How MarketMuse is Shaping the Future of Content with Jeff Coyle

All right, welcome to Hack to the Future with Kyle Roof. I am Kyle Roof and today we have Jeff Coyle, who's the co-founder and CSO at MarketMuse. Hi Jeff, thanks for coming.

Hey Kyle, it's great to be here. Great to see you.

So, Jeff, the first question is when you started MarketMuse, you were trying to come up with a name in 2015. Were you guys huge fans of the Brit pop-rock band Muse? And were you trying to fit that into the name to catch their wave?

Not. But I do not mind the band Muse. I'm more of a Pulp and Smiths fan if you want to go into more detail about that. I was working at a company called, my first business was called Knowledge Storm. We were sold to Tech Target, who many people are probably familiar with, a very large publishing network of B2B tech that I ran in-house, as well as social paid and community and several others. When I left TechTarget to go work at a private equity firm, just before that, I was looking for a solution that would classify documents. So a semantic solution would do auto-tagging on documents because we had hundreds of thousands of documents and taxonomies and those types of things at TechTarget. And so I was researching solutions that were in the knowledge graph space early on. One of them was PoolParty, for example, who is still around and doing amazing things. And I found an interview that this gentleman, Aki blog had done online somewhere where he was talking about it. And I called him up and I said, Hey, does your software classify documents? In the earliest version, he didn't even have a user interface for it. And he's like, no, it doesn't. But here's what it does do. And I'm like, my gosh, do you know what that means? Three days later, I was on a plane to Boston, and about six days later, I was raising money and joining as a late co-founder. So the original name of the software was Informite. Some people might not know that, I -N -F -O -R -M-I -T -E, but we saw in testing that people thought it was some sort of bug. We were thinking of it as some sort of, like we were supposed to say some sort of like element of information. That was the theory.

So then we switched, basically we were trying to get into content strategy before it was a thing. So it was like, we want to tell you, we want to be your muse for information that's going to allow your business to go to market. And that's where it comes from.

I like that a lot. It's good to test names. It's like the classic Nova, Ford sending it down to Mexico and it means no go. When you started MarketMuse, so you've got this great idea, you've got the software going, and you've got the interface. Did you ever have kind of like a now what moment? Or I'm like, how do we get people into the tool? How do we get them to adopt this concept when, as you said, people weren't doing this at all?

Yeah.

it was so painful early on. I mean, we were very reliant on networking. And I know you're something you're very keen on as well. It was many of our early customers, because the software was the core of the technology, what we have patented, and it was so strong. And it still is the foundation of the software in parts. The concept was so strong, but no one was doing it. I mean, it was very advanced publishers. You know, it was like the Trey Jones is of the world, the people who were thinking about this far before it was a thing when it was really easy to win. And, you know, major publishers, but it was our network. So it was, you know, customers of ours at previous companies. It was, you know, peers and friends who were early evangelists. And when they were the quick success that they saw from touching pages and making them better, but then also using the technology to advise on what they should create or update, what they should do. That was the key. It wasn't just the page level, it was more like that site section or the site level. Once they got that, it allowed us to get that network effect going. A great percentage of our customers in the first few years were word of mouth from existing successful customers or they changed jobs and brought it with them. So that was a big, huge win for us early on. But yeah, the adoption part was so hard and what we ended up having to do was win with people early on. So we had to provide services, content strategy, and content plans. We had to deliver content briefs manually. And we were building content briefs and templates manually just because the software hadn't caught up. It took about two years for us to get the first brief in the software. So, you know, there was the way we grew early on, unfortunately, was humans and manually writing stuff on Word docs and Google docs and, and, you know, selling that as, you know, technology-enabled services.

If you can make it on that, that is proof of what you have. You can kind of do it without. Once you kind of got into maybe you want to get past our network kind of phase, but you're still in that real early growth phase as well, was there something you found in marketing that was maybe unconventional or something you found that was successful or maybe something you found that wasn't successful at all?

Right, exactly.

Well, I mean, you know, I dislike the phrase, but it's a lot of eating one's dog food. So it's owning content strategy. It's owning, you know, content optimization. It's owning those topics. And, you know, part of that was also, gosh, go type in Jeff Coil podcast or Jeff Coil AI, and you'll find about 3000 hours of me talking on shows like this and, you know, being at events like Content Marketing World and PubCon, you know, were also a big part of this, a big part of our success. And I think people were fed up with the perception that copying others and things that were very derivative was kind of the norm, or like cheating the system almost. And so very early on, took the position of We're quality first, we're never going to give you insights about how to cheat. Because cheaters, it's time. What do they say? Google, you know, the Google research scientists that cheating is time correlative, right? Over time, we're gonna catch you. And that it's on a wall inside one of the offices over there cheating is time correlative. And so we were early on said, hey, we want to, we want to focus on quality, we want to stand on that hill whether we live or die on that hill, it's going to be about quality, it's going to be about differentiation. And that allowed us to be kind of the don't in the market. Yeah, so.

Sure. For those that are unfamiliar with marketing, can you give me the 30-second elevator on it?

Sure, it's a content intelligence software platform that sets the standard for content quality. So basically we can give you insights about how to build content that tells the story of expertise. And then we've expanded that to build, we have the best in class briefs, very comprehensive things. We're releasing a new thing coming in the next few weeks that's just gonna blow people's minds from a brief perspective. I'm very excited about it. We also have the only in-market solution that's site level or network level. So it will do some unique things and you'll probably be keen on this if we personalize metrics like difficulty. So we're giving you the recommendation that specifically says, Hey, for your site, this is gonna be hard. You know, you're a website about podcast microphones, and you want to own soft-serve ice cream? Well, it's gonna be tough. But if your podcast microphones, you might be able to cover you know, podcast studio and you can build that cluster of content a little bit easier. So we give that site-level personalized difficulty metrics and try to give estimates on how much content's needed so you can make those investments in less of a black box.

Would you say it's fair that a big problem that the tool solves is where, especially people that are kind of new to any of this, they use maybe like a keyword research tool and they get a list of keywords but then don't know what to do with it after that? And that's where some market, you can come in and say like, okay, sure, you've got these keywords, maybe you like this list, but now what's the next step?.

Yeah, I think the keyword research is a is, you three or four prongs, right? So you have the, you know, you have the substring matching, right? So you keyword variants, right? So it's you know, dog food, and it's like the best dog food. Well, the key is that variants tell you about, you know, they tell you about intent, right? So superlative best is a superlative. But the key to what MarketMuse also provides is if I were an expert on dog food. What are the things that my site would naturally have on it? Would I have coverage of particular types of dogs? So what we're trying to give is not only the page, this page but also what's the collection that tells the story that you know what you're talking about. Because what are the odds that you're gonna know, you're gonna be the expert on who makes the best dog food for elderly Boston Terriers if you don't have any other coverage about it? And the rest of your site is about, you know, Christina Aguilera. It's just you're you're not telling that story as accurately as you'd like. A lot of people ran into that problem early on with they got this list of keywords, they sort descend by search volume, we call it the sort descend Google GAKP. And you can see a page for each one of those words on their site. And they thought that that's what a cluster was. And then their site got walloped. And they're like, gosh, what did I do wrong? Well, you know, we were trying to answer that.

What are some lesser-known applications or benefits that people could tap into?

If there's one big problem we've had and it's that, you know, does a lot, right? The personalized difficulty is one where it's a key differentiator. You know, like I mentioned, but a couple of those site-level cluster analyses. So, you know, it takes about 10, 15 hours to look at a site properly and understand what an existing cluster looks like. And one of our solutions, our heat map, our site level heat map solution is about a minute and a half. So going from 10 hours to a minute and a half is pretty hot. And just knowing what those are, and then not just that, but to be able to grade those pages and understand how far away they are conceptually from the core. And there are some technologies in focus crawling. I had built a focus crawl platform in the mid-2000s and we used some of those concepts for radius-based crawling and some things that allow us to, you know, take a different approach to this than just, correlative, which can have a lot of value. It's just when it's paired with other concepts, it can be a beast. And that's that's the one-two punches is where we're focused. So I'd say site-level cluster analysis and internal linking is one where the current market is stuck on showing you your most relevant pages, but If you've been doing this for a while, you know that Google's internal page rank technology has connections to knowledge graphs and topic models. So another technology we have is we give you simple internal linking, but also a more nuanced, complex internal linking recommendation. So those are two things that we currently have. And then I'd say the third one I'd mention is our upcoming release, a re-release of content briefs is shocking. It gives me the chills. Just even when I look at it. I'm a little bit biased, but trust me, when you when you see these things, you're gonna it'll kind of shake in your tree. That's for sure.

I love it.

Having a link interlinking suggestion tool is fantastic because that is a problem that everyone struggles with. Especially if they've got an existing site and they realize that their internal linking has just been random at best. Just how to organize their site accordingly is a massive benefit.

We see a lot of cases where we call it the cliff. And it's the cliff that is the publisher's demon. So if you are a victim of this, this is we can fix you you publish new content and you link backward editorially, but you never go back and connect your old stuff to your new stuff, except through run of site links and more programmatic, you're not going back into contextual. So when we see a cliff situation like that, we know we're going to just blow it up for a client. So it's a fun nuance. So if you're a victim of, of the cliff, you don't go back, you know, you can use, there's a lot of solutions out there, for internal linking. Now, most of them are relevant space, but if you do take it a step further and look at internal PR, you can do some damage real quick. It's lots of, lots of fun.

I'm confident that's a problem that about 102 % of the people listening have. Nobody goes back and does any. No, I can tell you for a fact. For a fact. So now we're at the portion of the show where we ask, what is the best beer and wine? My crack research team scoured the top of your LinkedIn profile and found that you are the co-owner or co-founder, excuse me, of Silver Bluff Brewing. My crack research team scawered the top of your LinkedIn and found that you are the co owner of Silver Bluff. So which is the best beer that we should go for when we stop by.

Man, thanks for looking that up. That was a tap handle I just happened to have on hand. That's the original tap handle for the generic tap. So yeah, Silver Bluff, we're one of the fastest-growing breweries in Georgia. We won a US Open 2020 medal and a World Beer Cup Silver Brewski Beer of the Year last year. So our top performer in the competition, that one that won is our Mexican larger.

Hahaha! Nice. Nice. Nice.

Nice.

if you know beer, it's if you know what, like a Clara, like a Corona, we're like an ambar, like an obscure, a Bohemia, obscure or a Negro Modelo. We're, we're right down the middle. So it's, it's, it's, it's a kind of a new thing. We're, we're not as light as a Cam Corona. We're not as dark as a Negro Modelo. So that one's a go-to if you're, if you're really into the kind of, if you're outside in hot weather, which I know you are, that'd be the one you'd want to hot high on.

Okay.

We have a little bit here. We have a little bit here.

My current go though is we have a double IPA called Beekeeper. It's made with local Georgia honey. And it's about 8 .8. So it'll knock your teeth out. But if you pay point 80 alcohol by volume, but if you know, in small doses, it can be a delightful, delightful thing to have. You might want to end the night with that one. So Don't go for three beekeepers and we're carrying you out the door.

I've never gotten into the IPAs personally, but I am a huge fan of the Mexican-style loggers. So that sounds, the first one sounds perfect for me.

awesome. Cool. Well, don't, don't try, hey Kyle, don't try to compete. Don't try to compete for Mexican lager in Google. Cause I don't want, I don't want you competing with me there. Come on. That's one thing I do not want.

Are you taking that one over?

I was going to say, that I know a company that can help your website out with the clustering of which topics to cover.

Yeah, that's a hashtag rank brag for that one. For all of you who go into your Google and type Mexican lager and wonder, hey, why is this regional Georgia brewery ranking number one? There's an answer to that. And you're listening to it. Yeah.

That's awesome. Yeah, proof of concept right there. Yeah, proof of concept. AI is on everybody's mind. It's a huge topic right now. And I know that market use is one of the first tools to implement AI features and take advantage of the AI available. I know nobody's got a crystal ball, but how are you feeling? What's your feeling on AI in general? Like where it is right now, where might it be going?

it's, there's a combination every day. It changes. I mean, you get a fear. And yeah, so we were, we were an AI-first company. So, you know, the key thing is people think that AI is chat GPT, right? They're like equating those two things, but you know, artificial intelligence branches of AI have been around for a long time. And so our core is based on, you know, four principles of AI and topic modeling, graph, graph theory, link, link analysis and, and, you know, linguistics, parts of linguistics, and natural language processing. So we've always been thinking critically about that. We built a large language model technology ourselves. It was called First Draft. It came out about when GPT -2 came out. We killed it because it was too hard to maintain. It was ugly. It was so expensive. I mean, it would cost us about 20 bucks to make a page draft in the olden days. Now that would be a fraction of a cent, a fraction of a fraction of a cent, right? So fast forward to now, like the fear I have is frankly, it's user error. It's that people have this perception that they should jump to the end. But these editorial practices, these research practices, they haven't become invalid. And so when people are using AI at each stage to make that stage fast and beautiful, and they are keeping their knowledge in the loop somehow, whether it's through an agent, which means basically like an AI assistant that's touching a data set, right? Whether they're, if they're keeping people in, if they're keeping in the loop, what comes out of the end can be beautiful. It can be way better than you could have done on your own. I mean, we all use Grammarly. We all use Google Smart Compose, right? So those are AI, right? Those are AIs that make your stuff better. So there are ways of using it. My, so my biggest fear is user error and the belief that you can rush to the end and it's going to be okay. And people look at that final document, they're like, this is terrible. It's so mediocre. And they don't realize that they, you know, they could have done a better job to get there and still had great success. And, you know, that's so, so those are the things that I fear most. I'm very excited about what is going to come with general intelligence, and artificial general intelligence in the next few years.

how these technologies are going to make these things enabled. And then from the SEO and me, and you'll probably appreciate this, it's a whole different battleground. It's like a whole new gambit. It's a new melee, right? So fighting for the AI overview is just another data source, right? It's using retrieval augmented generation, but I'm fighting for that. It's a separate battleground. And so if there's one thing that, you know, good marketers do, Do they like more channels? So for me, that's how I think about it.

I completely agree. I also agree with the first part a lot. I have this feeling there's this perception that AI will help you do good SEO. And what I've said is that if you're bad at SEO, all it's gonna do is help you do bad SEO faster. You know, like it's not, yeah. I mean, that's all it's gonna do for you. Yeah.

That's the best thing ever. It might make you better. And I'll use I'll improperly use the grammar on purpose to say it might make you better at it. Because frankly, that's what's happening. You're going grammar south is a lot of these folks. And, you know, they're thinking that they're going to get more volume at the same quality. And the reality is, you can like 10 extra productivity and improve quality. You just don't know how to do it yet. But people will help you, you can get there. It's just a lot of hard work. And you have to make sure your company doesn't get poisoned by bad processes. Like the CEOs and the CMOs are saying, you know like search is dead or AI is bad. It's like, Nah, man, search isn't dead. And AI is not bad. It's just the things you've seen would make your company fail. And you've got to take a different approach to evaluation and get to understanding because you can't fear this stuff, it's gonna eat you.

You mentioned content at scale. Do you have thoughts on humanizing content at scale?

Yeah, that's exactly what I'm building right now. So, yeah, I stay tuned. Stay tuned. Well, you know, the humanizing content at scale is that's the master key, right? And it is, the master key requires, and I have a, I have a couple of presentations last year, my presentation that I had, that I gave at a make a marketing artificial intelligence conference and see content marketing world, two different versions was,

Okay. Stay tuned.

the rise of the subject matter expert, right? And then I gave one earlier this year that was called the subject matter expert has risen because that's the key. The key is humanizing humans in the loop. The key is you don't have the subject matter expert input. You're going to get garbage outputs. So you've got to find them, hunt for them, be them. Like if I'm, you know, my hobbies, right? You mentioned one of them root beer and brewing. But, you know, for example, if I think I'm going to write the world's best article about how to make adjustments to a totalizer, right, if I think GPT is going to write that article, you know, it's not going to give the nuance that somebody who's lived on a brew deck, you know, and and and dealt with a total that dealt with flow controls breaking, right now, what that experience was like. They don't know, you know, what it is. All they can tell me is to get to a point. And then if you've had those experiences and you have that expertise, that's your bullet. Those are the bullets in your gun. So I think, you know, the key is subject matter experts are all around you. You got to find them and you got to make sure that their stories check out. When you publish a top 10, I use this as one of my examples, Kyle, so you asked about beer. I just wanna top eight beers of this particular style, right? And two of them haven't been produced since 2008, right? Google's getting better and they're gonna be able to find that stuff. They already can to some extent, but when your story doesn't check out, it tells the linking of your site tells the story it doesn't check out, the structure of your pages doesn't check out, and when the content you know, follows through and it tells a story that you're, you know, using Google AdWords keyword planner for all your keyword planning. I mean, you're just setting yourself up for a bad situation. And that works with whether you're selling socks or $100,000 software. Subject matter expertise is a must-have. Hug those people, don't let them leave. Don't let them get frustrated. If they don't want to write, do interviews with them, capture their expertise, and get it to the page. If you got a product manager, if you got a CEO who doesn't have time, the head of your law firm, get every piece of information out of their squishy brain somehow on your site. That's how you're gonna win in the end.

The last question, Jeff, you speak at conferences. You mentioned a couple of conferences you just spoke at. You speak all around the world. Favorite conference or favorite conference location? Where should we go?

I wish I could get out to your neck of the woods. I've never actually been out there. I'm not in Australia, New Zealand, been up to, you know, Japan and stuff. I've never been in your neck of the woods. And I've heard amazing things., oddly enough, my favorite events of the year are in one of them I look forward to going to Cleveland every year for a marketing artificial intelligence conference, even though it's you know, people might hate Cleveland's great beer scene.

You should.

And I just, I don't know, I like the town. It's a beautiful place. But, Vegas is also one of my loves, and Rhodium, is the Rhodium community for listeners of your show. If you're not familiar with Rhodium, it's an amazing mastermind and community that was originally built on buyers and sellers of web properties and businesses, but it's expanded to founders, owners, and practitioners in our space. So if you have ever bought a domain, sold the domain, or rounded up a Business, you know, that Rhodium is an awesome group and a couple of our mutual friends are members as well. Yeah, it's just it's a good place. Good folks. And it's it's one of the most caring groups, caring and sharing groups. And I know that sounds a little bit silly to say in this world, but certainly, been most important part of my health and well-being is that community. So my favorite event is their event. And that's just because I, I know what I'm getting. And I know that I'm going to feel like a million bucks when I leave and I love Vegas too.

Win-win. Win-win. that's awesome. If people want to get in touch with you, how can they do that?

Jeffery underscores coil on the ex Twitter, Jeff at marketmuse.com. And LinkedIn, I'll answer everything. If I could give you my cell phone, I would I love this stuff. I live and love everything content, everything SEO, and I think product management. So no questions too crazy. I mean, you don't want to know how many people just say, Hey, can you check out this page? Can you check out this site? And they can't believe it when I do and I'm like, yeah, here's your five problems that I saved you 15,000 bucks. All right, see you later. Bye. So yeah, please give me a ring. any of those mechanisms, you know, and, or, you know, finding mutual connection, and yeah, check out MarketMuse. Some of the things we have coming in the next few, next months are, are staggering, market breakers in some cases. And, I'm so excited about it. I couldn't be more excited about what's to come at MarketMuse and then with our industry too, just leveling up the quality of the work being done. It just, makes me so happy. And really quality being cool. I never thought I'd get there. 10 years ago, I was like, quality is important. And everybody's like, yo, shut up. And now everybody's like, quality is a must, you must use linguistics. I'm like, my gosh.

I can't believe it. It's like, you know, I just got to make sure I don't get run over, right? When everybody else thinks something's cool, because, you know, there's just so many bright minds. But yeah, quality is important now. And, you know, just it's been a fun experience of being there as it gets there.

Well, Jeff, thanks so much for being on, and thanks everyone for listening to Hack to the Future and we'll talk to you next time. Take care.

Such a pleasure, Kyle.