team you had that could just go away and produce that content, right? And you could see the uplift after a while because, and that's all about having that focused attention on that.

Yeah.

Um, the downside of that is You would need to check in with them after a certain while or check into that data, right? Because what they were producing on month 1, things may have shifted. So by month 3, but it sounds like in Discover, it's almost daily looking at the data, right?

Yeah. Yeah, it's a lot because it's so volatile. It's just honestly, it's such a wild beast. Even like what's just happened over the last 12 months. the patterns that we've seen. So this time last year, I would say national publishers absolutely dominated Google Discover. We were getting— well, everyone was getting a hell of a lot of traffic on national news titles. But with the June core update, there was a big shift. Rather than just give that visibility to the national news publishers, Google decided to start giving more visibility to regionals and smaller niche sites. And where at Reach we have 70-odd news sites, I could really see that happening. So in order to make room on the Discover feed, because there's only so much room for the regionals and the more niche sites, they had to take away a little bit from the nationals. And so there was a little bit of a shift there. And this time last year, no one could have seen that happening. Um, and then I would— then even in November time, out of nowhere, uh, X/Twitter became the 4th most visible domain within the Discover feed. Then by December-January time in the UK, it was the most visible domain, overtaking YouTube. But then in February, uh, Google decided to press another button and it dropped out massively. X. And, um, then that's why it's a very volatile referrer. And that's again why I say that it's very difficult to forecast, uh, Google traffic or tell people what might be um, what's on the horizon for Google traffic.

Right. So just keep on looking at the data, right? And say, and, um, yeah, so that's what I tell clients is basically, so some, some, it's an interesting challenge for an SEO because some clients may have like mountains of spikes. So, you know, overall they've got a lot of engagement. So It's— but they've, you know, traffic may be going down, so that's why they're enlisting the help of consultants. So why is this happening? Um, so that's one way that you need to respond to that, because obviously there's been success in the past. Um, so it's around, as you say, it's showing that, you know, there are all these changes and it fluctuates all the time. Um, the other challenges around publishers that may not be seeing much traffic at all Um, so there might be the odd blob in Google Search Console and then nothing. So obviously that's then looking at perhaps a load of more things than just the content. It's also looking at the, the tech stack and the site performance, Core Web Vitals, et cetera. There's a number of different things, um, that often we have to look at. Um, but, um, so with regards to the recent core updates, um, have you seen any, um, Changes on that, Nicola? It's only been about 2-3 weeks since it's rolled out, I think.

Yeah, it's interesting. So, um, there's definitely like a lot of volatility at the minute. And I think if you look at any of the tracking tools, and Barry Schwartz has been reporting on this as well, like the volatilities are absolutely crazy at the minute. I'm definitely seeing that within the Discover feed. And it's funny because Google said that it's just rolling out to US sites to begin with. And I caught up with Google last week week. And I said, are you sure it's only in the US? Because I know— because I follow Glenn Gabe, I adore Glenn Gabe. Whenever there's a core update, I pay attention to everything he reports. But, um, I said, are you sure it's just the, the US? Because we're seeing a lot of volatility. And they explained, actually, when the core update rolls out, they don't just go from 0 to 100, it's like a phased rollout. So it might go from 5 to 10 to 50%. So that's probably what we're seeing. And they also confirmed that the UK and India are next on the list. So that's probably what we're seeing. So a little nugget of information there.

Already, I suspect, because I, I have been approached by an Indian publisher who I shan't name, who have, uh, who are suffering from this core update and as well as from the December-January core update that Google had rolled out.

Do you think there'll be Um, you know, as, um, Google, um, rolls out Gemini into more services, do you think there will be more personalization? What do you think will be future changes at all for Google Discover?

Well, with preferred sources, I think they're trying to give people like the option to make their feeds more, uh, personalized as possible. So they— I suppose they want— their aim really, what they want to see is the most engagement possible within that Discover feed. Probably they're going For advertising purposes. So I suppose it makes sense for them to make it more personalized. And that's why if you're a publisher, um, trying to encourage people to sign up to get to make your title a preferred source is really important. Be exploring different ways to do that.

Yeah, I don't think the uptake on that is particularly high though. Uh, you know, a lot of publishers have started integrating that in top stories as well as If they have an entity page in the Discover feed for people to follow them on Discover. And I'm seeing a fraction of a percent uptake on those buttons. So I think it is more of a way for Google to placate publishers rather than to give anything, any meaningful optimization for a publisher to put themselves front and center for their preferred audience.

So I put, made like the SEO titles like search engine land, my preferred sources. It does give increased visibility. But I suppose my issues with that feature is that they don't make it that easy. It should be just, okay, just click that, but you have to click that, then go through something else, then click something. You've made it too complicated. I think there's a way to streamline making a publisher your preferred source.

Yeah, I think someone once wrote, don't make me think, right? Just make it as simple as possible.

Steve Krug, the definitive book on UX.

So on the— just to finish off. So it's still rolling out. Google Discover is already a very, as we've been discussing, it's very volatile and fluctuates. So when should publishers act once it's rolled out?

We try to give, we give guidance regularly during core update rollouts because, you know, I know that like what most SEOs do, you kind of wait 2 weeks until after an update has completed, wait for everything to settle, and then you start drawing your conclusions. And we do that, but we do mini reviews in the process to show which sites are up, which sites are down, and we very much track what's just happening across the wider market and give that feedback so that they can act on that. And to be honest with you, discovery is so volatile. I think you can't just wait until the update has completed to give them that feedback. But that said, doing the deep dive, you should wait like 2 weeks after it has completed. Because even like with the last update in December, we're seeing something similar now. At first it was like, oh my goodness, traffic, but then it went right back up again and it was I was like, okay, so if we— and this is why I kind of like highlighted to our editors, if we had acted too soon based on what we were seeing initially, we would have made, you know, the wrong— we've been making decisions on incomplete data. You have to wait for it to finish.

I think, Barry, you can relate to that as well, right?

Oh yeah, it's never done until it's actually stable. I should also say, Nicola, you have a very privileged position in that you have direct access data to what, 70 websites that we have? And I think that gives you a lot more insight and also a lot more flexibility to change tactics on a per-site basis with minimal risk to the overall traffic level. If one drops, you can, you can make that up in another website to a large extent. I get approached by a lot of publishers who generally have one website and do all their eggs in that one basket. And I do feel the need to make this point very clear because I've made it several times. We talked about Discover. You said there's a very strong strategy at Reach to optimize for Discover. And in the context of Wix, for your large network of websites, that makes perfect sense. Absolutely. You want an editorial strategy for Discover that you can spread out across all your websites and try different topical focus for different websites, different strategies, different tactics. If you have just one website, the risk of focusing your entire editorial strategy on Discover at the expense of other channels is, in my opinion, too high. Because of the volatility, because Discover has a tendency to encourage very bad journalistic habits in terms of creating clickbait, thin content, reusing the same piece 17 times with a slightly different angle without really adding anything new. All those things that Google has been telling publishers to stop doing for the last 20 years. Discover seems to encourage that and doesn't really have an effective way of filtering that sort of content out of its feed. And this core update is perhaps a first cautious step to an improvement in that scenario. But what I see so far, it's not a meaningful improvement. Clickbait still works. It's just perhaps less space for spam websites, but, you know, clickbait headlines, deceptive headlines, rehashing the same story with a slightly different angle 17 times, it still seems to work just fine in Discover. And I think as a publisher, if your goal is to drive traffic from Discover, That pitfall is very real that you're just going to follow the tactics that you know work that drive traffic because in the short term they absolutely work. They will send you heaps and heaps of discover traffic and it becomes very addictive and that is not necessarily conducive to your long-term survival as a publisher because at some point inevitably that waterfall of discover traffic is going to turn into a dripping tap. And I have not yet seen a website who has been that badly affected by an algorithm update and lost 70, 80, 90% of Discover traffic. I have not seen a meaningful recovery yet. Even with dramatic changes in editorial stances, the timeline for recovery from losing most of your Discover traffic is at least a year and a half, 2 years from what I can say. You're basically starting from ground zero. Which if you have 70 websites, fine, you can survive that. If you have one website, that's an existential threat to your business.

Right.

And that is why I still say to this day, despite the fact Discover is the largest source of traffic to publisher websites on the web, to always treat it as bonus traffic, to especially if you have a single website to focus on, you need to have a viable traffic portfolio If you factor Discover entirely out of the equation, and only then do I think you are in a healthy position as a publisher. If you can literally take all your traffic, take out all of Discover, and look at what remains and say, "Yeah, okay, we can survive on this. We can maybe even thrive on that." You're in a very, very healthy place. If that is not the case, if you are very dependent on Discover, you need to be very, very careful in how far you go. To optimize that channel. You need to make sure you do not go down a path where you undermine the signals of quality and authority that Google is very clear about. They even provide you with complete checklists that your content should adhere to in terms of the kind of output that you generate. You do not want to compromise on that. And yes, that means you're probably not going to maximize Discover traffic in the short term. Because you're not going down the clickbait route, you're not going down the journalism route. But in the long term, it ensures that Discover stays a viable channel for your website and that you're not risking your entire presence in the Discover feed for short-term gains at the expense of your long-term survival.

Right. Because, you know, it's another example, isn't it, of publishers being led by an algorithm and not taking control of of it themselves. We had the big wobble in 2017 with Facebook, uh, Facebook turning off the tap to publishers. Um, all, all bets then went on to search. And then we've had interesting times over the last 2 years with the introduction of AI into search. Um, so it's— and now, and now the shift and, uh, focus on Google Discover. And I've certainly been working with, with teams where their, you know, kind of content like, um, latest shops that have closed on your high street, which was a trend, uh, at some point in Google Discover, then, then the whole team's writing about shops that are closing on high streets and what could, what shops could close on your high street. And then you're, and then you're left with, wait a minute, what, what are we doing here?

Yeah. Do you know what, when we say, I agree, bonus traffic is, or not all referral traffic technically bonus traffic. And just because something's a bonus doesn't mean you should Bonus is a nice thing to have, right?

I agree. I agree. You should optimize for Discover to a certain extent. I think the risk is when you base your entire editorial strategy on Discover.

Yeah.

That you go down a very slippery slope very, very quickly. And the risk is very real because it is such a powerful source of traffic because the tactics are fairly low effort sometimes for very high yield. But the risk is that you And because of the way the news cycle works, a lot of editors in newsrooms don't necessarily want to look at long-term statistics. They want the short-term targets to be met. Day-on-day traffic is still key in newsrooms. And I think that is sometimes, if not all the time, to the detriment of the output that the website produces. Now we are seeing more and more initiatives and perceptions that change that, that publishers seem to be very keenly aware they shouldn't be chasing after clicks and that they should be building long-lasting brands that, you know, immunize themselves as well from whatever the tech bros dream up in the next ketamine high. But on the ground in the trenches, there's still too many publishers I feel that chase after the cheap clicks with a very high disregard for the long-term damage that they're causing their brands. But on the other hand, I can't be too sad because for me it's a form of job security as well because inevitably when that whole house of cards comes tumbling down, they find consultants like Steve and me and others who then have to clean up the mess and tell them things that they probably already know, but that they need spelled out for them very clearly and say, yeah, you've been doing naughty things and now you have to clean up your act. And it's going to take a long time for that to yield results. But you know, it's a problem of your own making. So here we are.

And also, I suppose it's worth highlighting again that discovery is an extension of search. So, you know, you need to get search right if you want to get Discover right.

I would have agreed with you up till February 5th when Google said we have a Discover-only core algorithm update. I think now we are starting, we are at the start of an era where Discover is its own system, that Google is now crafting algorithms specifically and only for Discover. There was already a bit of a separation in search and Discover. If you, for example, compare Google News, the news tab and top stories with Discover, you'd see very different spreads of publishers and articles, which is, you know, Discover trying to give a broader range of visibility to different websites. And I think now this particular Discover-only core algorithm update is a very clear sign that Google sees Discover as a separate thing from search. That yes, it's built on more or less the same foundations as search in terms of crawling and indexing and initial quality and ranking signals, but the visibility signals on top of that You know, in search, Google has these things called twiddlers where they adapt, adopt rankings on the fly depending on all these different metrics that they use. And I think this Discover algorithm update is Google introducing a new set of twiddlers specifically for Discover, which starts to make it an independent ecosystem that is partially built on search, but partially its own thing. And we're going to have fun as an industry figuring out which signals specifically feed into Discover. versus search and how we can maximize those particular signals.

Did you say twiddler?

The technical term is twiddlers. They are called twiddlers. I did not make that up. And it's a bit of an inappropriate word for the British ear, but they're made up by Americans. So there you go.

Sounds like a children's book. There's a famous children's book, I think. Okay, cool. So let's wrap up there. I mean, Nicola, it's really good to go into the detail there with Google Discover. I think We've uncovered quite a bit about how newsrooms, you know, respond and optimize for Google Discover. And we'll have to get you back on at some point, maybe in the future. Maybe let's talk about 3 or 4 months' time and see how the algorithm sort of like evens out.

I'm sorry about my internet connection. That was so annoying. Like, God, what happened?

We are nothing but agile on this podcast.

We're just gonna have to give Steve more work with the post-recording editing. More beers for Steve.

More beers. So I think we always end with a classic shot of—

oh, oh, I left my glass upstairs, but oh, oh, there you go.

Yay, that works. Excellent. Perfect.

And I guess the last thing, uh, to quickly to mention before we sign off is, um, Barry, you have a, a paywall masterclass coming up.

Yes, on March 4th, I am doing a masterclass with Jesse and Shelby from What the Fuck Is SEO. And heavy clocks and wheels from the Telegraph on paywalls. And it's a 2-hour online masterclass you can buy access to. And we're going to cover all aspects of paywalls, including technical setup, commercial strategy, editorial strategy, and everything else with a Q&A at the end. So if you go to seoforjournalism.com, which is the website of Shelby and Jesse, you will find all the details about the masterclass there.

Perfect. And I've got one announcement, which I'll be running an SEO and AI visibility masterclass, a day event, which you can find out at Journalism UK, and you'll be able to book your session with me on that. We'll be covering many things that we've covered today in today's podcast. And I think it's just a reflection that I think a lot of things have changed over the last 5 or 6 months. So it's a good refresher on the core values of SEO and all those services that have changed now, how to optimize for them. Um, Nicola, do you have something to announce?

No, but if you see Reach sites in your Discover feed, definitely click on it. It's good for the environment.

Yeah, absolutely.

We want to reward our friends, so I will definitely click on the Reach website.

Okay, everyone. Good to see you again.

Take care. Take care.

See you at the next one.

Bye.