Raise your editorial standards – with Alex Moss

To stop making fluffy content, stop scaling content when you don't need to. Just make things that are concise, make sure the formatting and structure of the piece is actually there. Read more about journalistic editorial guidelines, like maybe look at what the FT think about guidelines when they're writing content or other top publishers and try and get as high up in standards as possible because the lower end standard, the billing something just for the sake of filling something. So I don't know, let's take the death of the browser as a good topic, right? Maybe I could write about that in 500 words, but don't force me to write 1200 words for the sake of making it a bit longer. If I only have 500 words to talk about it, then that's it. It's concise, it's well structured, it has its points, and it answers some questions that can then be synthesised into an answer. Hi, I'm Alex Moss and this is SEO in 2026. Alex, what's your number one SEO tip for 2026? Be concise and raise the game of your editorial standards is what I would say is the best tip for 2026. Why be concise? Because obviously we've perhaps lived through the era of creating as much content as possible. So why is being concise and perhaps less content nowadays more important? Well, machines now and LLMs and agents that they're all ingesting content using tokenisation. The more tokenisation that's used, the more processing power is required and therefore it's kind of like the content version of site performance and site speed at this point. If there's a slow loading site, then it's going to take a while for people to load which might impact UX and other metrics such as C, D, R and things like that. But when it comes to content, content content, say, which is a good phrase you use because that's was a bit of an old and gray hat technique of just galing content and just doing any old thing to get as much exposure and as many keywords as possible. But now being a bit too detailed doesn't always help. I think the key here now is to just when I say be concise is don't blabber on too much. I would say is the informal way of saying it. I think another good example is, and I'm sure you'll have found it too, especially with tech or startup kind of sites. You go to the about page and like what we do and you may read four sentences and you still don't know what they do. They talk it like they use very colourful words, you know, fluffy kind of stuff, sails the words to impress the human. But right now the audience is not just a human, machines are the audience as well and they're ingesting that content and they don't get sold by that kind of language. They get sold by direct language, things that get to the point, answer the questions, solve a problem. And that I think is the way in which content in general should be going. And I think we've let ourselves down maybe over the last few years by lowering the standards of editorial in general so that anyone could potentially make content say and I think now I think this is a very good thing because it'll take away content mediocrity and raise those standards not just in general content of everyday sites that we all work on, but even as far as you know journalistic and editorial online, we'll have to make sure that they're adapting to higher standards in order to have the best exposure. Yeah, I mean in terms of letting ourselves down, I think we've just been chasing the algorithm because 10 years ago we used to be able to say okay, right or 500 word article or 500 or a thousand word article that includes this keyword phrase and it's in a mixture you multiply, you know, add some LSI keywords or related keywords and do it for $20. So that's what I'm looking for. I'm going to publish that once a day and that works. You know, that could get you into the top page of Google, but it doesn't work nowadays and it's hard not to do something if it works. Yeah, and I think before everyone was not only chasing the algorithm, but let's face it, someone just trying to find a loophole and exploit the loophole in order to rank game visibility and get some extra sales or conversions and whilst that's good in the short term, that's essentially it's just a black hat technique. It's low end automation at scale and we've seen enough over this year in 2025 if people trying to scale content with AI and you know, the thing that annoys me is on X sometimes you'll have someone going, oh, I scaled up content and look at this, look at this graph that's going up isn't this great and I favor it them because I check in about six months later and see how it's going and not one of them are still gaining momentum or growing. They've all been penalized in some way or whatever scale they've done has to be gale back because it just hasn't worked in the long term. So what happens if you work for a business that has a lot of content that's been publishing for the last 10 or 15 years or so is the first step to actually review the existing content that they have and prune a lot of that? Yeah, I mean, that's kind of sounds quite familiar when HCU first started out right. There was a lot of content, cleaning, like, you know, bringing some pages together. If you've got like five pages that are kind of covering the same ish thing, but you were doing it content content sake back in the day, maybe now it's better to consolidate those things into one long form piece of content. Again, that's more editorially concise because you would take away some repetition and some filibustering in there just to pad out words and try and fill some keyword structure. But now it's not just about how many keywords and how many words and clustering things like that. It's not just about that. And internal linking has always been important, but it's even more important now for the flow of LLM's as well. And to make maybe not just a complete pyramid shape that keeps on going down and down and down to 7, 8, 9 levels, but maybe flattening out your IA a little bit more. So it's easier for a machine to navigate as well as a human. And that works with content as well, making it a bit more flatter in terms of going, not going into some some topic and then subtopic and subtopic of that and subtopic of that. Not many people can do that in a legitimate and concise way. So bringing that together and also removing any content that is a waste of time really to the internet as a whole is going to be a good thing. Because it is kind of, I would say some of it's like web spam. I've been talking a bit like this year about remember I'm sure everyone knows about the cold play concert where those two people were caught cheating. And the internet for the next four weeks was just full of it. And I thought that was a lot of web spam as soon as you've got like three or four publications talk about it. You don't need and then that annoying thing of brands started doing their own take on the comedic level of things. That to me is just a waste of time and to an agent and to an LLM is also a little bit of a waste of time as well. You just need to understand the source what happened and since I an answer. But we go over the top to try and get those visits and try and get some of the conversion going on. And I just don't think it's going to work as easily that way in the future in the longer term as well. So talking about LLM's you also say structure your content as if another lens your primary reader. And so what does actually mean that that actually mean in practice in terms of what you actually do to your content structure? Well firstly and it would be obvious anyway. And I've like I'm sure not just me but a bunch of SEOs have been saying in the past like make sure all of those H1H2s are all in there but it goes a little bit beyond that. Anything that can add formatting to text that may be relevant inside. I know it sounds weird but things like italics and bold and all and tables structure is all quite important now because LLM's ingest that they're not just reading words to understand those things. They also read a lot of mark down as well. So if you may be in a niche where where you're in a more technical vertical then it may be more useful to make mark down versions of that because that's what LLM's love. I think it's the one thing that's a bit unanimous in the world of AI is that LLM's like to ingest mark down more so than any other output including markup which to be fair is kind of mark down that's been made to be more appealing on the human eye but again the human eye isn't the only thing looking at our content anymore. There's a robot looking at it and they look at things and again when it comes to tokenization mark down and good structure helps the tokenization and you can go to I think there's a tokenizer open AI have their own tokenizer and you can like try it out for free when you copy and paste words and it shows you and highlights the way in which it tokenizes things and that's a good calculation you can use that as a good test. So being able to reduce the amount of processing power that LLM would have to do to ingest or all the content on a given page or a given site is going to be advantageous to save as much as possible on because LLM's they're cheap right or that's their mentality they'll try and find the cheapest way complete a process because that's the most efficient and cost effective way of doing so and fastest as well. And you also advocate exploring and experimenting with new open standards like LLM.txt. Well, LLM.txt is not standardized like robots.txt and is today or maybe an XML site map. It was a proposed standard written by someone who Ryan Jones, Ryan Howard I think it is Ryan Howard is based in Australia he's got one of those AI startups and he created the proposal and weirdly it's um to describe LLM.txt and then not shell think of it as just a summary of your site with good on-site references that covers what you do, you know, like what you're selling, what you're offering, and it's just kind of like a summary. That's all it is, but it covers everything that it needs to in order for an LLM to then dig deeper. Think of it as an open house when you're looking at a house or a first look at a house. You go in, you can check out all how big the house is, you can kind of gauge it, but you're not asking certain questions about what's inside the walls or what the roof made else of quite yet. That may be on your second visit when you delve further and investigate further into what you might once purchase. And lm.txt may be a version of that. And whilst some poo poo it may be by saying it's the new meta key words or it can be abused, I would counter argue that by saying that everything can be abused in the world of the web and owning a website. And of course, if you produce your own lm.txt file it will be subjective, but as long as you're honest and open in it and you're not trying to abuse the lm to make it do something it shouldn't, then it's a legitimate thing to do. And there's been a lot of discussion about whether it should be included. And my answer to that is, the longer you discuss it, the less time you're spending time on producing it. It can take you unless of course you've got a huge site that where it would take some time to produce an lm.txt file. If you've got a small site, it takes about 20 minutes. And in Yoast SEO, in free version, you can generate it automatically. So you don't have to do any editing whatsoever. And let's say you post 20 posts a day, it will dynamically update that. And whether or not a lot of lm's out there using it to me is immaterial. It's there and I consider it some insurance because the moment that one lm may do it, and I'm not saying that anyone's at the forefront of doing it, if one just says, you know what, we're going to do it right now. Everyone who has been criticizing lm's.txt will very quickly shut up and they'll also very quickly implement that idea. Meanwhile, me and all everyone who's got Yoast SEO who's enabled it already have it there. So it's like, it's just a form of insurance. It's there just because and it's there because it's an additional bit of help should someone need it or something need it. And in relation to that, why should SEOs work with mcp's? Oh, well, that's, I mean, I guess that's a different question to lm's.txt, but mcp's. Yeah, new technology, new standards, and obviously, yeah. Yeah, I mean, for an SEO, I'd say the best way of using mcp's is to connect data and be able to interact with it in a more conversational level and be able to understand big data in a way that you don't have to be, you know, a super advanced data analyst. You can ask it questions. It can analyze that data. And then it can bring it you back in again, in a natural language manner that you can then understand more and ask more questions. So for example, you can connect search consoles mcp to Claude or perplexity. And I think hope in AI do it now. I'm sure they do it now. You connect it, it's like putting in a USB thick into your hard drive of the lm. And now it has all of that extra data. In a previous talk, I, I compared it to in the first matrix when they were trying to escape on the roof. And Trinity needed, needed knowledge on how to operate a helicopter in order to escape. So with the pilot, on Nebuchadnezzar outside the matrix, literally uploaded all of that knowledge directly into a brain in like that. And then all of a sudden she knew exactly how to work and operate. The helicopter. Kind of like that. If I uploading all of that data and knowledge into a larger lm. So it could contextually understand that data interact with you on it. And you can do not just search console, but Google Analytics now have one. I don't think Sam Rushdo, but a hrefs do, Sistrix has one. And then you can connect all of those things and ask it questions about multiple things. And I think that's a really good thing. Because then it can also maybe answer questions that are meant for more non-data focus people, maybe more in the marketing team. In the content team, they may find it useful. For example, what was working this time last year during Black Friday? What made the most conversions? What were the biggest pages that were attracted this year that maybe I could use? What didn't work? And it can read the data and it can analyze it. And as well as that, you can upload your own data. So let's say for example, you had a Shopify shop. And in there, you could export all of your orders and then upload it as a document and then ask it questions about the order whilst correlating it with anything that the MCP's connected to. It sounds like because these machines will be able to make sense of so much data, so much content that there might be a chance that it's not necessary to be quite so concise with content moving forward. Do you think that might be the case? Do you think the fact that we need to be concise now is a relatively short-lived requirement in order to actually be more likely to rank higher in the next few years perhaps longer content might come into fashion again? It might, but again, long content can still be concise. It's like research papers, highly detailed research papers. If they're 20,000 words, it's not like saying that 8,000 of them is fluff around the 12,000 important words in that. It's still all concise and structured and done so. Like white papers are a good example of keeping to an agenda, not diverting too much away, not waffling too much, which is why do, per-bally, it stays on track and stays in its lane. And again, that's ingested in a better way. But in the long term, things are unpredictable, right? Like, I've had conversations with some people internally in Yoston engineer who is convincing me more that the browser may die in the next few years. And that's an interesting thing, because the browser is made to bridge a gap between the content and styling and the human who's looking at it. But now that a machine might be the primary audience of who looks at your site in the next couple of years, I'm now thinking that websites may, and especially going back to Markdown, think of a Kindle. You've got a Kindle file. You can put 10,000 bucks onto this one device, and it's outputting a very flap way. It's text, but it's got all the formatting with it, to go with it. But they're just text files. That's essentially Markdown. And I believe that that's what apps and LLMs may turn into is a Kindle management service for the Open web, where they pull information off on a more basic level, and then merge that with media, for example, that you may have inside. But I don't think humans are going to be looking at websites in the same way that we have been. And I think that's a shift that we're all going to have to adapt to in the next couple of years. Perhaps. This is theoretical, right? The browser may be along lived another 25 years, but I just believe that the way we're ingesting content and sites and searching for things and discovering things at the moment, whilst it's in stepping stones, is going to change in the longer term. Interesting answer. And I like your thoughts about Null Fluff in content, and it doesn't necessarily mean that you shouldn't have lengthy pieces, but as long as it's not padded out with lots of fluff. And it kind of reminds me of the answer that I have when people ask me what's the best length for a podcast episode, as if there's a magic number of 20 minutes, 25 minutes, or something like that. You can say that, OK, the average commuters is a mount, so maybe you should aim for that. But generally, my answer is, as long as it needs to be, it could be a nice, concise 10 minute episode is enough for the message that you want to convey. But it could also be the fact that a two-hour long discussion is appropriate for the conversation that you're having. Consistency is best in terms of what audience has in terms of expectations. But there's no definitive answer necessarily. It depends on the message that you want to convey, and that applies to written content as well. Yeah, which is also interesting, because long form podcast interviews are a trend right now. And that very long sum of them, I think, there was a recent interview with Lex Friedman, and someone in it was 10 hours. And in my head, I'm like, how do you even sit and talk for 10 hours? That is a feat within itself. Then you've got to think of, well, who's going to listen for those 10 full hours? Probably not everyone, unless you're very, very, very focused on that interview. And then you'll get summaries or you'll get chop and changing. And I don't know now that things are in chapters on YouTube. You may only be interested in certain segments of the interview. And again, LLMs will be able to take the transcript. They'll be able to take the chapters. Soon they'll probably be able to understand, like, mood, and identify sarcasm, things like that that can, again, add to it. But while long term pod, long form podcast is kind of fluffy, isn't it? Because you can talk about all sorts. You also get unadulterated content with long form podcasting. So if we were talking now for three hours, I think we'd go into much more granular detail about things because we have that time. Here, this one may be 30 to 60 minutes. And therefore, we have limited time. So again, I can't go on too long. You'd have to control me. Otherwise we'd be here for three hours talking about really detailed stuff. But again, it's about just being concise, getting to the point, and LLMs will even, from a long form podcast, will soon be better and better and better understanding what the summary is and where things are important. And more so, once we get to connections, how it's connecting us and our personalities, things like Google Drive, and it understands more and more about us, it'll probably be able to understand those 10 hours, which are the bits of that 10 hour interview with Lex Freeman is actually applicable to me. What am I going to find interesting based on what you know and what you've seen I've been doing with my browsing history and Gmail activity. And I think that's going to be really cool, but it also makes it much harder for us a lot of SEOs because personalization is the norm now. You know, I remember PWS equals zero on Google and when personalization started, you wanted to check the ranking of something. You would go to whatever the search term was on Google and you would just add the UCMPWS equals zero and that meant that personalization is disabled and you would see completely different rankings. But that again, people aren't talking about that because no one's got depersonalization anymore. To the point that even on OpenAI, you can have zero account activity, you can be logged out, you can prompt or ask the same thing as me and we'll get a different result. That's how personalized it is. There's nothing flat anymore and it makes it much, much harder to attribute what's our rankings. It makes it much, much harder for rank tracking to be more accurate and more valuable, you know, everything's up in the air in the next couple of years, I'd say. The shortening tail, echo chambers, the conversation could go much deeper in this particular topic. Hopefully some of the time soon. But let's be different of about things in terms of a takeaway. So Alex, what's the key takeaway from the tip you shared today? Read more about journalistic editorial guidelines like maybe look at what the FT think about guidelines when they're writing content or other top publishers and try and get as high up in standards as possible because the lower end standard, the filling something, just for the sake of filling something. It's concise, it's well structured, it has its points and it answers some questions that can then be synthesised into an answer. Alex Moss was a principal SEO at Yost and Co-founder of Firecast, find out more over at AlexHeifernMoss.co.uk. Alex, thanks so much for being part of SEO in 2026. Thank you very much. I've been your host David Bane. Get your copy of SEO in 2026, the book over at SEO in 2026.com. [BLANK_AUDIO]