Beyond SEO: How AI Is Changing What Websites Need to Be

Music

Sandra: Hi everyone, and welcome to another episode of Greyd Conversations. Today’s topic is one that I’ve been looking forward to for quite some time. We all know AI is changing almost every aspect of our industry at the moment, and we’ve already covered topics like AI powered website building, AI tools for agencies, maintenance, security, and automation. But one area we haven’t covered yet here is how AI is changing the requirements of websites themselves. What makes a website discoverable? What makes a brand visible? What makes the content trustworthy nowadays? And what role does a website even play when more and more answers are generated by AI before someone even visits a web page? To help us answer some of those questions, I’m joined today by Emma Young. Emma is Head of Organic Marketing at Hostinger where she leads SEO, Content, YouTube and Localization. And some of you may have seen Emma’s fantastic presentation at WordCamp Europe in Krakow, which in my opinion was one of the strongest talks of the entire event. So I’m very happy to have you here. Emma, welcome to the show.

Emma: Thank you for having me. I’m also excited to be here.

Sandra: Before we dive into AI search, would you mind telling us a little bit about yourself and what you do at Hostinger?

Emma: Yes. So, hi, I’m Emma. Like you said, I lead the organic marketing teams at Hostinger. So those four channels, a little bit of socials in between and some shorts. But in practice, it’s about like 65 people right now across multiple channels. And I think all we’re trying to do is answer one question, which is: how do we get the right people to find us without throwing money at it? So it’s very different from the paid marketing side of it. And I’ve been building this function for a few years now, and I think what kind of keeps it interesting is that organic used to just basically mean Google, and now it means everything. So a lot of the stuff that we’re going to talk about today. But I would say that my job and my scope has just exploded alongside AI search and everything. There’s just more and more and more of everything. So a lot more cross team collaborations. I think where we might used to have been at the end of the marketing loop, like the plus one or the enabler, our organic teams are now at the kickstarter, the early signals and then signs that it’s working in addition to that enabler. So this AI search piece has been really nice for my position. I like to juggle a lot of things at once, and it’s something that I couldn’t really ignore when I started seeing it show up. So yeah, it’s kind of what pulled me into this space. And it’s a lot of collaboration and consistency. We have this joke like evangelism of SEO but evangelism of AI search. So a little bit of that.

Sandra: That sounds like a lot and super interesting. And you also already hinted at some of the questions that I’m going to ask. So let’s maybe start with the big picture. One of the notes that I took from your presentation at WordCamp Europe was: your future customer searches in places where your brand doesn’t exist yet. That sentence kind of stuck with me because it completely changes how we think about websites and SEO. So could you explain a little bit to us how the search behavior has changed over the last, let’s say, 2 to 3 years?

Emma: Yeah. So that line specifically came from something that I’ve just been seeing in our data, and there’s been a lot more case studies that have really been released over the last couple of years. So like traffic patterns that didn’t match the old model anymore. We could see like the great decoupling. But also I felt like people were arriving at Google having already kind of half decided what they wanted or what they didn’t want. And so we were trying to figure out where they did this research, where else did they learn about some of our cool new products or trends? And why was Google only becoming the thing where they went to either confirm that or to buy something? So I think three years ago, the funnel still started on Google for most people. You had a question, you googled it, you found an answer. And that’s just not how it works anymore, especially for younger users. Like discovery now starts with TikTok or YouTube or some short social media type things, even LinkedIn. But Google still gets that click. I would say it still is often that last step, but just not the first anymore. And I think what’s really opened my eyes was, I visited home and my parents, who are in their 70s, it made me laugh when they started saying, I’m going to google it, I’m going to google it. But this last time when I went home, they’re like, I’m going to ChatGPT it. So I was like, what? I think that also just shows that it’s not just the younger ones. It’s everywhere now.

Sandra: So yeah, I think the biggest key takeaway here is that traditionally there was that one platform, that one algorithm, that one ranking system that we could optimize for. And today we have all these different platforms, AI systems and discovery channels. So how should companies rethink visibility in this new reality?

Emma: Yeah, that’s a loaded question because, I feel like, this landscape of where people land on Google or on some other platform has kind of changed from this ladder. I talked about this ladder in my slides as well. A colleague of mine does it. She kind of put this idea in my head and I think this ladder used to make sense, but now it’s like a map, and you have to be everywhere and consistently everywhere. So it’s like, that question is really hard to answer in just one sentence, but I guess the TLDR is: you have to be everywhere that you think your potential customer or client will be searching or just be, and you have to have already anticipated what they might be looking for and answer that so that you can be somebody that they consider when they finally do want to make a decision.

Sandra: Yeah, I think for me personally, since I’m also in marketing, that’s one of the most difficult things of this entire change. Nowadays it’s not even the user anymore who has questions I have to answer, because most of the time AI search, ChatGPT, Perplexity or whatever is the one answering the question. So I often get asked, if AI does all the answering, what does my website even matter about? Is it enough in the future to just have like the pricing and buy button? Because that’s what users are looking for at the end when they finally hit Google. Or what role does a website play?

Emma: To me, I think websites still matter. But I think the job description has changed a bit. For the longest time, the website was the destination, the whole point of SEO, for socials, for email marketing. All of that was to get somebody to your website and land on a page. And I think that’s still true, but it’s no longer the only truth. So now your website is almost like a source. It tells the AI system what to crawl, what to evaluate and like, decide whether or not to trust it. So it’s still the foundation that determines whether you get cited or referenced or recommended, even if nobody clicks it. So when people ask, do websites still matter? I think they didn’t become irrelevant, but they became part of the infrastructure. And it’s nothing glamorous, but nothing works without it. It’s kind of like websites now doing two jobs for the price of one, which is very relatable for the rest of marketing.

Sandra: That also raises the question, how important is the design of a website? Because before it was humans visiting the websites, and the websites had to be built in a way that it is attractive to them. They want to go for the content. The content is also structured in a way that they can visually see what they’re supposed to do, where they’re supposed to click, etc. And now it’s mainly AI. They’re definitely all the structure part. We’re going to talk about the technical aspects in a minute as well. But the design part, how relevant is that when one of the main target groups is not a human, but an AI anymore who probably doesn’t even care what a website looks like? So today a lot of the focus on websites is on design. Do you think that will change?

Emma: I love it, it’s like a spicy question, you know, like you say the wrong thing and all the designers are going to come at you. But I love it. So this is my bit: I think design still matters a lot. But the definition of good design is maybe what’s shifted. So today, good design means your website works beautifully for humans who arrive with high intent, and it’s also structured well enough for machines to extract that same meaning. But I don’t think those two things are in conflict. They can work alongside each other. They just need to have some collaboration there. So yes, there are websites that are maybe too focused on design, like beautiful websites that win awards but are an SEO nightmare, you know, like super JavaScript heavy or load too slow. But then there’s the websites that are the complete opposite, that are winning in an AI first world, but then they’re really ugly to anybody who has a little creativity in them. So I think it just requires designers and developers and CEOs to be in the room together from the start, not one make something up, and then the second tries to fix it and the third tries to implement it.

Sandra: It probably, which just came to my mind, will be a little bit like when the mobile responsive topic came about, because at the beginning we were like, okay, we’ll build a website, then afterwards we do a mobile version. That’s kind of how all this started. And today that just happens together. It’s just the fact that websites need to work on any browser size or any device. And that just became a combined process and not two separate steps that probably were even done by two different people, or one step completely forgotten. So it has to be done together, probably something like that.

Emma: I mean, I’m sure it can be done separately. It’s just really annoying, really frustrating. But I agree, mobile performance is a perfect example. All of us are just scrambling to try and fix it now and it’s like, man, wouldn’t it be nice if we thought about this like 15 years ago? But who knew?

Sandra: Also, another thing: I just came upon another post on LinkedIn, I think yesterday. I don’t remember what the company was, but the big news was like, yeah, we completely ditched Google. We are now completely optimizing for AI visibility only. But you also covered that in your talk at WordCamp Europe, that Google is not irrelevant. Because a lot of the AI visibility also is based on Google’s index. Could you explain how important Google still is and optimizing for Google?

Emma: Yeah. And it’s really funny because I remember I had this chat with someone like a year ago or maybe longer than that. And we both went down this conspiracy theory existential crisis ramble. We’re like, yeah, we got to reduce our dependencies on Google. But then both of us, the next day, were like, we were crazy. It’s just like energy feeding off of each other. So I said, basically: if you’re not in the index, you don’t exist in the answer. And if you’re not in Google’s index, you don’t exist in AI search answers. And I think this is very important because a lot of people, it surprises me how many people assume that AI search is separated or lives in some other universe that operates independently from Google. It’s not. Most major AI systems are either directly using Google’s index or training on some type of data that uses Google’s crawlers. So if you’re not in that original index, you’re not going to be in the conversation later on. So you literally cannot be cited in an AI system that has never encountered your content. And that only is encountered through this index. So when I say, if you’re not in the index, you’re not in the answer, it’s not in a metaphorical way, but in a more mechanical way. The index is your entry point. And this whole “SEO is dead” and “traditional SEO has moved on” stuff, those things are still very important. All of the things that you showed that you have to do to show up on Google pages are the same exact tactics that you would use for AI search. Then you do everything else on top of that. But the first hurdle you have to do is actually be indexable.

Sandra: Yeah. So that also means that there’s no magic thing happening, that you didn’t get it to work out for you to appear in Google, but then suddenly you do everything right for AI, and AI will put you in the overview. That’s not going to happen.

Emma: No, definitely not. It’s not some magic box that like, okay, I messed up on Google, so now I can show up on Perplexity. I wish, but no.

Sandra: Yeah. And probably that not only applies to the content. That probably also applies to the entire technical foundation, right?

Emma: Yeah, yeah. For sure. All of the technical SEO things that I don’t even want to call technical specific, but the things that you used to do or you always will do for SEO, like your site architecture or your schema, how you set up the site in the backend, all of that stuff is still very important. Whether you want to show up on Google or on AI search.

Sandra: So I assume things like semantic HTML and structured data are probably even more important today than they were before. Because it’s not just the humans, but also the machines visiting the site.

Emma: Yeah. Like semantic HTML, I don’t want to say it’s more important than it was in the past, but it’s one of those things that sounds super dry until you understand what it actually does. And then it becomes a little bit more obvious why it matters. So we need to use semantic HTML properly, with the headings, the hierarchy, all of that stuff. You’re giving the machines a very clear way of reading and ordering your content, so that they very clearly know what is a heading, what is a paragraph, all of that stuff. And it’s very helpful for a human to read your stuff, but also for an AI system to pick up the meaningful parts of your page. But then compare that to a site that completely ignored semantic HTML in the past, that just has divs and inline styles, where everything kind of looks visually okay. That was still okay for a human. You could still read it. Fine, but it made no sense to a crawler. It can still read it. It just has to work much harder. And I think there’s more chances for it to get it wrong. So the way I like to say it is: semantic HTML is like having good manners for a machine. The machines are your audience and you want to just have manners to them. And that’s like the easiest way to kind of explain the importance. But then there’s structured data, which I think is also something that I’ve seen a lot of LinkedIn posts about, where people are like, oh, structured data is going to magically help me be more visible on AI. And that’s also not it. It’s just the communication tool. It’s how we are speaking to the machine so that they can understand. You can have bad structured data, and it won’t destroy your website, and you can have good structured data, and it won’t save a bad website, but on a good website, it’s a very good, meaningful signal. So it’s lots of little things to do.

Sandra: Now, what about accessibility? Do accessible websites naturally perform better in AI systems?

Emma: I think so. Accessibility is finally getting its spot. It’s one of my favorite things to talk about recently because it reframes accessibility from this compliance check into a strategic advantage. Especially when, like, the Accessibility Act in Europe came out, I can’t remember the acronym for that. Everybody just went through and was like, okay, we can do this. Check, check, check. But we didn’t actually make it into part of our strategy for showing up. So of course, alt text on images, those clear heading structures, those type of things like captions, those all help an AI system understand your content better. And they shouldn’t be a separate requirement. They’re the same requirement, but kind of approached from two different angles. So a screen reader and an AI crawler are solving a similar problem: how do I make sense of this page without relying on any of the visual cues? And I think if your website works well for one, it should work well for the other one. So the companies that really invested in accessibility because it was the right thing to do, I think definitely have an advantage. Now it kind of feels like justice.

Sandra: Yeah. Kudos to everybody who did that. And the good news is, it’s not all extra work. If you’re doing a good job in terms of accessibility, it automatically supports your SEO and your AI visibility, and the other way around. And that has always been the case. But like I said, so far it was mostly about compliance and legal requirements and not so much about the opportunities that having an accessible website could actually offer to your company.

Emma: And I think you actually explained it very well. None of these things are new that we’re doing. They’ve kind of been invisible in the background. Being a human, you kind of can fill in gaps. And so you don’t really notice all of these things happening. You just kind of fill it in and you have this instinct to weigh heavier on something that you’ve seen that you trust more, and you give a little grace for something else. But an AI doesn’t do that. So now all of these things are being flagged a lot more, where you’re like, oh, this isn’t a priority right now. And I think all of those things are just surfacing more and more like, oh, crap, we need to do this. We should have done this at the beginning.

Sandra: Now let’s move from the technical side to the content part. One of your slides said: “write for the model that summarizes, not just a human that clicks.” What does this actually mean in practice?

Emma: What does that mean? What was the little ones like? My stack had a lot of cocktail imagery. Maybe I had too many margaritas while writing it. No, but I think, kind of continuing the conversation that we were having, it means the goal of your content has slightly shifted. So for years we optimized for the click. We wanted that clickable headline, that curiosity gap, some meta descriptions. I haven’t read a meta description in a really long time. But I know in the past I felt like I was always optimizing for somebody who would choose that over the one above it, because I used a better word or something. Get them to the page, then worry about the rest. But now if an AI system is summarizing your content before anyone even clicks, then that’s what matters: whether your content is clear, extractable, and trustworthy enough to be cited accurately. And the model isn’t going to pick your headline just because you spent 30 extra minutes on it. It’s going to read all of your answers and decide, is it even worth repeating? So I guess in practice it means: get to the point faster. Answer the question directly before you add context and nuance. And again, these are all not new things, but use clear language over clever language and structure your content so that the key insights aren’t buried paragraph after paragraph, but at the top. And I think what’s most important is that writing for robots, like AI, doesn’t need to be at the expense of humans. It just means writing with enough clarity that both can extract the value from it easily.

Sandra: For me personally, it kind of reminds me of the techniques that we learned for writing a press release, because I have a background in PR. They taught us that a journalist doesn’t have time to read all of the content, so it’s super relevant to put all the answers and all the relevant facts in the first paragraph, and then you can go into more details if you get their attention. Probably you learned it in school: building some kind of attention in the content that you write, not giving all the context at once, but building it up and explaining it. That’s not going to work for a press release. And that’s very similar to what you’re describing for how to write for an AI to understand your content.

Emma: Yeah. Similar to mine, I didn’t jump straight into marketing. It was more journalism in my background. And I think it’s easier for me to go back to that. Now, I have my personal go-to sources that I trust more, and I weigh what they say heavily versus something else. That’s how AI is working. Whereas SEOs are having to unlearn a lot of stuff and relearn this PR way, this journalistic way. Because there’s so much ambiguity in each of them and how they’re weighing this source versus another.

Sandra: And probably also the content formats that work best shift a little bit, because AI loves those tables, overviews and these kinds of formats. I mean, if you really did a good job at writing good SEO articles, that probably was even part of that. But I think there’s also a change in general about what good content actually looks like in terms of its characteristics and formats.

Emma: Yeah. Especially stuff that’s getting cited. I agree that direct answers have always been something that we’re trying to do. But I do think that now it’s more possible to get your original data and research picked up. So, if you can pretty much publish anything that somebody else couldn’t do, because you have either this opinion or this internal data source, these I would say are a new format that’s getting picked up more. It’s not a new format, but AI search is liking it more. We’ve always done structured content. We’ve always done FAQs. We’ve always done “get straight to the point.” But I think AIs are liking this heavy data, opinionated, your-point-of-view type stuff, because it’s something it can’t create on its own. And those are the new kinds of formats and characteristics, I think, that I’ve just been noticing more.

Sandra: Absolutely. One of the most fascinating statistics that you shared in your presentation was that roughly 85% of AI citations come from outside of a brand’s own website. So that’s a huge shift. And first question: if the citations do not primarily come from a company’s website, where do they come from?

Emma: Yeah, I was also like, oh damn, when I read this. All this work we’re putting into a website. But it’s just third party sources. So publications, review sites, forums, social media news, other people’s blogs, your partnerships, YouTube. Anywhere that isn’t owned media where you’re talking about yourself from your own domain. And when you think about it, 85% makes sense, because you can’t own so much of the landscape. So 85% I kind of can come to terms with that number. I was just like, I now need to pay attention not only to our content, but everybody else’s. But when you think about it, AI systems are all trying to give, they’re allegedly all trying to give, trustworthy and unbiased answers. So if every citation came from our own brand’s website, we would basically have AI serving up marketing copy as facts. And these models are supposed to weigh more towards trustworthy and independent sources, because it’s supposed to be more credible. But I actually remember at a WordCamp a few years ago, a few of us were walking outside to the conference, and someone said, so I spoke to a liar yesterday, and I was like, oh my gosh, what happened? And they were like, oh, they said they have 100% uptime. And I was like, how? That’s BS. There’s no way you can have 100% uptime, from a hosting company. And he was like, yeah, but they said, if you don’t consider this or this or this, but they say it. So I was like, oh man, imagine nowadays if everybody was like 100% uptime, but then in secret saying, if you don’t consider this or this or this. It would just be lies on lies. So when I look at those parallels, I’m like, the places that are doing some heavy lifting where they let you know that it’s not actually 100% uptime, like Reddit threads or Trustpilot reviews or YouTube videos, that’s where real people are talking about real opinions and real experiences. But yeah, it’s funny that we spent so many years trying to get people to our website, and it turns out we should have just been everywhere else.

Sandra: Yeah. I was just about to ask, what does this mean for marketers? I mean, it’s a difficult enough job to manage all that in your own environment on your own website, but you can’t really control everything that’s happening out there. How does that change our job?

Emma: Our scope just got a lot bigger and a lot less controllable, which is really uncomfortable for most marketing teams. For a really long time, digital marketing kind of had fairly logical boundaries. You owned your website, you owned your own channels, you ran your campaigns, and then you measured whatever happened within those walls. It was clean, mostly attributable, and manageable. But we can’t follow this model anymore. Like we were talking about, the conversation about your brand is happening outside of those walls, on Reddit, on LinkedIn. And if you’re not paying attention to that, then you’re actually not managing your brand reputation anymore. You’re just managing your own content, which is 15% of it. So practically, having share of voice in these third party spaces needs to become a metric that marketers actually care about and resource for. Not just like, are we getting press or are we getting mentions, the old way with PR, but in places that AI is listening. It’s a much bigger job. It’s a lot more collaboration. I feel like I have the matrix running in my head. I have our goals, plus each team, each channel’s goals, and then they all have to interconnect. And then it’s like, did you ever see that movie The Adjustment Bureau?

Sandra: No, I don’t remember when it came out.

Emma: All I remember was like, they had these lines of where your life was supposed to go, and every little thing you did either moved it this way or this way. And I feel like I’m always just tweaking things like this. So you kind of have to have somebody with… to be honest, I have ADHD. I think it works really good for people who have brains like mine because it keeps it all in check. You have to know PR, you have to know SEO, content, socials, community. But then it also requires letting go of the idea that you can actually control the narrative. Everything has to go or it’s not going to work. So I don’t know. It’s our time, people who have ADHD, to pull it all together.

Sandra: Yeah. And like you already said, you cannot fully control it. Especially with platforms like Reddit being so important. I mean, I think it’s kind of obvious why Reddit is cited heavily, because it’s not the company itself telling their story, but others sharing the story they had with a company or a product or whatever. But I think this is one of the examples where it’s really difficult for marketers to actively participate, because the platform is meant and built to not be for marketers sharing their content, but for actual humans sharing their experience. So is there even a point, or what’s a good way to have Reddit in your strategy?

Emma: Yeah, I wish I knew the 100% reason why Reddit is cited so heavily. I have all of my assumptions and guesses that kind of reinforce each other. It’s been around a long time, so there’s a lot of volume and longevity there. And I agree with you, there’s a lot of authenticity. Their reputation, you can do upvote and downvote, and the community moderates all of that. So anybody who puts promotional content, it’s like, no, you’re out of here. And if you see somebody get vetoed out because they said something bad on some subreddit, I feel really bad for that person. Because Reddit’s answers are weirdly beautiful and specific. And I think it covers a lot of things. So when I hear brands saying, oh, we should do Reddit, I’m like, but don’t ruin Reddit. Like, come on, guys, this is my safe space. I work in marketing, so I don’t trust a lot of stuff, but I will trust some random username that tells me I shouldn’t buy, I don’t know, this health tracker ring, because they tried it for three months and they hate it. So because of these safeguards that Reddit has and because it’s been functioning like this for so long, it’s hard to infiltrate it as a brand. Which is good. The communities have been dealing with it for a long time. But if you show up with this corporate voice and super promotional language, or some obvious agenda, you’re going to get called out publicly. And to be honest, that might be worse than not showing up at all. Trying to undo all of that bad press is hard. So the easiest way to do it is: don’t send marketing to Reddit, send people. We do do Reddit at Hostinger. But the people who want to do it. We have employees, we have some of our heads of product, our experts. And they go in and they try to answer some of the questions that people are struggling with. It’s not like, hi, we’re Hostinger, we’d love to help. It’s like, oh, crap, that’s broken. This is how you can fix it. And then maybe we update some content on it so that people could have done it themselves. But I think a good rule of thumb is: would this comment be useful if there was no brand attached to it? And if yes, it’s probably fine. But if the only value is your brand being mentioned, it’s just marketing pretending to be participation. And the community will smell it immediately and then you’ll regret it. So it’s a long game. Don’t jump into it and expect a bunch of citations. No, actually, just do something good.

Sandra: One funnel that’s probably a little bit easier for marketers at least is LinkedIn, which I think Perplexity for example is heavily relying on. So do you think personal branding or employees being on LinkedIn is becoming even more important nowadays?

Emma: I wouldn’t maybe say more important, but it’s not either/or. But I would say personal branding is punching way above its weight right now, in a way that companies should pay attention to. So, for example, I reshared a vibe coding stat piece that we wrote on our tutorials, on a LinkedIn post, because I just thought the statistics were very interesting. And I’m ranking for my LinkedIn post on Google and on Perplexity, but not the original article URL. So maybe because I talk about statistics and content marketing, it’s given me a higher bar than this article. I can’t figure out all of it. It was just repackaged advice. But with personal branding, I think AI systems right now are calibrated a little bit more towards authentic human voices. So if somebody shares their point of view and they have a track record, that carries that credibility over time. That’s something that a company page, especially on LinkedIn, can’t replicate. Your LinkedIn company page is almost understood to be like a marketing channel, but your personal profile is like a trusted source. So right now, I’m trying as well to push the people in your company who actually know things to share your content, and not from your brand account. You can reshare it from your brand account, because there’s that compounding effect. But I think it’s more beneficial if a person builds authority in a place, then they’ll bring the brand alongside with them, not the other way around.

Sandra: Yeah. You’re already hinting at the next thing I want to talk about. So far we’ve covered search, websites, technical foundations, content, and these kinds of things. But what I took away as the biggest challenge and the biggest shift is maybe the organizational one. AI significantly shifted the responsibilities not just in marketing but in the entire company, because historically, there was maybe a social media team, somebody responsible for PR or maybe an SEO agency, a person, a content team, and everyone had their own KPIs and could work next to each other. And today everything seems so interconnected with each other. And the first question that comes up is: who owns AI visibility, or who should own AI visibility inside an organization? Because it doesn’t seem like there’s this one discipline that somebody can be responsible for. It’s kind of like everybody’s involved.

Emma: Yeah, that’s the million dollar question these days. To be honest, nobody owns it clearly or cleanly. Most companies are probably doing something different right now. But that’s a problem and an opportunity. Because this AI visibility sits at the intersection of SEO, content, PR, socials, and brand, and not to forget to mention the alignment across partner teams and product. It does actually require inputs from all of them. And like I said earlier, the outputs benefit all. And if one person doesn’t pull their weight or isn’t accountable, then it could ruin it for everybody else. So where most organizations are failing is they fall into this gap where teams are waiting for somebody else to pick it up, while everyone’s quietly hoping that it’s going to resolve itself. What we do is, we have an appointed person, like a DRI for it. Sometimes it was the head of SEO. Then it was our organic marketing strategist. Now we have just a DRI, and they built a cross-functional working group. So they have their squad, they work with every single one of these teams, but one person drives it and makes sure that it keeps moving. We tried four different versions of it. It wasn’t just like, oh, we got lucky on the first time, which is why it was head of SEO, then one person, then somebody from organic. But my honest opinion is that it should live close to whoever works on organic strategy, because there’s a skill set and a mindset, and most of the metrics overlap heavily. But they cannot be siloed. Like you were talking about, everybody used to be able to work in these teams. Now PR has to be in the room. Content needs to be in. Brand for sure needs to be in the room. You need the key stakeholders to really understand the importance of this, and also understand that if one falls, we all fall. What I’ve noticed and heard a lot at WordCamp after this talk was everybody agreeing to do it and everybody agreeing it’s important, and then nobody being accountable for it. And then they’re like, why didn’t it work? “You’re the DRI.” And it’s like, yeah, but I can’t actually go in and do it for you. You have to do it. So there’s a lot of education, evangelism, educating the company and the team on why it’s important, and then having the conversation earlier.

Sandra: Yeah, it also makes it a lot more difficult, for example, for agencies who before could just add an SEO service on top of building the website. That could include the technical foundation, maybe also content, keyword kind of tasks, and these things. But now this has become just such a small piece of the entire big picture, that I suppose it also requires agencies to rethink what exactly they can offer to a company. Probably they can’t offer “we will make it visible in AI citations,” because that just doesn’t work.

Emma: Yeah. I think it is harder for agencies. And there will have to be some evolution in the way that you offer different services. I was kind of thinking about this, and where they could make a little pivot is brand coherence auditing. You could offer not only to build your website, but actually go across the surfaces that your clients exist on, whether it’s their own content or these third party mentions, even press coverage and socials, and actually start to map out what AI actually sees when it looks at them. I don’t think a lot of agencies’ clients would have done this, obviously, because they want the agency to do it. And I think they would be surprised what they find. So a nice high-value service that could be added. But something a little bit more on the websites and technical side, for the ones that maybe write blog posts or help with SEO, is offering more content architecture consulting. So not just “we write your blog post,” but “we’ll also help you structure your entire content ecosystem so that it’s readable, sizable, coherent.” And then in addition to your linking strategy and your cluster building. I think that’s a lot more like strategic engagement in addition to the production. And then, I don’t know if a lot of agencies want to do this because people go to maybe social media agencies or something like that, but try to help them build their presence in these third party spaces. So even if it’s like, we’ll write your blog post, but we’ll also help you write what you should write on LinkedIn or on your Reddit posts. This is where that PR and SEO community management could really work together. But I’m not sure. I would guess that most agencies aren’t set up to do it well yet. They just have to be curious and start to see, okay, I can actually handle this.

Sandra: But this is also a great opportunity, because a lot of agencies, especially web agencies, are struggling at the moment with, okay, where is this whole AI thing going? What will my future look like? How do we need to change our services? And this might probably be a good opportunity for some of them to change their offer in a way that AI is not a threat anymore, but can actually be a new part of the services.

Emma: Yeah, I agree.

Sandra: You mentioned coherence a couple of times. One of my favorite messages from your talk was: incoherence is invisible until AI makes it loud. Could you explain that?

Emma: Yeah. I swear, I think this is like, Margaritaville. I was like, that sounds so good when I was writing my book. But what all of that mumbo jumbo means is, the idea came from something that I just kept noticing, that brands have actually been quietly inconsistent for years, but it didn’t really matter all that much. We all kind of had this different messaging on our website or on your branded pages or your press releases, but they were all kind of saying the same thing in a similar field. They just maybe had slightly different value propositions or slightly different target audiences. But it was just this slightly different version of the original story. It was just fuzzy. Nothing catastrophic. And for the longest time, this fuzziness was invisible because we as humans are remarkably good at filling in gaps. Even in the way that we speak sometimes, a lot of us just don’t finish the sentence because we know that the other person can fill in the rest of where we were going. But you’d read the homepage, maybe you get a vibe and you move on. These inconsistencies existed, but nobody was synthesizing everything simultaneously and then pointing out all of the little contradictions. But then enter AI, and AI does exactly that. It pulls all of these sources that have already existed, synthesizes it into a picture of who you are. But now, if those little sources are telling just the slightest bit different stories, the picture that it’s producing is fuzzy or blurry at best. And if it’s contradicting itself, that’s the worst case scenario. So incoherence was always there. It’s just now AI has like a boombox, a loudspeaker, saying it to everybody being like, they messed up.

Sandra: Yeah. Also, we all know it. Either we’re an agency or we are working in a company with a lot of websites. There’s that one landing page that somebody created for a specific project, and then it’s completely forgotten, which is not just a security risk, but now there might be some very outdated messaging on the landing page that nobody even knows exists anymore. Or like you mentioned Reddit, some customer providing wrong information on a post two years ago that we might not even know about. So I completely get how this is a completely new challenge. And one that’s particularly interesting, for example, for our customers, because many Greyd customers are enterprises, universities, franchises, or multi-location businesses or organizations that often have dozens or even hundreds of websites. So what I took from this is that content governance has just become way more important, because nobody might have noticed that outdated landing page that everybody forgot about. So companies, especially the bigger companies, should put a much bigger focus on making sure their content is consistent across all their patrons, all the websites.

Emma: Yeah, 100%. Content governance is one of those things that sounds super uncool until you understand what it’s protecting you from. It’s the system that will ensure that everything is accurate and consistent and on brand. And I would say that some organizations have weak content governance because it’s just unglamorous infrastructure, the content police, and it’s very easy to skim over that when there’s something more urgent to prioritize. I would say content governance is like the simple answer. You can have a brand book, you can have a style guide, but then you need somebody to point out and direct people if they’re kind of going off course. And there was this one sentence I used to use as an editor, when I was teaching writers and editors: if you keep changing one word over and over again in a sentence, eventually the meaning will change. You don’t notice it because you are the one that’s typing it. So I used to use the sentence “a blue boat sailing across the ocean.” It’s like, okay, this is a very cute little blue sailboat sailing across, you picture something sunny. And then an editor comes through and they’re like, I don’t like blue. I’m going to change it to navy. And then they do another round and they’re like, actually ship sounds better than boat. And then they read through it again and they’re like, let’s just change ocean to sea. So when you started with a blue boat sailing across the ocean, all happiness turned into a navy ship sailing across the sea. You’ve got this boat going to war, but in your mind it’s still the same thing. So I think that’s what makes it so uncomfortable for brands. They’re like, this inconsistency was never intentional. It’s just this drift that was happening. So even if you have this content governance, your messaging can evolve, priorities can change, your product can change, what you do can change. You just have to be aware of that now and try to fix it, because it’s still living out there. But it also changes a lot of how we saw things. Some people have blogs which historically are supposed to get outdated. Like, from ten years ago, it’s okay, you know, but I said that in 2005, that’s 20 years ago. Geez.

Sandra: Ten years ago it was still 1990, right?

Emma: Definitely. But how blogs are supposed to be outdated, and AI doesn’t see that. It’s like, no, they said this in 2005 on their Myspace channel, like, it’s facts. So now we have to do this technical governance and content governance over the stuff we’re keeping alive now. But we also need to maybe go back and see all of the things that we said in the past, which is something we’ve never had to account for. So yeah, I went on a circle there. Sorry.

Sandra: No, I mean, it just emphasizes that. And that’s the one thing that everybody can agree on, that AI will just change a lot of how we work, what’s important, what parts of the job become more important, or where we should put a little more focus on. Let’s do a quick fire round. I’d like to mention a couple of topics and would love to hear your take on whether this is something that companies should keep, ignore, or maybe even prioritize. Okay. So let’s start with core web vitals.

Emma: Prioritize. I think speed, stability, there’s a penalty for ignoring them that’s not so quiet anymore. So definitely prioritize.

Sandra: What about internal linking?

Emma: I’ll prioritize. It’s free. It’s powerful. Most sites are doing it really bad or unintentionally. So maybe do a whole audit of it.

Sandra: llms.txt?

Emma: Keep it. Don’t let it distract you from your fundamentals. But I think it gives a good signal of intent. It’s not a visibility strategy, but it gives you good signals. It probably also has been a little bit overhyped. So I keep it, but not put all the efforts on it.

Sandra: What about the importance of sitemaps?

Emma: I would prioritize it. It’s again this uncool thing, but it’s essential. It’s not hard. It’s just underrated. Just do it and then forget about it.

Sandra: We kind of already covered it. PR?

Emma: Yeah, definitely prioritize it. I think your PR strategy and your AI strategy are the same.

Sandra: I think a lot of my former colleagues will love that, because over the last couple of years before AI, I think PR was in some kind of crisis because it seemed like an easy thing to save some money on and to skip and just focus on the other marketing parts.

Emma: So they will love this. They should know their worth and be like, remember me? Fired me. Yeah.

Sandra: What about video transcripts?

Emma: Prioritize. The transcripts for sure, immediately. If you have any videos without transcripts, you’re leaving discoverability on the table. So just either use AI or something. Definitely prioritize it.

Sandra: I was just about to ask, because at the beginning, when this came up with accessibility a couple of years ago, that you need to have those transcripts. At the beginning it was just so much work. For example, if we have a conversation like this, I remember it took us hours to go through the transcript and optimize it for an hour video. But today, I think most AI tools are very good at that. So it’s not much work to fine-tune a little bit and just go over it and see if AI got anything wrong. Other than that, it’s kind of easy now.

Emma: Yeah. And I would say that we’ve also become more forgiving about it. Like if a word is not 100% correct, it’s okay, it was AI. We used to be so judgy about that in the past. And now it’s just like, oh okay, typo.

Sandra: Yeah. What about content hubs, like topical authority kind of thing?

Emma: I would say definitely prioritize that, because that’s what AI systems are trying to identify, if you have this authority. So if you can give them any type of clear signal, don’t just scatter high volume pieces, like we used to. Nobody cares about that. That’s a strong thing to say, nobody cares about traffic. But do a content hub, get to this topical authority, and then go on to the next one.

Sandra: That probably also provides some guidance for all those small teams that are sitting out there and saying, okay, there’s so much new tasks and additional work, like how do we start? Where do we put our focus on? So would you say, rather than trying to cover a lot of different topics and be cited for a lot of different topics, focus on the ones that are most important and really put some effort in there to make sure that you’re cited on that? That’s probably the better strategy than trying to go broad, right?

Emma: 100%. And I would even go even more granular than that. Just do one thing and be the master of that, and then go into the next one. If we had to start over at Hostinger, I would do that for sure. It wouldn’t work the old way.

Sandra: In general, that might be good advice. Let’s say there’s a company that, as always, has limited budget, resources and time. What would you suggest as the first three things they should take care of, maybe some short term things, and maybe things they should think about or implement on a more long term basis?

Emma: That’s a good question, because I think I didn’t tell you to ignore anything. I was like, prioritize, prioritize everything. So if everything’s a priority, then nothing is. Okay, so three things to do like next week. Oh okay. What would I do? I think the first thing I would do is crawl my site. There’s free tools, you don’t have to pay. Like Screaming Frog has that free version, Google Search Console is free. But understand what’s actually indexable, what’s broken, what’s blocked. I don’t think you can make any good decision without that visibility, without knowing what AI systems can and can’t access. So it could take like one afternoon, one person, and then you’d have basically your plan for the rest of the week. And then I would Google myself. That’s funny enough, we’re talking about AI search, but I would Google myself, and maybe also on AIs as well. Like a vanity search, you know, but a very systematic one. Search your brand name, whatever your product or where you’re selling, your categories, your use cases. And then look up what comes up on page one on Google, on what AI says about you, all the different AIs. And then start to map out maybe your gaps between how you describe yourself and how the internet is describing you. And that gap will be the start of your strategy. And then I would say, on the more content side, like this content hubs clustering thing we’re talking about, find one internal expert and get them to start publishing things, so that your brand can be known for that thing. And don’t do it from your brand account. Get your head of product or one of your executives or something like that, a real person with real knowledge in your space. And then start being consistent about giving their point of view, sharing your insights. And I think that will slowly start to compound while you’re doing these things from step one and two. Those things could be the highest leverage but lowest cost, like crawl audit, brand gap analysis and expert voice. It can all be done in one week and you don’t have to put any extra budget into it.

Sandra: Yeah. Maybe just make sure that, that’s the stat I think I see a lot of the times, that a company tells an employee, whoever that is, head of product or whatever, to also do some content, also post on LinkedIn now and then, but they’ve completely forgotten to also allocate some time for that person. So the natural thing that’s going to happen as soon as something big comes up is that that part gets completely forgotten. “I didn’t have time to post” or whatever. And sometimes they might also need a little support, because especially when you’re from a tech company, that’s also something that I experienced personally. If you’re in a tech company, the people who have the knowledge are probably the best to share that knowledge. They’re not naturally people who are used to sharing their knowledge from a content perspective. So they might not really know what to do, where to start. So maybe provide them a little bit of support, not in terms of phrases like this, because then it will just end up like marketing copy again, but getting them started and giving them a little support on what they should actually do. What you mean by “can you put up some content there?”

Emma: Yeah, definitely. And even just helping them structure their content, like unformatted, or put it into words that maybe somebody not so technical could understand. It definitely gets easier over time. But I agree. It’s one of the first things that will get cut if they don’t see immediate return on it. But it’s important.

Sandra: Thank you so much. I think that’s been a lot of very valuable advice that you shared. I feel like we could add three more sessions at least and dive into a lot of these topics way deeper. But I think we’re at an hour here, and like I said, you already shared a lot of very valuable content, both for companies with products and for our agencies. So thank you very, very much for joining us today.

Emma: Thank you. It was fun. I love to talk about this. So if you want to do a re-record, just let me know.

Sandra: I was just about to ask, if people would like to follow your work, where should we point them? Where can they find more and learn more from you?

Emma: Yeah, LinkedIn. I go through phases. Sometimes you’ll see like 50 posts in two weeks, and then I just won’t be on there. But yeah, LinkedIn is the best. I put a lot of my internal thoughts into my posts. Not all of them are tested and validated. So take that with a grain of salt if you want to find me. But yeah, that’s probably the best way. Just Emma Sophia.

Sandra: Perfect. We’ll put also the link in the description, for everyone who will listen to or watch this episode. Thank you very much for being here. If you enjoyed it and haven’t yet, please remember to subscribe, share it with the team, and let us know which topics you’d like us to cover next time. The next episode will also be a very interesting one, which will be about WordPress products and how they can survive or even excel in today’s economy. So I can highly recommend that episode as well. And again, Emma, thank you so much.

Emma: It was a pleasure.

Sandra: And see you in the next episode. Bye.

Music