Why Topic Clusters are Better Than Keywords

Hi, guys. How's everyone doing? Welcome to Inbound 2017. Is it anyone's first inbound? Anyone's second inbound? Third inbound? Well, awesome. Welcome and welcome back for all of you who are returning. My name is Leslie. I am a marketing manager at HubSpot. As Aideen said, part of the marketing team since March 2015. So before we get started, I have a little bit of a riddle for you. Started out, if you've heard this one before, where is the best place to hide a dead body? Anyone know exactly page two of Google? Because no one's ever going to get there. So that's all. Thank you all for coming. My slides will be online. No, I'm just kidding. This is. So let's dive right in. In this talk, we are going to be talking about, obviously, topic clusters, what they are, why they matter, why they're important for your content strategy, and most importantly, how you can use them to empower yourself and your team to make the most of effective content possible. So we're gonna take a little bit of a trip down memory lane back to what HubSpot was doing in 2015. Back when we were an early blog, the strategy of our SEO strategy wasn't super set. We're gonna talk about what we were considering then, what we're considering now, and then I'm gonna teach you guys exactly how you can take this strategy back home to your teams and start implementing it today. So let's go to March 2015. I had just started at HubSpot as a staff writer on the sales blog. I was our second full time hire. The lovely lady to my right is Emma Rudner. She was our editor. You probably have recognized her name. And this is the HubSpot blog. We were in really, really, really early innings. We started the sales blog back in July 2014. That was our first full time hire. And at that point we were very new in the sales market. We were pretty established in marketing, but not so much in sales. And so we were doing a lot of things all at once. We were kind of throwing topics at the wall to see what stuck. We didn't really have a good strategy. We didn't really know what we were doing. We were just testing. We were really, really in those early phases. So were all of you guys at Brian and Dharmesh's keynote yesterday? Awesome. So if you were there, you remember that Dharmesh talked about how the HubSpot blogs used to publish a lot of content. I was definitely part of that problem, hopefully was part of the solution. But this is exactly how much content we were doing on a weekly basis. This is a screenshot of our real editorial calendar from 2015. We were publishing five days a week, four posts a day, 6 37, 38, 30 12. I can still see this calendar in my mind when I go to sleep. I still have nightmares about not filling the slots for the next day. We were just a two person team, super, super scrappy. And when you have to publish 20 new posts a day, or sorry, 20 new posts a week, you have to make some compromises. So we were publishing old posts and new posts, we were updating old posts and re optimizing them for organic. So here's one example of a post that we republished and tried to get to move up the ranks, but we were also publishing a lot of net new posts. And so on this slide I've circled a few of the posts that, let's just say if there was a Pulitzer Prize for content marketing, these would not have been the posts that we submitted for consideration. So afternoon energy boosting foods, that's relevant to the sales audience. Maybe it's something that's helpful, but it's not something that's going to be long term valuable for us. Same with an infographic about coffee. Something that's fun, a little bit shareable, but again, it's not going to be doing work for us in the long term. We also had posts though where we were being deliberate about how we were optimizing the post, how we were structuring it and what we were publishing. So these posts that are circled are posts that we thought had potential value to rank long term, that we thought that they would rank for long tail keywords or a broadhead keyword and would just rise up the serps over time. So we were doing some keyword research, we were optimizing each post as best we could. People like to say. And at HubSpot we like to say that blogging is really a set it and forget it channel. You write a post, you publish it, you optimize it for organic. It rises through the SERPs, it gets shared, it gets backlinks, and over time it accrues traffic, it accrues leads, and you really generate leads in your sleep. Right? That's the dream of blogging. And definitely we were a set it and forget it channel. But in 2015, what that meant was we would set a topic, we'd write the post, we'd publish it, and then we'd forget about tracking it because we had too much to do. We were a two person show publishing 20 posts a week. Thankfully, we brought in a full time writer in July 2016. It lightened the load a little bit, but we were still working seven days a week. Emma got promoted and so it was two people again. And you know what, it's a little bit of a sob story on my part, but my job was actually easier than all of you because I had the huge luxury of being part of a huge marketing team. All I was doing was creating content. So I just want to get to know a little bit about who's in the room. Can you go raise your hands? If you're a marketer, I assume it's most of you. Cool. If content creation plays any role in your job, put your hand back up. Can you keep your hand up? If you do content creation and at least one other job, Maybe it's social media, email nurturing. Exactly. What about two other jobs? Three other jobs? All right, now everyone put your hands down. Who is lucky enough to just be a blogger and editor? Exactly. No one. And that's the power of this strategy. The thing that's really, really, really hard is when you're at a company that doesn't have the luxury of having very specialized roles, people in very specific categories of marketing, everyone has to wear a lot of different hats at once. And that means that you really, really, really have to be effective with the time that you spend creating new content, creating new campaigns, creating new posts for social. Studies have shown that we're only really effective for 40 to 50 hours a week. That means you don't have the luxury of throwing stuff at the wall to see what sticks. You don't have the room to say, oh, I'm going to create 20 pieces of content. Maybe four or five of them will rank in the long term, maybe they'll do work, but maybe these other 15 won't do anything. That's not a luxury that most people have. The topic cluster strategy is a way for you to take control back over your day. Hopefully you'll get to enjoy your nights and weekends again. You'll get to create content with purpose. When you implement this strategy on your blog and everything you do will have a reason behind it. You'll understand why that new site page exists. You'll understand whether it's a lead generation play, or if it's purely a brand awareness play, or if you're just filling a gap of a topic that you haven't written about, but that you should, because that's the industry that your company is in. This is about really reclaiming your time, making sure that what you're doing is valuable, is deliberate, is high impact. And it's really, really hard when we're sitting at the top of the funnel sometimes to quantify what we're doing, to show the value of what we're doing and what we contribute to maybe a sales team down the road, how we affect revenue. And so this is really a way for you to quantify that and to prove your worth to yourself, to your boss, when you're looking for the next promotion to your site, when you're trying to quantify at the end of the year reporting on what you've done. So if you're like me, when you think of keywords, SEO, you think about keywords, you think about on page and off page optimization, you think about backlinks and anchor text and all that stuff is important. But that's just not enough anymore in 2017. So in October 2016, we kicked off what we at HubSpot call tuna season. The last couple months of every year are reserved for planning for the next year. Everyone does this. And so Emma came to me, she was my manager at the time, and she said, leslie, you're in charge of figuring out your organic strategy in 2016 or in 2017. And. And I am not an SEO expert. I am not a technical consultant. I have never spent any time in the search industry. I am someone who was a project manager for eight months as my first job out of college and HubSpot was my second job. And all I ever did was write and edit. So I had no idea what was going on. I said, ok, cool, Emma, that's great. I'm a big believer in faking it till you're making it. I told her I got this, I went away, I tried to formulate a plan, and so I assessed what was really important to us at that time. So on page and off page optimization. This is literally a slide that was taken from a HubSpot training deck back in 2015. This is what I was taught. You have to optimize your title tag, you have to optimize your page title. You put that head term at the beginning of your URL, right? You have H1 tags, you have alt text on your images. Make sure to write a good meta description so that in SERPs, people know what they're getting. So this is all really basic, this is what we were taught. And also backlinks and anchor text, all of that matters. We were taught to write one post for one keyword. So the overarching Strategy for SEO on the HubSpot blog was we would think of a term we wanted to rank for. We'd go to Google Analytics, we'd go to maybe Semrush, and we'd poke around a little bit, see what the serps looked like, see what other people were writing, and try to make our post the best in that category to try and win that category. And from then on, we would remember that head term that that post was written around. And anytime I wrote a new post that had that anchor text in it, I'd link to it no matter what it was about, whether or not it was relevant. I said, great sales qualification. I have a post about that. Even if this topic is about closing, even if it's about phone conversations, I'm gonna put it in there because that's an anchor text link that's valuable to us. That wasn't true. So before we get into the nitty gritty of this strategy, we need to really do a micro lesson in site architecture and how Google looks at your site when it's deciding what it's going to rank and what it's going to surface in search results. So who here is familiar with the term PageRank? Most people. Cool. For those of you who aren't, PageRank is Google's official term for what very technical and savvy marketers like to call SEO Juju. It means the more PageRank your page has, the more likely it is to be near the top of search results. And so when you look at a typical site, the homepage starts at the top. It has the most page rank. This is typically because on every page of your site in the header or footer, you link back to your homepage. It has the power of hundreds and thousands of links all pointing back to it. The next level of pages. These top level categories are links that exist on your homepage. These are the places that your homepage links out to. So when you move from that first level to the second level, your SEO juju gets redistributed among every single link on that page. And then all those subsequent pages link out to other pages and other pages and so on and so forth. So this is what the HubSpot blog looks like, site architecture wise, and is probably what a lot of your sites look like as well. We had HubSpot.com, we had a link to our blog in the header. We had a link to each of our specific blog sections on that blog homepage. And then we would get to the blog posts. Now we were Publishing on a crazy schedule for 10 years. This is no joke, you guys. We have an ecosystem of about 14,000 blog posts. So by the time you get down to that level, that page rank is being sliced and diced in a million ways. And it's probably true for a lot of your sites as well. A lot of you guys probably have a site where you have sort of flat hierarchies of pages that have just maybe 0.1% of the original page rank that they could have had. And so this really isn't an efficient way to organize your site. A lot of pages are fighting for attention. Google doesn't know what the most high value pages on your site are, and so there's not a coherent way for them to understand what on your site to surface. And so when you come up against a competitor that has a better organized site, or anyone who has any form of site organization and you don't, you automatically get pinged and you automatically lose out. And so all of that work goes to nothing. Luckily, we have topic clusters. So before I jump in to what a topic cluster is, I want to show you guys the power of the site architecture that exists when you are under a topic cluster model. So each of these webs represents one topic cluster. And we'll go into what each of these components means and how you optimize them. But as you can see here, the homepage will link out to 10 to 15 topic clusters. On this slide, it's 6. And so that second level, you're immediately already getting to those high value content pages that you're writing to rank for really broad competitive keywords. So because you're moving everything up one level, that page rank is getting distributed among fewer pages. What you then want to do is pick supporting content. 15 to 20 posts, 20 to 30, depending on the size of your site. Those are called cluster pages. And then the page rank is distributed from your pillar pages down to those pages. So way fewer levels. It's a much cleaner site architecture. It's much easier for Google to see, hey, this cluster of posts is relevant to this keyword. This cluster of posts is relevant to that keyword. And because you have more PageRank, it becomes easier for your entire cluster to get that organic lift. So now let's dive into what an actual topic cluster looks like. This center page, or the center circle in this diagram, is a core topic. At its heart, A content cluster is really just a set of pages. There's nothing crazy about it. There's no crazy coding or fancy modules or interactivity or video that needs to happen. It's just, just an architectural definition. And so you have this center page. It's a pillar page, the page that you've created to rank for your head keyword. And you have it surrounded by. You see these hollow circles with dots inside? Those are subtopics or cluster pages. And those cluster pages are subdivisions of that head term. So let's say you're an orthodontist and you're blogging and you have a page that's really, really great about the vast varieties of toothpaste to buy. It's like all your recommendations, where you can get them, how much they cost. It's a beautiful page. And so each of those subtopics could be variations on that long tail keyword or on that head keyword. So long tail keywords such as best toothpaste for people with sensitive teeth, best toothpaste for people who are looking for natural organic solutions, best toothpaste for people who are prone to cavities. All of those things riff off that original head keyword, but they take one specific section of that topic and dive in really, really deep. Your pillar page, which we'll go over in a little bit, should be broader, should be designed to be sort of that authoritative source on your head turn. So when you're thinking about SEO in 2017 with a topic cluster focus, there's a couple reasons why this is important. The way to think about this is that on page optimization still matters. Don't stop doing any of that stuff. Topic clusters just add an extra layer to your content strategy that really give it power. The way to think about this is back in 2007, backlinks, title tags, all of these things were new. People were just starting to enter the search industry. People were just starting to build careers out of it. And so doing on page and off page optimization got you a level above what everyone else is doing. That stuff only gets you a seat at the table today. It does not get you a level up. Everyone is doing it. The cat's out of the bag. And so this is really that next thing that's going to take you to the next level. Because most people are not architecting their site in this way. The second thing to know is that topic clusters should have a self contained internal linking structure. That is you have one pillar page, 20 supporting pages, and those pages should only link to each other and to the pillar page. There's a couple reasons for this. The first one is that backlinks are valuable. Remember, every backlink distributes a certain amount of page rank to the page that it links to. And so when you're linking within a cluster, you're sort of keeping it all in the family. You're telling Google, hey, this is where the page rank is concentrated for a reason. This cluster is valuable because all of these posts relate to that one topic. The second thing is that Google can only do so much when it crawls a site. We sort of think of Google as this omnipotent force. They read everything on the Internet, they know everything that we're writing, and they do to an extent. But Google can only crawl your site in certain detail. It can't read a blog post and understand it the way a human does. It has to look at the sort of other signals that you're giving them. Backlinks are one of those most important signals. The reason for this is because Google uses click through rate as a proxy for value of their content. When you look at a SERP and you know it's the first 10 results, you look at the most relevant one, Google's gonna say, hey, the people are clicking most on, you know, search result number three. Let's move it up. That must be the most valuable result. So let's go back to that orthodontist. He's writing about best toothpaste varieties, but he has a lot of pages on dental health. He has a lot of pages on maybe one to brush your teeth or what toothbrushes to buy. If he's linking to that page about toothpaste from all these other different pages, Google doesn't want to surface a page about dental health, about food, about toothpaste. When someone's specifically asking, sorry, about toothbrushes, when someone is specifically searching for toothpaste, they only want to serve results that are related and that are going to provide value for the user. Remember, it's always about what is most valuable for your reader. And so when you have a self contained internal linking structure, you say to Google, all of these posts are about the same thing. They all rank around one general topic. It's safe for you to surface all of these in SERPs. If your internal linking structure is a mess, Google will ping you for it because they have no idea what your site is about. They don't know what pages are about, what. This linking structure sends a very strong signal that all these pages are related. And so it gives all of your pages a lift and it makes it easier to win that keyword. That's the third thing. You rank more easily for broad competitive head terms. When you have A deliberate strategy, an internal linking strategy is the only thing that keeps your site organized. And so head terms, head keywords that have 100,000 searches a month, 200,000 searches a month. Everyone is trying to rank for those. People are buying adwords, people are buying sponsored ads, everyone is trying to get in line to be number one. But no one else is organizing their site this way, so Google doesn't know which one to surface. If you have this architecture, Google knows that you're the most valuable and that you own that entire sphere of the topic. The fourth thing is that you can measure and track performance on a cluster basis rather than on a post basis. Anyone who spent any time in content creation knows that it's very, very hard often to track the impact that one individual URL has on revenue. Sometimes you'll have those posts that are magic lead generators, but more often than not, you don't know sort of the collective impact of what you're doing. And when you don't know that impact and you can't measure trends over time, it becomes really, really difficult for you to evolve your content strategy. You can't pivot and say, hey, we've written 10 posts in this format and here was the collective impact. And we wrote another 10 posts trying out a different variable, and that was the impact there. So we know that option B is the better option. If you're tracking on a URL level, it becomes very, very difficult to see those trends. You really need a more collective impact to make decisions. And so when you have clusters, and I'll get into how you track this in a little bit, when you have full clusters ranked around head terms, it becomes much easier to collectively and understand your overall impact. The last thing is that a topic cluster strategy really provides you a strong path to conversion and acquisition. All of us are marketers. We all know that the end goal of content, whether or not it's beautifully written, beautifully designed, it's to get someone to convert. We want to get someone to book a demo to give us their information, fill out a form, whatever it is. And so when you're able to measure topic cluster. Sorry, when you're able to structure topic clusters to user intent and you're getting that lift in organic traffic, you're getting a lot more traffic to experiment with what CTA works the best. How can you reach statistical significance on this or that test? And when you have sort of all these posts that are related and users are clicking through your site in a way where they're learning the most about a topic, it Keeps them on your site longer and it gives them more value and so they're more likely to convert. So those are the five reasons that topic clusters really are extremely valuable when you're thinking about SEO. Now let's dive into pillar pages. If you recall, the pillar page is that center circle. There's five things to do when you're creating a pillar page. A pillar page is just a page. It can be a site page, it can be a landing page, it can be a blog post. It's any page that has writing on your site. But it has to be designed to rank for a head term. So number one, it has to be broad. It has to address every facet of a topic to some extent. It should be the authority. Think of it as sort of the appendix to a certain head term. People should be able to come to that page and say, okay, I feel comfortable reading this page about, say, landing pages and understanding the basic components of a landing page, what they're used for, their role in marketing, maybe a little bit about optimization. And they should be able to come away from that page feeling relatively informed on that topic. Secondly, it should be. So that's. Sorry, that's number two. Number one is that it should be broad and evergreen. The big play of a pillar page is that because it is so broad and because it is an authority for a head term, you also want to write it in a way that attracts other sites to link to it. Remember, backlinks are one of the most effective ranking signals that you can give to Google. And so if a post is about some pop culture fad, that was in 2014, and that's what you're trying to get to rank for a broad term, that's not going to work. People aren't going to be able to link to that because it's not going to remain constant over time. All of those people who are linking to your post, they're content creators too. They don't have time to go back to their post two, three years from now and change out that one link. So they're just not going to link to it in the first place. So make sure that content is evergreen for the most part. Third is that it can be relatively shallow. You have all your cluster pages that are going to support that pillar page. Those cluster pages are the place to really dive deep into the long tail keywords. The long tail questions that people have optimize for, questions through search is getting more and more conversational. People are asking Siri for things, they're asking Alexa for things. And they're talking to Google in the same way that they're conducting voice search. They're not necessarily googling that, you know, two word head term. They're saying, what is this? What is that? Why is this? How does this work? And so rather than optimizing one term and just hammering that term over and over and over again, optimize for those questions, optimize for the human variations of what people are looking for when they come to your page. The last thing you can do, and this is sort of a longer term play, is once a pillar page is really set, once you've set a topic cluster, you have the power of all those links pointing at it. You can then distribute that pillar page's page rank. It's going to gain page rank over time, it's going to gain juju. And that means you have a little bit to spare. So there's a couple things you can do here. You can link out to a new cluster from that pillar page if there's a new cluster that you want to rise and rank more quickly. If there are certain valuable cluster pages that have high volume long tail keywords, you can link out to those as well. And that's something that you can do as a longer term play to set yourself up for future success. So you set up a topic cluster, you set up a topic cluster, you wait a little bit for Google to kick in, recrawl your site, and then over time you're really continuing to waterfall down a virtuous cycle because you can keep on linking out to your pillar pages, your new clusters, and give them PageRank from your existing content. So here we go. This is the meaty part of the session. So HubSpot, this is how we did it. A lot of the times our scale for us is an advantage. We can run tests more quickly because we have a larger audience. We can get more, we can get more boosted from one email than maybe one of your companies would be able to. This is a perfect example of how our scale really came back and bit us in the buzz. A lot of this process had to be manual. This is a very sort of new topic. There's not a ton of great solutions out there on the market that can automate the work that needs to be done to architect your site this way. And so we had to go through all this manually and do it. But if you guys have smaller sites, start doing it now. The earlier you start this, the easier it will be. The fewer pages you have to manage, the faster you will be able to get up and running with this strategy. So back in October 2016, Emma told me I was in charge of organic. And as I said before, I'm not an SEO expert, but luckily there's a ton of people at HubSpot who are very smart and willing to take coffees with me. So I talked to them, I did some research, and this is what we came up with. Project Christmas Parrot. If you haven't spent any time with HubSpotters, you should know that we are huge, huge fans of quirky internal project names. It's how marketers get their kicks in. Are there any Slack users in the room? Okay, a few. So for those of you who aren't and maybe aren't familiar with the party parrot emojis, they're a little fun family of just little dancing parrots. The Christmas parrot is my favorite. He wears a little Santa hat. He grooves around. So this is a way that for me to embark in a fun way on a project that I knew was going to be very stressful and long. So the first thing that you have to do when you're starting this project is define your cloba clusters. And there's two paths that you can go down here. Before you do this, you really want to look at the existing universe of content that you have, everything that exists now. But you also want to look at all the stuff that you should be writing about that you haven't yet. So things that are in your space that maybe you don't have a specific feature or offering on, things that you know your users care about that are in your space. You want to own as much of your industry as possible. So there's two paths you can go down when you're defining topic clusters. The first is the features Topics angles path. So what I mean by this is take a look at your product or your service offerings and think about what specific problems those features solve. What specific problems are your service offerings aiming to solve? And so figure out related angles, sorry, related topics. And then from there you can go drill down one level further into the angles that will become eventually your cluster posts. So, for example, this is an example from the Trello board that I was using. Take Tasks is a feature of our HubSpot sales product. It just allows salespeople to set themselves reminders to send that email, call that prospect. And so I thought about what people would be using tasks for. You can assign tasks to people, so team selling is on there. You have to manage your time in sales, so you don't have time to really rely on what's in your head, you really have to write it all down. So time management and sales is up there. Task management is another obvious one. And so those were our feature topic clusters. The second angle you can go down is the consumer demand path. What I mean by consumer demand is think about who you serve. What do they care about? Is it necessarily related to the service that you offer? It can or it can't be. So back to our orthodontist. You know, an orthodontist is someone who aligns teeth, sets, braces. He's not necessarily someone who does dental cleaning, oral hygiene, but obviously everyone has teeth, everyone cares about them. And if you want to be an orthodontist and be really successful and own your space, you should be writing about dental health too. You shouldn't just be writing about braces and retainers and Invisalign and everything else that's out there. And so this really, really makes you that authority. And you will be more likely to rise and surf every time if you were able to capture a larger share of the market. So that's the first thing. Define your topic clusters, write them out in a list. The second thing you want to do, and this is the part that is kind of going to suck if you have a really, really large body of work, you gotta tag your posts. So this is a screenshot of what I was doing back in October. I had a list of all the URLs on our blog. I clicked into every single one and I tagged them with a cluster. Now, if you remember earlier, I said a cluster probably should not be more than 20 or 30 pages. This is because, again, it keeps that page rank distributed to a subset of pages that is relatively small. We have 14,000 posts on the HubSpot blog. So as you can imagine, we had something like 400 clusters, at least to start. And that was just the ones that we could identify, sort of relying on the power of our own editorial experience. So this is going to be a fairly manual process. I will warn you, decision fatigue definitely plays a role here, but you're going to have to go through and tag everything to the best of your ability. The best thing you can do here is keep the group of people who are doing this relatively small so that you know that you're all sort of aligned on how you're thinking about a topic cluster and what is defined as what. And that way you can keep that ranking or that categorization relatively consistent. Then you take all those URLs and you feed them into. In our case Semrush. Now I don't know if anyone is here from Semrush or if any of you guys saw the sponsor booth. They did not pay me to say this. I did not even know they were going to be a sponsor at this event. This tool saved my life. There are so many things you can do with Semrush. What we primarily use it for is running posts through running URLs through there, seeing what URLs are ranking for 1 to 10, 10 to 20, 20 to 30. This lets us know where our low hanging fruit is. This lets us know what keywords we're already owning and don't really need to do much more work on where our gaps are and where we do need to do a lot of work on. But just take an audit of the health of your site. Pick out pages that you think are pillar pages, things that are already ranking, things that seem to have really good content, and see how they're being evaluated by Google. What you can do here is don't worry if you don't have a pillar page for a cluster. Don't worry if you only have a pillar page and no supporting content. That comes later in the content creation stage. What you really want to do right now is audit the landscape of what's out there. Auditing the landscape of what you have to work with. The next thing that you're going to have to contend with are gaps and duplications. So this is a sales blog post. 29 sales email subject lines. I don't know if any of you work closely with salespeople, but their jobs are really, really hard. Anything you can do to provide them quick value is always going to be a win. And we figured that out pretty early on. So we have 120 posts on the sales blog that are dedicated just to sales emails and sales email subject lines, which is insane. There's not 120 different things that you can write about subject lines. So we had a lot of duplication. What you can do there is use a 301 redirect. So again you're going to feed your URLs from your cluster into some kind of tool that's going to tell you how they're ranking. Look at the post that seems similar. Look at the post where there's content overlap, where the topic is really just the same thing. I guarantee you probably have written the same post twice without even knowing it and redirect those posts. There's two ways you can sort of prioritize this. If one of Those duplicate posts is already ranking. Redirect the rest of your pages to that post that passes back all of that link juice to that post that's already ranking. Gives it a nice boost, moves it up the serps. If nothing is ranking, and this is the case for a lot of our posts as well, redirect everything to that oldest post. Google always rewards link age, and so if nothing is ranking, you can't really go wrong by just redirecting everything to the oldest post. Also, what you have to do in this process is a little bit of editorial work. So not everything is going to be exact duplicate. When you're redirecting those posts, there may be content that gets lost. Just do some editorial work. Update that main page to make it better. And so in that process of streamlining your site, you're also improving your own content. Next thing you want to do is optimize within a cluster. So what do I mean here? This is where you get to relink your entire site. This is a beast of a project, so I would suggest you get people to help you. Or a gallon of coffee. What you want to do is pick your pillar pages, define your clusters, and then make sure that all those clusters link back to that pillar page. You should also, for every post, look at the three or four other posts that are in that cluster that will provide your reader the most value. If they landed on your blog post on your site page and were clicking through your internal links, what's going to be the continuation of that story that you're telling on the page they land on? How do you get them to stay on your site? To visit more pages, to keep your bounce rate down, to keep your conversion rates up? You really want to provide a valuable experience for your visitors, so. So be strategic about what post you're linking to. You don't want to just look for the anchor text that you think a post should be associated with and put a link in wherever that exists. It really is about that reader experience. The second thing you want to do is when you're optimizing, when you're relinking, remove all other internal links. You don't want to link outside your cluster. You can link externally, that's completely fine, but internally it'll confuse Google. And a lot of people ask me this, so I wanted to just get it out in the beginning. When you are relinking these posts, there's a good rule of thumb, no more than one link per 150 words. SEO is not a science. It's always changing. But you don't want to link stuff. You don't want to go completely crazy when you're re linking your website. We talked about consolidation already. We talked about combining that duplicate content. And the last thing you can do is historical optimization. So if you've spent any time at inbound anytime on our blog, you're probably familiar with this term. For those of you who aren't, historical optimization refers to the process of looking at your old content. Maybe posts are getting a lot of organic traffic, but they're not converting super well. Maybe a post has been around for a while and it's converting pretty well, but it's just starting to rise in organic search and it's still maybe on page seven. Posts that are starting to trend upward are a great opportunity for you to update that content and republish it under the same URL. The reason for this is because Google rewards freshness along with linkage. They want old links, but they also want those links to be constantly refreshed so that they can be assured that the information there is the most relevant to their readers. So when you see posts that are a little bit out of date but have evergreen content, when you see posts that are just starting to get organic traffic and they're not great content wise, you can re optimize them, republish them, and that's historical optimization. And then the last thing that you're going to discover when you're going through this process is that you probably have a lot of content gaps. There's probably areas where you really haven't written a lot. We internally call these the eat your vegetable posts. You know, you only have so many hours in a day. You're only going to spend time on those things that are going to be a quick hit of traffic in email, a quick hit of traffic on social, whatever it is. But you have to have those vegetable posts. You have to own your entire space. What the topic cluster strategy lets you do is it exposes those gaps in your content strategy. You can prioritize them by search volume, by, you know, maybe you have a product roadmap that is tied to these topics you've written about. It really allows you to prioritize how you're going to create that new content. And then the big kahuna of the question, how the hell do you track this? So as marketers, we all know this challenge. A lot of our work sits at the top of the funnel. A lot of our work is very, very, very difficult to quantify. If you work on brand awareness, if you work on Anything that's not lead generated user acquisition, it can be very hard to see the impact of what you're doing. There are a couple different tools that you can use out there. The one that we use internally is accuranker. Accuranker lets you track the rankings of your posts over time. It lets you track the rankings of your competitors URLs over time as well. It will send you alerts if there's a precipitous drop, if there's a dramatic increase in your rank and you can dig in and do a little bit more work there. So the way that we track clusters is really by using the pillar page as a proxy for the health of the entire cluster. Once you get everything set up, once everything is pointing back at that pillar page, you should have the optimal structure to sort of tell Google, here's what's important, here's what you should surface. It takes a little bit of time to kick in, but once that starts working, you're going to start seeing your pillar page URLs rise in the rank. And so we use three different factors to keep track of when we're thinking about rank. One is just volume of traffic, is volume going up. As you rise up the serps volume is linearly correlated. We all know that. The second thing is, is the relative rank. Is the rank going up and the rank should be. And the third thing is click through rate. Look at the click through rate of your link from serps. If people are not clicking through, maybe you're going for the wrong term. Maybe the content isn't really up to snuff. Maybe your meta description, like we talked about, not up to par. So there's a lot of things you can do from those three metrics that will tell you how your cluster is performing. If you are seeing dramatic changes, if you are seeing changes for good and bad, you can then dig into those individual links and see, is there something going on here? Did I forget to hit publish on these new links? What can I do there? But the only way to really keep yourself sane is to track the rankings of those pillar page URLs and dig in when necessary. You don't want to be looking at a spreadsheet of 20,000 URLs every day. You're going to go insane. So that is a process that we are using today. And you know, every slide or every presentation at Inbound ends with a final thought, ends with, you know, something uplifting or inspirational. And I don't do TED talks. I'm not, you know, a super experienced public speaker, but I'm going to try to leave you with this. Back when, two years ago, when I was new at HubSpot and I was writing seven days a week and I wasn't, I had no idea, you know, what purpose my writing served. I didn't know whether this post was going to be, you know, a hit in serps. I didn't know whether it was going to do work for us over time. I was just guessing and checking. That's really, really, really frustrating. All of you wear multiple hats. All of you have a lot of things to do. And the worst thing to feel like is that you spent a week on an amazing project and you thought it was going to be great, but there was no strategy behind it. There was no sort of reason for what you were doing. And so you end up doing a lot of stupid shit. You end up doing a lot of stuff that wastes your time and. But you don't know that. You don't know that until after you've done it. And that's a really, really hard learning curve. That's a really, really big waste of your time. Dharmesh and Brian talked about having vectors aligned. You can't align your vectors when you don't know the universe that you're working in. We re architected our site, we re architected our content gaps. We understood where we needed to write new posts where we already had excess content. Now, every single post that we write on the blog, every single landing page we create, every single site page has a purpose. Purpose. We know whether this is intended to be a pillar page. We know whether this is supposed to be a supporting page. We know what pages to link to. We know what pages to link from. We understand how to expect this trend of traffic to go up over time, to go down over time. We know what to look for. And when you know those early warning signs, when you know the trailing indicators of success, you become that much more powerful. You stop doing the stupid shit. You stop wasting your time on things that might not work. And you get your time back. You get more control over your role. You get more control over the impact that you have at your company. And hopefully that's a mission that all of us can get behind. So thank you so much all for coming.