 Hey everybody, it's me again. Ready for the next speaker? He is one of the most trusted scientists across the entire SEO industry. Dr. Pete Myers, his data insights have been used by thousands of SEOs, including myself. He is the inventor of Mozcast, the original Google algorithm weather gauge. He's an expert on Google updates. I often have a lot of questions for myself. For nine years, he's worked as a Moz scientist at Moz. Moz scientist, mad scientist at Moz, because he's truly mad. And he also, an interesting tidbit, he also just earned his black belt in Taekwondo. So don't mess with him. So Dr. Pete's title is called rule your rivals from data to action. I don't want to give too much away, but Dr. Pete's been working on new ways to do smarter competitive research for SEO. It works really cool. If you've ever wanted to outrank your competition, you're going to enjoy this talk. Please welcome Dr. Pete Myers. Hey everybody, I'm Dr. Pete, the marketing scientist for Moz. I welcome you back to a second year of MozCon virtual. Thanks for joining me. I keep feeling like I should say something profound about the last year. And to be honest, whenever somebody asked me even the simplest question, like how are you, it turns into a 25 minute monologue and I only have 30 minutes. So I'm not going to subject you to that today. I just want to say I'm glad you're here, not just for the show, but I'm glad you're here in the broad sense of the word. Unfortunately, a lot of people aren't with us. And I'm hoping that we can see each other in person when the time is right and that that won't be too long. So thanks for coming. My title of the day, rule your rivals is purposely a bit dramatic and hey, I'm a marketer. But the reality is I've been exploring the competitive analysis space for the last three or four years now. And I've come to a couple conclusions. And one is that competitive analysis really isn't about beating the competition. I don't care how much traffic and sales someone else gets, right? I care about what I'm not getting. And yeah, there can be a connection between those things, but a lot of us are smaller medium fish in a pretty big pond. And so it's really about what am I missing? What are the gaps in my content? What are the gaps in my keyword strategy? Why can what other people are doing inform me about that? And not just about what they're doing, but where I'm underperforming. So Maz is a great example because we've been around for years. We've been writing a lot of content. There's very few topics we're interested in or our audience is interested in that we've never talked about, right? But a lot of times we do have content that's outdated. Maybe it wasn't that great. Maybe we've written about 20 times and just alluded things. And so sometimes our competitors can be a way of revealing that weakness in us where we can see that, hey, you know what? They're ranking three or four for this pretty competitive thing. And we're down at number 27. Why isn't that thing we wrote in the past still doing well? And what can that tell us? And so I think there's a lot of strategy to be learned from competitive analysis if we don't focus on that vanity side of doing better than someone else because it's really about outperforming where we're at now and finding the gaps in our own strategy or the pieces that might be missing. So I want to talk about kind of our traditional approach and I'm going to call it the avalanche because that's where we end up. We end up buried in data and we think that's good. And even as a data scientist, I got to say quite often what I hear from people is we did all this work and we created this competitive analysis and we handed it off to someone or some team or the CMO and nothing ever happened with it because it's fundamentally inactionable and I want to talk about that for a minute. So let's say we're just getting started. We pull up our favorite keyword research tool. Obviously I work for Moz, but I know you use other tools. That's okay. This is the keyword gap analysis tool, the current one. I think I have a handle on our competitors. Oh my God, I'm going to put in SEMrush. I'm going to put in Ahrefs. And I'm going to look at where these two intersect, right? And I get back 23,000 keywords. I think great. Oh, I'll just data. And what's funny is some of you are probably sitting there right now thinking 23,000. Come on, 50 or 100 or half a million. You know, you're scoffing at this. And we get carried away, right? We get caught up in those numbers. And so I want to talk about what is the value of that? What is the value of that big number? Well, what do we do? We export it, right? I don't want to scroll through all these things on the UI. And we put it somewhere like Excel or Power BI or GDS or in a SQL database. And we feel like we've accomplished something. And then we do all the work of filling in these columns, right? Volume and difficulty and various traffic indicators maybe. And I can show Maz's ranking. I can show the competitor's ranking. And I think I've accomplished something. And yes, this is a more flexible, usable format. Certainly, I can filter and search and do the things I need and sort. But there's no insight inherently gained by this. And I would actually say that these tools sometimes lead us astray in the sense that I can look at this and think I have a handle on it because I can view these 20 or 50 or whatever I can fit on the first screen and I can run these filters. But they don't really draw insight from that entire list. You know, they really don't make this manageable. It's a tool and that's great. But I haven't really accomplished anything yet. So I did something a little strange just to illustrate the point and kind of for myself as well. I took all these keywords from the export and I pasted them into Word. And it's hard to tell from the image, but I picked a pretty reasonable font size. I think 11 point, pretty normal font, one and a half spacing, kind of like if I were writing a draft of a long document. And what I ended up with was a document that was 96,000 words and 257 pages. So at the end of this competitive analysis, what I've really accomplished is to write a really shitty novel. And you can't read it, but it's quite bad. I've written some bad stuff. So, you know, this is extreme, obviously I'm exaggerating, but what have we gained just by this giant dump of data? And are we fooling ourselves by putting it into tables and adding a bunch of columns, which takes a lot of work too? And so we get to the end and we say, hey, good job team. Good work. You know, we're buried and we can't breathe. We don't know what to do next. And somebody needs to help us out of here. What have we really accomplished? You know, the task is over, but it's something going to be done with this. I want to focus just quickly on a couple of core problems with these large data sets. One is that obviously we focus on volume a lot, make sense, we want more volume to get more traffic, but it tends to lead us towards very broad terms. So if I look at this intersect of our competitors, I'm getting things like HTTPS and EAT and difficulty. Why know why we're there ranking for EAT, right? Expertise, authority, trust, we all know about that. But EAT with no context. If it's not, how does Google use EAT or SEO for EAT? This is completely worthless to us. This is actually going to be extremely competitive and get nothing back because when you go to that SERP, what are you going to see? You're going to see putting food in your mouth and chewing because that's what EAT means to normal people. I'd say we're not normal, but hey, you know, you know as well as I do. So this is not something we want to compete for. So we're going to have to wade through that noise. We're also going to see probably a lot of brand terms. So terms like SEMrush blog or SEMrush login or to lead a project in a trust. Unless I'm just trying to make trouble, these are not things I want to rank for. And that's not just altruism. That's reality. Like if I pull someone in under SEMrush login free to Moz, what am I saying to them? I'm just saying that I'm a liar, right? I'm tricking them. They're not going to convert. They're not going to trust me. This is a poor use of my time. This is counterproductive. And then finally, I think a big problem is this is not really 23,000 distinct concepts. That I'm going to go off and write 23,000 pieces of content for even if I wanted to or had the budget to do that. If you look at something like alternatives to Google search, this longer tail phrase, I have a ton of variants of that in this data set. And those are really all one related concept. And we're not great at combining that data. And yeah, we could do some simple filters, but look down around line five search engines other than Google. This doesn't have the word alternative in it at all. And yet it's fundamentally the same concept. So how do we wade through that? How do we wade through the reality that this set of 23,000 phrases is really maybe dozens or hundreds of concepts that I could target of which I may be picking a handful and that the value of this is not in the giant dump of data but in those final targets? All right. First, I want to start with an assumption. I think we all think we know who our competitors are and our clients and our bosses and whether we're in house or agency doesn't matter. They think they know and certainly we want to respect our institutional knowledge and their institutional knowledge. But quite often what we see in SERPs and what we see in organic is a different picture than what we're seeing in maybe the brick and mortar world. And so I want to give you a few examples. I'm going to show you some prototype screens today. These are not, these are for internal use only. I have some small announcement to make at the end about this but this is a tool we've been working on for a while called True Competitor. We're trying to analyze the SERPs and see who someone's really up against and we have a score called rivalry. I'm not going to talk about that too much right now but basically we're balancing your overlap between keywords, how attractive those keywords are, how relevant those keywords are and whether you're in an authority window that you could compete. You know, if you have a DA of 20 and you're trying to up against Google or YouTube, what's the point? So we're trying to balance all that. But here's the example for State Farm, right? And I go in and I see all State and Geico and Progressive and Nationwide and Farmers and I think, okay, that matches my reality even as a consumer. I have insurance, I watch the commercials but that seems to make sense. And then I see these, Nerd Wallet and Investopedia and I think, okay, this tool is dumb and wrong. And okay, we went through some iterations that were dumb and wrong but I'm much more comfortable with where we are now. And so we dig into that data and we go, okay, Nerd Wallet, are they really a competitor? And what's interesting is if we kind of expand just a sampling of those keywords, we see that all State and Nerd Wallet, they're ranking for pretty similar broad insurance queries, right? Look at Nerd Wallet, life insurance, renters insurance, car insurance quote. These are things that State Farm wants to compete on. You know, these are things that anyone selling product would want to compete on. And so Nerd Wallet is writing content that's potentially covering our products, our competitors' products, covering how-to information, covering general questions about insurance that we really do need to be on top of and we need to know what they're saying about us and we might need to be providing that information. So we shouldn't be dismissing this. There's another interesting example, Tesla. Nowhere on this top list is Ford and Chevy and Rivian and some of these up-and-coming competitors because that's not who Tesla is competing against in the SERPs for the most part. That's not to say that those other competitors aren't real. Obviously those are very important competitors but from an SEO perspective, what I care about is what's appearing on Google. And so I'm really competing with car and driver and motor trend and electric vehicle blogs like Electric. So am I producing the kind of content that's asking people, answering people's questions which may be less about how do I buy a Tesla and maybe more about how long does the charge last and where can I charge my car and understanding electric vehicles which people are still going through the process of. And so it's a different buyer's journey and if I'm competing with the wrong people online, I'm not going to tap into that opportunity. And then finally, there might just be things where they don't all fit into one bucket. So if I look at Moz, okay, we have product competitors like Ahrefs and some Rush, great. We have content competitors like Search Engine Journal and Search Engine Lamp. Those are different strategies that we would pursue. And then we have multiple Google sub-domains we compete with. Well, why? Because we are a product that not only is built on Google but competes with Google. Google is a partner competitor and that's a third different world. And none of these windows are incorrect. Actually, these are three different opportunities that we should be exploring. Finally, I want to point out that we do have in STAT the competitive landscape tool which helps to kind of give a share voice view. One thing that's interesting about STAT's approach is that we do look at the feature landscape and so what you'll find is that your top competitor is almost always Google, which is sometimes a little depressing but is a good reality to be aware of. So check that out if you're a STAT customer. All right, how do we turn this into something actionable? Right, this list of keywords, even if we filter it down, it's just one step. And so we started to build out a prototype to understand the keyword intersect a little better and not just kick you back a big chunk of keywords. And there's a couple of things that we learned along the way. One is a lot of times STOs look at, hey, tell me what I'm missing that my competitors overlap for that I don't rank for at all. And quite honestly, that data can be interesting and is one useful view, but it has a ton of notes. It has a ton of things I just don't care about ranking for that I am purposely not committed to or things that are really much more brand or product related. And so what we found is taking this, what we call the keywords to improve approach that basically says, hey, you know what? I want to know where my competitors are doing well, where I'm showing up, but doing poorly and where I have opportunity to improve that's tangible because I know I can rank for this and I know that I've written about this in the past and I know it's probably relevant. And so this turns out to be a really interesting view. The other thing we wanted to look at was, you know what, if I rank number two currently and my competitors number one, that's necessarily not necessarily a journey I want to take or that's going to have high ROI. But if I'm 13 or 15 or 27 and my competitors number three, that could be something where I really have potential to jump up to page one without a ton of effort and to get salad ROI. And so we started to look at it from the perspective not of total volume, but of lift. And I encourage that, you know, look at what you gain from improving because if you're already near the top, it's not a big deal. If your competitors are at 15, 16, 17, it's probably not that relevant, right? You know, they're not getting that much traffic. This might not be something going at worth going after even if you don't rank at all. But we still have this problem, right? You know, we have Google advanced search, advanced Google search. They're basically the same thing. And in fact, Google often interprets different orders not differently at all. So how do we kind of combine that? And what we realized is that even as we're exploring things like clustering and NLP and some of the more advanced ways to look at how keywords relate, we have a way to kind of cheat. And that way is the content that ranks for it. And so we started to look at not just the keyword gap, but the content gap. And so here we have an example of hrefs have to post on Google advanced search operators. And here's a sampling of the keywords that that ranks for. And that could be ranking on dozens or hundreds of variations. And so what if instead of focusing on single keywords, I focused on that page or on a set of pages to see how they're performing? Well, it turns out we can kind of feed that back in. I'll make that minute. Sorry. We don't have these tools quite available yet. And so you may be wondering, okay, how can I see? Like what is a competitor's best content for any given topic? This is a labor intensive way to do it. And we're hoping to make your job easier in the near future. But this is useful in a way to get at it. One trick I use within Google, take your topic. You might need to run this through a handful of keywords. So Google advanced search, for example, and then use the site operator, but combine your list of competitors, two, three, four, don't go crazy with or. An or has to be all caps in this. So for example, site colon, subrush or site colon atress. This will return the pages ranking for that set of sites on that term. You could do this one by one and just do keyword plus site operator. But what I find useful about using or is now I get kind of a relative comparison. Now I know that that atress page is outperforming these two subrush pages. And I have a sense of where they sit in the hierarchy and I could hit three or four competitors and kind of see what I'm up against. So I find this useful. Do this for yourself. This is a big part of the journey for me. It's quite often we overlook what we've already written and we overlook where we might be covering the topic, but not well. And we end up cannibalizing ourselves and I'm going to do that in this talk and show you how that mistake is made. Quite literally. I also find this trick useful. The dot dot operator looks at a range. So if I say something like Google Advanced Search 2019 dot dot 2021, I'll get anything that mentions 2019, 2020, 2021. There is a date range operator. If you've used it, it's hard to use and it's kind of quirky. It relies on Google knowing when something was written. What I find is that four-digit years are unique enough typically that this is a little better approach that's going to sweep in more. So if you wanted to say, hey, you know what, I only care about what I wrote five years ago. You know, you could say 2010 dot dot 2016 and see if there's something updated you want to update. So I found this useful. Even now, if you use Security Explorer, you could take that sample page you found, that Hress page or whatever it is, plug it into our tool. Some people don't know that in addition to looking at the keywords that a subdomain or domain ranks for, you can target an exact page, same tool, same box to type it in. And we could get back 1100 keywords that page is looking for. And now, yes, this is still a data dump and yes, we still need to filter it. But these 1100 keywords are fairly tightly focused on one topic because we're using all this NLP power Google has, right? Google has determined that this is all relevant to that page. And we can see 172 of those are in the one to three position, 302, four to 10. So if we look at page one, it's more like 470, right? And so now we have a very much more manageable set of keywords that are all relevant to a single topic. All right, so I put this through the test and I wrote three blog posts. Actually, I wrote two and I'll get to the third one in a minute over the last couple of months to try the strategy out. And what I'm hoping to land on and what I landed on for each of these and what we're trying to provide the tools to do, we're not there yet, is to get you back this action plan, get you back what someone on my team calls a content brief. And essentially, it's a set of three things. It's a set of the keywords I'm looking to target. I'm only listing 12 here, but basically trying to show a variety. This could represent hundreds of thousands. But this is just a sampling of what I'm going after. My competitor's best pages. So I'm showing one for Matress here, one for Yoast. I could have four or five, right? It's up to you. But I kind of know what I'm going after and those are a good target for seeing what they're ranking for and kind of what they're doing and what their focus is, how niche did they go and are these top everything guides or are they very focused on a single subtopic? And then I want to understand what my current best piece of content is that Google thinks fits this because I need to know. Is this something I need to rewrite? Is something I need to update? Is it totally irrelevant? And I'm just ranking on the fringe. And so I think this content brief, this overview, can become very actionable. And so in this case, Rand had a Whiteboard Friday. I should have known better not to compete with Rand and Whiteboard Friday. I didn't. It was five years old and I thought, you know what, I've got a fresh take on this. I can compete with this. And I wrote a new post called Long Tail SEO in 2021. How you can have a dollar die trying. And it's really about the evolution of what Long Tail means and how these phrases have both diverged and kind of reconverged. A couple points about the content creation I'm not going to dig into how we wrote all of these blog posts. First of all, I'm kind of outing myself for keyword stuffing, but I want to point out to you that you can cover a wide variety of variations on a phrase. So for example, Long Tail keywords, Long Tail with and without the iPhone, Long Tail of SEO, Long Tail of Search, using natural language and talking like normal people talk that doesn't seem like keyword stuffing because that's what we do. We don't say the same phrase over and over and over again. We say things different way and we use variations naturally. And I think this is something Google is rewarding more and more and makes it easier for us to target variations of a phrase in a single piece of content. We don't have to write 100 pieces of content for 100 Long Tail variants, right? We can do this with one or a couple pieces if the topic focus is good. I also introduced this new thing. You know, we've got the head and the Long Tail and I've heard a lot of words for what's in the middle and I came up with the chalky thorax and I was all proud of myself and oh boy, I ranked right away because nobody had ever talked about this before. So congrats. It's always a fun game for SEO but it leads to no real results at all. And then the nice thing about having this action plan is I know what to track, right? I've plugged all these keywords into my ranked tracking and I created a campaign for MozCon with labels reached of these experiments and I go in and go, oh, wow, number one, number one, number one, number two, I'm feeling good and then I realize almost all of these are Rand's posts of WIPER Friday. And so I really had gone against my own advice and cannibalized myself but it's okay. And what I like about this approach is I know what happened. I know right away, hey, you know what? That post I ran, it's still outperforming me and I need to figure out why. Do I redirect it? Do I link it? Should I be focusing on something that's maybe more of a big picture resource that we could channel all these things into? So I have next steps I can take even if this thing didn't do exactly what I wanted. We're quite often, we come out of competitive analysis and say, I'm gonna write these 10 posts and they don't work and we just move on to the next thing. And I think that's a failing across a lot of content strategy like knowing what's next. So I knew what I was aiming for and I know why Rand's ranking here and I can think about what I want to do to improve that. Interesting to know here that actually a lot of cases Rand's posts and mine were both ranking and so we were double dipping. It's good, well, last might not but you know, it's not all bad. We don't always see that in ranked tracking tools. Okay, second case study. I'm gonna give you the same kind of action plan format. I was going after a post on Google search operators. I've written a number of posts about that and then Google obviously has a resource that ranks very well. But I decided, you know, I'm gonna get a little more niche. I'm gonna talk about kind of a focus topic which is in a meta way competitive research using Google site operators. So I also ran into a problem which is worth noting if you're too focused on one type of content like your blog, I might have missed, for example, that we have a search operator post in the learning center and this is meant to be more permanent resource. So I don't want to compete with that, right? I want to keep that active and relevant. And so my goal is a little different than if maybe I was trying to update something old. I'm really trying to add to the relevance of this topic and keep writing about it and keep it fresh but support these other pieces. I'm gonna talk about that in a second. So I wrote a post called Google advanced search operators for competitive content research. I'll read it if you want to. I'm not gonna dive into that. Once again, I pulled up my rankings and thought, oh, looks pretty good. None of them are my page but a couple of weeks after this they were on the learning center which had in some cases moved up because what I'd done is try to make this part of the picture. I wasn't trying to compete with that content. I linked this to the learning center. I linked this to my older posts and so I'm trying to create more of a content hub and I'm trying to say, hey, we have fresh content on this idea but we also have this evergreen thing. So if I didn't know that, if I didn't know what was out there and I didn't really understand my own content space, I would have just tried to kind of outperform this juggernaut and failed and instead what I was able to do is build something that helps those other pieces and creates a topic cluster that can help Google see that we're still relevant and this can improve the ranking of those evergreen pieces. A little trick has nothing to do with competitive analysis but sometimes I get carried away. I like to throw in a little flash fiction in a post but what I found is that that lends to unique phrases. So here's one where I have a phrase shards of a thousand synonyms. I can use that to very quickly see if this post is ranking but also to very quickly see who's copying my content. So just something to play with. It's kind of fun. It doesn't even have to be fiction. You could just use a unique catchy phrase about something and then that can become a nice signal beacon for maybe who's scraping us or you know how fast things are ranking. And then finally I'm going to push through this one. We did a post on disavow. Marie Haines had a great post but it was seven years old and you know, Marie's written some great stuff on this but so much has changed in that seven years. And so in this one instead of writing this myself I handed it off to my boss Tom Kaeper. Here's a Slack message. I won't read it to you but it's pretty short where I'm basically saying, Hey, we have a post from 2014. It's good, but it's outdated. I don't want to write an ultimate guide. I want you to just refresh this topic with what they need to know now. You know, Google's made a lot of changes to this process. And so here's kind of the topic. Here's our current top page. I'll list a couple. Here's four competitor pages and here's the main keyword ideas I'm going after. And so this turned out to be a really great way to hand this off to somebody and get them moving. Time wrote this post, long story short. These all were rankings for Tom. So I feel a little bad. I didn't do so well. Tom actually did well and part of the reason was we 3.0 and redirected that old post because that was our strategy to update this content. So now we went into this with a more knowledgeable approach. All right, quick recap. First of all, don't make assumptions about your competition. First of all, you might be wrong. Your cert competitors and your in real life brick-and-mortar competitors might be different and you might just not have one set of competitors. You might have different windows like we have here from Oz that are all worth exploring in their own way and all require different strategies and different approaches. Don't write a shitty novel. That's what we do. It doesn't get you anywhere. It's just not actionable. Come up with a plan. This is my draft of what a plan looks like, you know, it's the keywords, the competitors and my current content space that wraps around that. There are other approaches to this but I think this gives you a great context to work with. And this also gives you a really easy starting point for knowing what you're tracking, knowing why it might not work and having a sense of what to do next. And I think that's really valuable in content strategy. If it doesn't work or it does work, what should I do? Maybe I want to link my post to an old post. Maybe I want to link the old post to my post. Maybe I want to 3-1 redirect something that's outdated. Maybe I want to build a higher level page that actually combines all of these different resources into one resource group. If you come into this with a content strategy, you can make those decisions and it doesn't just end with, this didn't rank for what I wanted to. It ends with something that can evolve over time and really perform for you. So real quick, we are going to be launching the True competitor beta, private beta and competitive analysis. We call keyword gap after that. We're hoping, I'm pre-recording, but we're hoping by Moscon that we'll have this. So we will have this form live. It's mz.cm.com-research. I'll post that in the chat when the show goes live and feel free to join this list and we will inform you about these coming products and we're hoping it's not long after Moscon. Thanks a lot all. Thanks for joining me and take care.