In a world where search engines and AI tools are deciding who gets discovered, simply adding more keywords is not enough. Everett Sizemore breaks down the SEO framework that helps brands create clearer meaning, stronger visibility, and content that both humans and machines actually understand.
https://page2pod.com - In this episode of the Page 2 Podcast, Jon Clark and Joe DeVita sit down with Everett Sizemore, a 15-year SEO veteran and fractional SEO director at eSizemore Consulting, to unpack how brands can create clearer, more structured content for search engines, answer engines, and AI-generated results.
Everett explains why semantic triples — simple subject-predicate-object relationships — are becoming increasingly important for helping machines understand what a brand is, what it does, and how its key entities connect. The conversation explores why many SaaS companies struggle to describe themselves consistently, how unclear language creates confusion across search and AI systems, and why “cutting the fluff” is more than just a writing preference — it is an SEO strategy.
The episode also covers practical SEO tactics including content pruning, building content inventories, using tools like Screaming Frog, Google Search Console, Google Analytics, and Ahrefs, and revisiting foundational assets like FAQs, persona pages, and HTML sitemaps.
🧠 In This Episode
• Why semantic triples matter for modern SEO and AI search visibility
• How brands can define entities more clearly across their websites and profiles
• Why SaaS companies often confuse users, search engines, and LLMs with vague positioning
• How plain, structured language helps search engines and answer engines understand meaning
• Why content pruning starts with a complete content inventory
• How Screaming Frog, Google Search Console, Google Analytics, and Ahrefs can support pruning decisions
• Why FAQs, persona pages, and HTML sitemaps are making a comeback
• What SEOs can learn from the shift from traditional search to AI-generated answers
• Why the best AI-era SEO strategies often come back to fundamentals: clarity, consistency, and discipline
This conversation is a practical guide for SEOs, SaaS marketers, and content teams who want to build clearer, more machine-readable websites that perform across both search and AI-driven discovery.
Subscribe to the Page 2 Podcast for more conversations with SEO leaders, digital marketers, and search innovators helping brands move beyond Page 2.
Comment below: What is one part of your website or brand messaging that could be made clearer for search engines and AI tools?
🛠️ Tools & Resources Mentioned
• Everett Sizemore on Linkedin → https://www.linkedin.com/in/everett
• Everett Sizemore Official Site → https://esizemore.com/
• 2016 "Defining and Mapping Your Brand for Google" SearchCon Presentation → https://www.slideshare.net/slideshow/defining-mapping-your-brand-for-google-a-slightly-technical-guide-for-growing-your-brand-as-an-entity-in-the-knowledge-graph/54936502
• Everett's Semantic Triple Recommendations → https://www.linkedin.com/pulse/how-why-create-list-semantic-triples-your-brand-voice-sizemore-8pjbc/
• Cruftfinder → https://youtu.be/6Sf_xkoHIUI?si=p0gFjEH-ZNFqwsut
• LLM Bottom of Funnel Approach → https://www.linkedin.com/posts/everett_at-a-certain-point-someone-needs-to-buy-something-share-7331506379997794304-YXrc/
Jon Clark (00:00)
Welcome to the Page 2 Podcast where we uncover the strategies, systems, and tactical decisions that move brands beyond Page 2 and into real visibility across search and answer engines. Today's guest is Everett Sizemore. He's a 15-year industry veteran and today, through eSizemore Consulting, he works as a fractional SEO director helping SaaS teams. This episode is about clarity and the frameworks that make it possible. Everett breaks down concepts like semantic triples, simple subject-predicate-object relationships that help machines understand meaning, and why most brands fail to define themselves consistently across the web. He explains how those gaps lead to confusion in both search rankings and AI-generated answers, and why fixing them often starts with something deceptively basic: writing in plain, structured language. We also get into the tactical side of that shift. Everett talks about content pruning and why things like persona pages, FAQs, and even HTML sitemaps are becoming more important again. Through it all, there's a common thread. Most of these new strategies are really a return to fundamentals. The industry likes to keep adding layers 'more tools, more automation, more AI 'while Everett is pushing for constraint, consistency, and discipline. And that disconnect between complexity and clarity is what makes this conversation so relevant right now. I've personally found myself rethinking how we structure content and even how we describe our own brand after this one. If you learned something new today, take a second to subscribe to the Page 2 Podcast, leave us a rating or review, and let us know what resonated. We'd love to hear your thoughts. Okay, let's get into it.
Jon Clark (01:37)
Welcome to episode 118 of the Page 2 Podcast. I'm your host, Jon Clark, and I'm joined as always by my partner at Moving Traffic Media, Joe DeVita.
Joe (01:46)
Hi everyone.
Jon Clark (01:46)
Our guest today has influenced our own internal tool sets in a very literal way. So two ideas for tools that we use at Moving Traffic Media, Semantic Triple Generator and an AI Anchor Text Suggester, both came directly out of his thought leadership. So if his ideas are already showing up in our work, we figured it was long overdue to have him join us on the show. Everett Sizemore, welcome to the pod.
Everett Sizemore (02:07)
Thanks, thank you for having me.
Joe (02:09)
We want to jump into the hard stuff. But first, I think you will receive the award for our most well-traveled guest. So I just want to get you to comment on 'it looks like you've lived in multiple states, multiple countries on different continents. And you do a lot of writing, which we appreciate. Your career started in journalism. So I wonder if you could just comment on how all that travel you've done has influenced your interest in writing.
Everett Sizemore (02:38)
I'd love to talk about that. Just a quick correction, I probably am not the most well-traveled. I was listening to your episode with Lily Ray, the last episode you had, and she's like, my favorite place to go dancing is Berlin, or you know, just to go out for the night. But I started off, I guess when I was about 18, I moved to Hawaii, and when I was there, I wrote a couple of articles about unseen Hawaii or something like that for a travel magazine. I was like, I really like doing this. So I did a bit just visiting places. I moved to Indonesia for a while and wrote for a surf magazine there called Surfer and a few other small magazines, the kind you get for free. And that went pretty well. I had to come back to the US after the nightclubs and bars I went to started blowing up after 911. And I was like, okay, well, what am I going to do now? And a lot happened. I started an e-commerce business. I was the jewelry supplier, the organic piercing jewelry supplier for Hot Topic for about six years.
Jon Clark (03:41)
Oh wow.
Everett Sizemore (03:41)
That's how I paid my way through college. So I moved to Australia to go to school and I went to Bond University there and majored in journalism thinking that I would come back here and do that. When I got back to the States, even well-seasoned journalists at newspapers were getting laid off. It just was not a good time for that. But fortunately, Associated Content had started up in Denver. And they were looking for somebody to sort of help them scale those efforts. And it was kind of the first big citizen journalism play. I guess you could call it like an early Substack or something like that. And that's when I really started. So I had the e-commerce business and I learned, yeah, you have to do certain things to get found online, but I didn't know what SEO was back then. I'd never even heard of it. This was like 2005, 2006 or something when I started working at Associated Content. And through that, that's when I really started diving into SEO, just to help them scale that out. And I loved it, so I kept doing it. Plus it paid better than writing.
Jon Clark (04:44)
Well, when Yahoo acquired it, it eventually became Yahoo Voices. Is that right? Yahoo Voices? What did you learn about user-generated content way back then, when the rules were, I would say, much more loose in comparison to sites we've seen hit more recently? What did you learn back then that is transferable to today's environment where we're seeing content scale, but through maybe LLM models and not necessarily human-created?
Everett Sizemore (05:22)
That's a good question. So we tried to avoid any kind of programmatic content SEO back then. There were just so many writers who were willing to do it 'it was based on a rev-share model. You got a rev-share of advertising on your content. And so I would do basic keyword research. We would figure out what people were searching for and what the big topics were, what was trending. And we would put out a call to those writers. And because they got a rev-share, some writers were really motivated to write about topics that were popular or trending. The content that came out, I think, was 'even if it was written by citizen journalists, people who just don't have a journalism background 'the quality I think was still better than what I see coming out today from LLMs. Maybe not as grammatically correct, maybe a little more repetitive or long-winded, but it just sounded more real and it was very obvious from the writing and the way we presented the authors. If you want to talk about E-E-A-T, they had author profiles, links, bylines from every article that would go to the profile page. They could write stuff there about themselves and then it would have a list of all the articles they'd published. You could just tell that these were real human beings writing first-person perspectives. And I think that really made all the difference. The other thing that was really helpful at the time, which you can't do now, is you could have five or six indented results from the same domain. So we would put out a call for articles and get 10 or 15 or 20 written about this certain topic. And a handful of those could all be on the first page. And you would run into some pretty serious keyword cannibalization if you tried to do that now.
Jon Clark (07:07)
So Rand Fishkin highlighted a slide that you had shared back in his MozCon presentation back in 2016 around quality being sort of a site-wide metric. Take us through how you arrived at that 10 years ago when today it's a pretty common known.
Everett Sizemore (07:23)
I have to give credit to two of my colleagues that I worked with at Inflow at the time, Dan Kern and Rick Ramos. They were working on a project for a client and had cut out like 80% of the content. Just really bad old blog content and things like that. And I'm not sure if the goal was to improve rankings so much as, like, this is just a bunch of stuff we don't need. And when they looked at the results and saw, okay, there's a clear correlation between these improved rankings for everything else and when we remove that content, then we started testing on other clients and started replicating it and sort of refined the strategy. And that's when I came up with, okay, quality is a site-wide metric. I forget how I put it, but essentially that is the case. We know from more recent documents that were released in Google court cases that site authority is a distinct part of their algorithm. And at the time there was a lot of back-and-forth arguments in the industry about whether that was just kind of a side effect and we were confusing correlation and causation or what, but yeah, we happened to be right in that case. I definitely can't take all the credit for that.
Jon Clark (08:32)
I assume you're still using that tactic today. Can you maybe take the audience through what a pruning decision looks like in terms of benchmarks that you look for or certain types of content? For example, tag pages on a WordPress site. Are there common things that you can quickly identify as areas to prune or is it really data-driven?
Everett Sizemore (08:52)
I'd say it's both. You do enough of them and you realize, okay, I'm on WordPress, it's these pages. I'm on Webflow, it's these pages. There are always outliers, but you get a general sense depending on the CMS or how the site was built. But generally speaking, I would make sure that I have an inventory of all pages first, so the content inventory is the first step. I would expand not just URLs that I find during a crawl. I would also include any URL that's in the sitemap and any URL that 'and of course, Screaming Frog makes this a lot easier now connecting to those APIs 'but anything that is in Search Console or Google Analytics for the past, I usually use a year. I think it defaults to 30 days or 90. So pull in all of those URLs, obviously deduplicate them, get rid of unnecessary query strings and things like that, pagination, stuff like that. And you end up with this inventory of all the content on the site. And because, especially again with Screaming Frog, you have all of the metrics, I connect it to Ahrefs as well. So I know how many clicks and or impressions or visits each page got. I know conversions and I know whether it has backlinks. So generally, it's just a list of metrics all in one place in one spreadsheet where I can see how this content is doing compared to the rest of the content on the site. And the threshold that I would draw really depends on the scale of the site. So on one site, you might say, I'm just going to get rid of any content that has had zero visits in the last year and doesn't have backlinks. I'll just allow that to 404 and I'll update my internal links to get rid of links to those pages. For another site that is, let's say they have a million pages, you might draw that line at five visits or something. It kind of depends on how it all adds up. Generally, if I'm going to do the audit, I'll limit the base recommendation down to: improve it, leave it as is, or remove it. And if it's remove, then there's another column that I'll put in the sheet or wherever you're doing this that says, okay, I'm removing it. Does that mean I'm consolidating it into another piece of content? Does it mean I'm no-indexing it, but it's going to stay around? It doesn't mean I'm deleting it and just letting it 404. So you make those determinations depending on things like, okay, it's helpful for users. A good example of this would be I see a lot of SaaS companies publishing their user help database. And most of those questions aren't relevant to people who aren't already signed up for your product. And yet, they're getting indexed. They're competing with some of the same keywords that your transactional or informational content on the main site ranks for. So a lot of that stuff, we don't want to delete it, but we want to no-index it. So it would say remove with a no-index tag. If it's improve, okay, well, what exactly are we improving? Is it the text? Is it the layout? Or maybe there's some other piece of content that is set to remove that is going to get consolidated into this piece. And so it would be improve it by consolidating this other stuff into it. Alex Juel, who I also worked with at Inflow, had at one point created sort of an automation that had a bunch of rules around it, like if this then that. And that was pretty cool. It got you like 80% of the way there, but there are just so many outliers that I think only experience can really teach you about.
Jon Clark (12:21)
I was just going to ask about that because we take a similar approach with our content audits where we pull in a bunch of data and then we've tried different weighted models where we weight links differently versus traffic or we take an average of all visits that are landing on the URLs collectively and then make decisions on URLs that fall below that or above that. And then we take the human element to sort of validate or not some of that information. But I'm always curious about folks who go through these exercises and how do you look for things that could be rewritten? Because it could be a very top-performing post, but there could be a whole bunch of just old data or dates in there. And so sometimes it really does require a manual review of the content in some cases. Have you figured out any ways to automate that side of thing, or is it really just the manual one-by-one sort of final validation of what you're recommending there?
Everett Sizemore (13:18)
Well, I'm constantly sorting and filtering. I tend to work in spreadsheets a lot. And so it's not necessarily that I'm going through one by one. For example, I might look for any title tags that have a date in them, or I might look for words in the content like yesterday or last week or this year. If it's from 2014 and it says yesterday, you know, does that content need to be pruned out because it's no longer relevant? Or can it be made relevant by adjusting the language to be evergreen? So, in August of 2024, as opposed to yesterday. So I look for dates. I look for certain words that indicate that the content is time-bound. And then I just look at the metrics as well. If it 'most of this stuff is going to fall in the realm of it gets some traffic, it could do better. I also pull in rankings for the top keywords and things like that. So I would know like, okay, well, show me all the pages where the top keyword is on page 2 or below position 5 and that has traffic between 10 and 100 visits. It really depends on the client and the site. You would look at them, or at least a subset of them, one by one and determine, okay, well, this entire paragraph at the top is just fluff. It can be cut out. This could be a bulleted list or a numbered list. You start seeing things like that. And then that's where I would leave the comment: Okay, well, this needs to be improved. It has a little bit of traffic. It has potential. And these are the things I recommend changing. I could probably do some, make some Claude skills or something that helps this. I haven't yet, but essentially I just have to look at it.
Joe (15:13)
You have this message that comes through in your writing of cut the fluff. It comes through in a lot of what you do. But you've been a vocal proponent of getting back to basics, straightforward writing styles. A lot of the writing you've done on semantic triples, I think calls that out really clearly. I wonder if we could maybe shift the conversation more toward the importance of understanding entities because I think the relationship patterns that you can build quicker around an entity come from more straightforward, simplified writing. So if you're okay with that, maybe you could, for our audience, let them know how to think about an entity and then how to structure content around an entity they may find important for their company or brand.
Everett Sizemore (15:59)
Yeah, that's a big question. That's a good one. I think probably the first thing is just what are the entities and defining them, and then working with leadership to come up with kind of a canonical description of that entity. Like, let's say you have the brand as the entity. Well, how do we describe this brand? And what often happens is it'll be described on the homepage this way, and it'll be described on the about page this way, and the social profile this way. And sometimes not only are they written differently or focused on different things, but they can actually be contradictory. So there's the concept from local SEO, NAP: name, address, phone number. So one of the things that those folks do, or at least I haven't been following that sector, but I'm assuming they still do it, is they try to standardize everything. So if you're going to have your phone number as in parentheses versus a dash or vice versa, try to standardize that. If you have corporate offices and you have some brick-and-mortars, you need to be pretty specific about which address goes to which. If you are calling it Acme Inc. in one place and Acme Co. in another and then acme.com, LLC 'so you just need to sort of consolidate all of that towards a single entity with a canonical description. It doesn't mean you have to use the exact same description every time, but they have to be fairly consistent. And you can do that in the organization schema, homepage, about page. You can do that on social profiles, third-party sites like G2 Crowd. So there are lots of places where you have a bit of control over the description of your brand or whatever entity you're working with. That's probably where I would start.
Jon Clark (17:45)
We had Darren Shaw on the show last week and we use a lot of his citation cleanup services. And when that's done, you're exactly right. You basically have a sort of: here's an official way to talk about your address and reference your business for local companies. And that becomes like their Bible. So it seems like you've been preaching this concept of a semantic triple way before we were even using that phrase as well. And you had a great article on LinkedIn around semantic triples, what they are, how to apply them. I think Jason Barnard has done some work there as well. How are you working with clients around this concept? It seems very basic, but when you use terms like semantic triples and things like that, sometimes it can be difficult for clients to'I don't know'be willing to standardize the way that they talk about themselves across everything. Is that really what you have you handled'some of the challenges of talking through that concept with folks?
Everett Sizemore (18:41)
I look forward to hearing that episode with Darren Shaw. I'm glad you brought up Jason Barnard. I actually got in touch with Kalicube, I think last year or the year before, because I had two entities that were being confused as a single entity and I needed to disambiguate them. The budget didn't work for the client I was with at the time, but they did give me some advice and it worked. I forget exactly what it was. But you have some places that are really powerful 'and we're kind of on a different subject than semantic triples, but back to the entities for just a moment 'if Wikipedia says that you're a blue balloon, it doesn't matter what you say on your site, you're a blue balloon, right? So there are some sources of information that are more powerful than others. So you don't necessarily need to standardize it everywhere, but where you have control over it. Back to semantic triples 'I don't know that I've been preaching them for a long time. I actually remember other people talking about them many years ago when Google started bringing in algorithm pieces that used natural language processing. And it's just a very easy way for knowledge graphs to understand and store information. So you have subject, predicate, object. Jon Clark records a podcast, right? And it's never the other way around. It's never like: this is the podcast that Jon Clark recorded. Humans can understand that, and certainly machines can understand that just by talking to AI. It's not difficult for them to understand that, but it's just a little bit easier for that knowledge to be stored in some knowledge base. I start with clients by saying, okay, well, let's define what your brand is when it comes to that, not just: the brand does this. And it sounds remedial 'sounds, of course we're going to say what our brand does. That's one of the first things we do. But then you go to most SaaS websites, and you read what they are, especially if they're in tech. And unless you know those industry-specific terms 'a lot of times I'll come away just going, I have no idea what they do. I've just read a whole page of content and I don't know what they do. And maybe it's just because I'm ignorant to their industry, but ideally anybody would be able to come and understand what you do. I had an article that I published on LinkedIn with a prompt that would ask an LLM to give you a certain amount of semantic triples about your brand or products and services. And it would give you some for your services, audience, value propositions, and differentiators. So it's not just: we do this. It's: brand X helps these people do this, or product Y is for these people, right? So it doesn't have to be just this short little sentence every time using the same thing. You're describing all kinds of different things. In the example I used, it was a differentiator between LinkedIn and other social networks saying LinkedIn caters to business professionals or something like that. So that in itself is also a semantic triple.
Joe (21:48)
So the entity could be the company, the company's name, but can the entity also be the service or product category that they want to be known for?
Everett Sizemore (22:00)
Absolutely, it could be any product or service. So your brand is an entity, that product or service is an entity. An entity could be the industry that you work in. It could be a person, place, thing, idea, any of those things. It's important sometimes to clarify or disambiguate how those entities connect to each other. I was working with a client recently that has two sides of the business. One is B2C and one is B2B. Both are named the same thing. They don't differentiate. So you have people on Reddit saying, I'm confused, I thought the brand did this. So those are the kinds of things that you would need to clarify to say, okay, well this part of our business does this. And the hard thing is, well, in order for you to separate out that part of the business, it kind of has to have a name, because that name and the context around that name is going to become some kind of mathematical 'some kind of number, whether it's in a knowledge graph or in some kind of multi-dimensional vector space. So you need to separate those two things. So we just added a little bit after 'instead of just brand X, it was brand X for business or something like that. Speaking of that, one of the best places to start kind of work is Reddit. People have questions about brands on Reddit. So you gather those questions, you categorize them, and discover the things that people are confused about. I have a client who, since they started up like seven or eight years ago, has only served US clients. And they recently acquired a business that's overseas and have opened up to clients throughout all of Europe as well. That was 'three months had gone by. And if you were to ask ChatGPT, "Hey, what does this business do?" it would tell you they work with US customers doing this and this. So it hadn't picked up on that yet. So I used a tool called Waikay, which I think others may have mentioned on the show before, but it's sort of a hybrid prompt tracking tool. But the thing that I find interesting is it will tell you what facts are known about your brand or your entity or whatever it is you're looking into. And it will tell you where those came from. So I was able to track down multiple sources that had text on the page that said, this company does this for US people only or exclusively. I'm looking for words like that. And I approached some of them. We have direct access to social profiles and things like that. Others, it would be somebody who had written about it and I could contact them and say, hey, would you mind updating this information? Some of them would be business listings and things like that as well. And so we corrected that on the website. There were still plenty of places on the website that still implied that it was US-only. And then we had that corrected on a multitude of sources out there that were being cited by LLMs. And within like a week, you could go to ChatGPT and ask, and it would tell you all over the world. So that's the sort of thing where semantic triples and entities and AI and SEO kind of all come together.
Jon Clark (25:16)
Is there a PR relationship in that ecosystem as well? You could imagine press through some sort of PR arm could also influence what those citations are as it relates to where that business was operating globally. How do you see PR fitting into the current age of SEO that we're in today?
Everett Sizemore (25:38)
I always tell 'so I hate link building. I just think it's a different mindset and a different practice. I hate that we're the ones 'yeah, it's almost like cold calling, right? Reach out to a hundred for one. So I always tell clients like I would rather have one halfway decent PR person than a dozen SEO link builders any day of the week. It's just much more powerful. One of the things that happened with that other client I was talking about that had acquired the other business and entered new markets in Europe 'yeah, they issued a press release saying, hey, we acquired this business. But in that press release, it didn't say, now we serve these people in Europe. So that kind of wording wasn't in there. Definitely, it doesn't even have to be a big PR hit. There are 'we could talk about PR all day long. I'm not in PR, but I love it. And I really respect those professionals who know how to do it well. But it could be a data journalism piece that you pitch to a dozen different newspapers with local spins on each, or it could just be something you put out on PR Newswire. It's a good shortcut to say this is the language we want LLMs to use. It's just a good way of getting some of that language out there. Guest posting, same thing.
Jon Clark (26:56)
I guess it's fair to say SEOs might still misunderstand or maybe not take advantage of the PR channel as much as maybe we should in today's world.
Everett Sizemore (27:07)
Yeah, I would love to partner with a PR person. If I could offer that to my clients without taking half of the invoice cost and giving it to someone else, that would be great. Every time I've worked with a client that is working with a good PR professional, it's just been so easy. I don't have to worry about backlinks. I love it.
Joe (27:26)
We kind of have this thought experiment with a lot of our guests: if money wasn't an issue, how would you build this perfect team for this new type of optimization? It always starts with an SEO fundamentalist. But then it's like, well, I'd like a piece of a UX person, I'd like a piece of a PR person. It's an ecosystem now. It's less of a channel. It's really broadening the way we have to think about digital marketing.
Everett Sizemore (27:53)
Absolutely. A lot of walls have come down in the last couple of years.
Jon Clark (27:56)
I wanted to talk a little bit about HTML sitemaps. We've been a huge proponent of HTML sitemaps basically forever. And you've been pretty vocal that they're sort of the comeback story 'they're maybe not given the credit that they deserve. Can you take us through how you're thinking about utilizing HTML sitemaps to influence either AI citations or maybe just generally content discoverability?
Everett Sizemore (28:19)
Yeah, it's not a huge part of the strategy, but it is part of it. And it seems like a lot of websites, a lot of SEOs, just don't create them anymore. You have expanded footers and you have XML sitemaps and people started using LLMS.TXT files, stuff like that. But none of them do exactly what an HTML sitemap will do. The first important thing about them is they can be very simple: minimal design, no lazy loading, no images, no JavaScript. It's just text paragraphs, straight HTML. And that's very easy for large language models and the systems that feed them that information 'it's easy for them to process. It also gives you an opportunity to describe what can be found on the site. I did a little experiment with them 'it didn't work, but the idea is you have your navigation, which doesn't always link to the most important pages. You look at any website's internal linking structure and things like terms of service and privacy policy have more links than your most important piece of content. It allows you to separate out 'that stuff is needed for navigation, I'm just going to focus on the important pages. So I don't link to every page in the sitemap. It's also different from the XML sitemap, which includes every page. And so now, again, just focusing on the important ones, but also providing context. So let's say you have a section of the site. It's not just going to list out every single page there, it'll list out the important pages, but above that, it will describe what this section is. These are our frequently asked questions about this and that. These are our most popular blog posts. So you can provide some context there as well. It's just another small piece to the puzzle, and I think it's important because most websites these days, it seems like, are just loaded with JavaScript. There are very few pages like that that can be just simple HTML outside of serving markdown files.
Joe (30:41)
Optimizing content for bottom-of-the-funnel queries seems to be the best opportunity today for us as an industry. I don't want to talk about listicles anymore on this podcast, but are there types of content you're suggesting to clients or you're interested in exploring to help with bottom-of-the-funnel visibility with LLMs?
Everett Sizemore (31:03)
Yeah, specifically talking about on-site content. Let's not talk about listicles. Some things I 'and I work with a lot of SaaS companies, B2C and B2B, and they kind of follow a framework, but most sites are leaving out a few important things. So you have home, about, contact, those kinds of pages, then you have your products and or services. So each one will have its own page. And then you have potentially features. So these could be features that span different products or it's just a feature that comes with the product that's important enough to have its own page. You have solutions 'you could also call them use cases. You have case studies, comparisons, reviews, and then what I would call persona-based needs pages. And often in navigation, those will be represented either as "who we help" or "who we help" and then there's a dropdown that says "by industry" or "by job title" or "by role." So those are like persona-specific landing pages. I think those do really well now. They are the types of pages that I tended to ignore in the past, and I'm kind of kicking myself because the old-school way is you do your keyword research and you create the content that fits the keywords. A lot of those types of pages don't have much search volume, but it's important when somebody goes to an LLM and they say, "I am a marketing director" or an HR professional or whatever, "I'm looking for something that does this, I need it to integrate with this and this." Integrations is another good example. It's good for the LLM to be able to have that page to reference and cite and say, you know, actually, this company makes a SaaS product that's built for you. So no, it doesn't get a lot of search volume. You're not going to have a bunch of organic traffic coming to those pages, but it's become more important than ever to have them, I think. Pricing is another good one, and quotes and calculators, things like that.
Jon Clark (33:10)
I was about to mention pricing. I think it was Kevin Indig that did a pretty broad study that showed basically if you don't include the price in your pricing pages, a lot of LLMs will just hallucinate something and put something out there. So there was a benefit to being willing to put that out there and say, this is what our price is. You've talked a little bit about how FAQs can be used around brands, like your own brand, and how those can be used to influence LLMs and sort of what their knowledge is of that brand. I'm curious about the other pages that you're talking about. Do those also warrant FAQs on those specific pages as well? You know, you've seen some sites where they get FAQ-happy and every page has an FAQ. Is that something that is reasonable in today's day and age where we need to be able to funnel up that important content in some way? Or does that branded FAQ better live on its own page or the About Us page or something like that?
Everett Sizemore (34:08)
Different FAQs for different pages. And also the FAQs shouldn't just basically reiterate what's above. You start an article with "What is a blue widget?" and then you get down to the FAQs and there's a question, "What is a blue widget?" And it's especially dangerous if you describe it differently there than you described it up here, because which is which? I think the mistakes people make with FAQs where they say 'it makes sense to have them on every single page. It can, but they'll put informational FAQs on transactional pages, for instance. Like let's say I have a page: "I do blue widget repairs, my blue widget repair package costs $59," and I give all that information on that service page. And you get to the bottom and there are FAQs like, "What's a blue widget?" You wouldn't be here if you didn't know that. Or, "Should I get a blue widget or a red widget?" Those kinds of questions don't belong on a transactional page. They should be questions about: "Does your blue widget service include a warranty? Do you have a service fee?" Things that are tied directly to whatever that product or service is.
Joe (35:11)
How long does your widget repair take? I want to stay in this topic, but I really want to get you to comment on chunking because you have been really vocal about the back-to-basics writing style. And in one article I saw you wrote: "Can't we just get back to high school English class? It's really that simple." The chunking 'it feels like you're saying chunking is the same as writing a paragraph with one main idea. Is it really that easy?
Everett Sizemore (35:40)
Sort of, yeah. I think the way people explain chunking and that concept, that's often what they're trying to explain. It can mean other things. First, I just want to comment 'it sounds like I'm really vocal about a lot of stuff. I should probably start watching my tone on LinkedIn. I think I've become this cantankerous old fart after all these years.
Jon Clark (36:04)
I'll be honest, it doesn't come across that way to me. It just seems like: "Hey, I have an opinion. Here's what I think. And here's what I've seen work." I think that's the most important thing 'there's proof points typically to what you're saying. And so it's not 'we've all seen the SEO bros on LinkedIn that are just spouting off nonsense and there's no real data to back it up. I think in many cases, you've earned the right to share whatever you're sharing.
Joe (36:25)
The other thing is you look at the comments of people who react to the material you're putting out, they're all the leading names in our industry. So I would say you talk and people are listening. We're excited to have you today.
Everett Sizemore (36:39)
Wow, I appreciate that, especially given the people I've seen come onto your show. Will and Tom and Lily. Did you have Gianluca Fiorelli on yet?
Jon Clark (36:49)
Not yet. That's on our wishlist.
Everett Sizemore (36:52)
Yeah, he'd be great. So anyway, enough about me being vocal about chunking. Is it as simple as just writing short paragraphs with one idea? For the most part, yeah. But it also kind of has to do with 'you have to forgive me because I am not an AI expert 'but from my understanding, when a page is processed, the content is tokenized. And so this paragraph could be a token. The paragraph could even potentially be split in two and there are two different pieces of information or data. So it kind of is important in that way. So you would want to separate those two. Like in the case of an FAQ, one thing I tell clients is: make sure that the answer includes the context of the question or the question is implicit in the answer. And that's because that content could just get tokenized differently. You could have the question go here and the answer go here. It's just important for each of those pieces to stand on their own because of that. So if I were to say: "Is it possible to have a blue widget and a red widget on the same thingamabob?" I wouldn't want to answer that as "yes" or "no" or "it depends." I would want to say, "Yes, it's possible to have a red widget and a blue widget on a thingamabob at the same time." That's a very short, succinct, unambiguous answer. And then after that, in a new paragraph, you could get into the details. Like, it's possible in this specific circumstance, but nine times out of ten you don't want to do that 'your thingamabob will fall apart.
Jon Clark (38:34)
I thought Steve Toth always had a great recommendation for that sort of structure, especially around comparison content, like widget A versus widget B. And his recommendation was to start it with the main difference. "The main difference is X versus X." So you sort of reframe the question in an explicit statement, and that way you get that context really succinctly in a single sentence oftentimes. He was having some success getting that pulled into AI overviews and quick answers and things like that.
Everett Sizemore (38:58)
Love that. Yeah, when you're comparing multiple things, you start off with the differentiators or by disambiguating those two things. That's great. I like that. I think just sticking to one major point per paragraph is like high school English or maybe middle school English. And so the amount of people who are saying, "This is chunking, this is this new strategy or tactic" 'I've been recommending to my clients that they write like that on most pages, not every page, but on most pages, they write like that for many years just because it's good writing and it's scannable. When people are on their phone or on a desktop, it's different than when you're reading a book or a magazine or something where you have these long paragraphs and that's fine. When you're on the web, there's something about our attention spans that just go out the window and people start scanning 'they start scrolling down the page. And I think breaking it up like that makes the page more scannable. So there's a UX aspect to it as well.
Jon Clark (40:04)
Are you scoring the content on a readability level? I think Screaming Frog has the ability to do that with some of their JavaScript options. There are other tools that do this as well, but is that something you're looking at as well? Like, "I would prefer all the content be at an eighth-grade reading level" sort of thing. Or is that something that you're not really looking at?
Everett Sizemore (40:23)
I haven't really used that feature. I've been aware of it, and every time I run the "export all internal HTML" or whatever reports you can, you can see that metric. And I've always found it interesting. But you have some clients in a B2B SaaS industry that has a 'maybe they have a lot of terminology that most people don't know about, and they're writing to their audience. So maybe I should start looking at it in the sense of consistency. "Okay, well, you're talking to this audience and you sell things to rocket scientists. You're not going to have an easy readability score." But it might be kind of weird if you just have these three or four pieces of content that are sixth-grade reading and you go there and you look at those and lo and behold, this is a guest post that we allowed somebody to put up in 2014. That's interesting. I haven't used that, but maybe I should start looking at it.
Joe (41:22)
We've only got a few more minutes left with you. And I want to talk a little bit about your consulting practice. You've mentioned working with SaaS clients. We know you're pretty particular about who you'll work with. You don't like to work with 'well, I won't put words in your mouth, but maybe you could tell us about who you like to work with now. And I'm really curious about when you make a decision to move on from one client. If you're not going to just keep building more and more clients 'you want to keep a core set of clients that you know you can spend enough time with 'how do you decide to move on from an existing client?
Everett Sizemore (41:56)
I don't often do that. And maybe it's just because I tend to work with clients that I like. I'll often send a lead somewhere else if it seems like it might not be a good fit or something. I'm trying to think if in the last six years of being a consultant I've divorced a client, and I can't really think of one. There have been a few where it's gotten to the point where I don't enjoy working with them anymore because of a change in leadership 'different things happen. But I still try to do the best I can for them. But what ends up happening in those kinds of situations is the client senses that as well. And I don't do contract lengths. A client's not stuck with me for six months or a year. It's always thirty days' notice at any time. So I think if the relationship is not working 'and it is a relationship 'either party should be able to leave at any time they want. I don't want a client continuing to work with me just because of a contract length. The other advantage of that just from a business perspective is I don't 'I think it's crazy. I think that building in a contract end date is building in a day when your client has to ask themselves if they want to keep you around. And I don't want that. I don't want to say on April 15th, "I'm going to have to justify my existence to you," and then I'll have to do it again in another three or six months. It just rolls on. And so it's month-to-month from day one. And I find that works really well for me and my clients appreciate it as well.
Jon Clark (43:26)
I think the interesting thing about that is that even if your relationship with your day-to-day client is really strong, if that contract has to go back through procurement and legal, there's all these other eyeballs that are like, "Wait a minute, this line item might be easier to cut than to renew."
Everett Sizemore (43:42)
Yeah, and I get it in an agency model where you have to predict cash flow and things like that. But if you're just a consultant, I don't see what advantage there is to having a contract length. There are only disadvantages.
Jon Clark (43:57)
All right. A couple of really quick rapid fires if we have you for another five minutes or so. I found this amazing video, "Cruft Finder." And I was curious 'I can't remember the exact number of years, I think it was eight years ago maybe 'but it would seem like that tool still has a ton of value, or the concept of it. And the Google Doc is actually still downloadable. And it would seem like you could take that and move it into some sort of online tool with the use of AI to create or something. Have you thought about improving that tool, putting AI on it?
Everett Sizemore (44:42)
I would say that that tool belongs to Inflow because it was created when I was there. But now I'm going to go check it out now that I know the link still works and the spreadsheet and everything. Cruft is essentially referring to all those pages that in a content audit you probably want to remove and prune out.
Joe (45:02)
Is there an author or a small group of writers who influenced the way you write?
Everett Sizemore (45:07)
I don't know about the way I write. There are people that 'in terms of authors, I don't read a lot of books in our space. I've 'from time to time I'll read one. If I had to choose a book that influenced me most, it would probably be Flow, and I can't pronounce the guy's last name 'there's a lot of consonants in it 'but his first name is Mihaly. And it's called Flow: The Psychology of Optimal Experience or something like that. And the reason that's been so important to me is it keeps me small. It's really tempting sometimes to 'if I have a full client load and then say I'm very lucky and blessed and I have another lead come in and then another one 'it would be tempting to say, "Okay, well, I'm going to hire somebody or I'm going to start outsourcing a bunch of stuff and really grow." Instead of having five clients, I'm going to have twenty. If I'm making this much money, I can make this much money. It reminds me that my flow state is doing the work. My flow state is in a spreadsheet or looking at code or using Chrome Inspect and solving puzzles 'things like that 'or analyzing data, looking for patterns. And if I started getting into a world where I have a little mini-agency or something, then I'm going to worry about putting food on other people's tables. And if I don't do this, then I'll have to let people go and this and that. And so I would end up doing all this stuff that I don't really enjoy. And so every once in a while, I read that book to remind myself: I want to spend as much time as possible in that state.
Jon Clark (46:49)
I love that. That feels like the perfect place to wrap up if that works for everybody else. That's a good way to end things. I have to check that book out.
Everett Sizemore (47:00)
It's great. It's been around forever.
Jon Clark (47:02)
It's a sign of a good book. Well, Everett, this has been great. I appreciate you taking the time to dig into some of these questions with us. If folks want to connect with you online, where can they find you?
Everett Sizemore (47:03)
My website 'you've heard the phrase "the cobbler's kids have no shoes" 'it's basically an online business card, but it's esizemore.com or my LinkedIn profile. I think it's linkedin.com/in/everett.
Jon Clark (47:25)
Okay, perfect. We'll add those to the show notes as well. Thanks again for joining us on the Page 2 Podcast, and for the folks listening, if you enjoyed the show, please remember to subscribe, rate, and review. We'll see you next week. Bye-bye.