Today I have a very special guest: Jeff Coyle. I’m not going to tell you what he does. I’m going to have him tell his story and we will be talking about AI and content marketing and AI-based content management tool.
So welcome, Jeff! So happy to have you on my show. Why don’t you tell everybody, who you are and what you do? I know you have an awesome, awesome product. And I have seen the demo in the past and I loved it. I can’t afford it, but I love it.
Jeff Coyle: Well maybe we’ll talk about that today too. I’m Jeff Coyle, I’m the co-founder and Chief Product Officer for Market Muse and a little bit about my background. I’ve been fortunately or unfortunately in the content strategy, content marketing lead gen and search engine optimization game for about 21 years.
Pam Didner: All good things. I love everything you said: lead gen, content marketing, the demand gen. Yay. My favorite topics. Hard to do it well, but you have a lot to share.
Jeff Coyle: Yeah, I worked at, I’m one of the first people who were selling leads to B2B technology companies before he believes that people even had content. You know, we were convincing large multinational B2B companies to write white papers and syndicate them like, cause they even have them at the time, you know.
Pam Didner: They probably focus on sales, collateral. That’s the only thing (laughs)
Jeff Coyle: A lot of white papers and that’s it. Right? And so we kind of enhanced that space. And then we were acquired by a major publisher. I worked with them in the house for a number of years and during the late phase of that, I was doing basically editorial teams, content teams, product marketing teams, and SEO teams don’t always work together. Pam, are you aware of that? They don’t always work together?
Pam Didner: My God. Don’t talk to me. Whenever you talk about those topics in terms of how to create good content and the multiple teams that need to come together. Yes, I understand. I understand the pain so well.
Jeff Coyle: So I went through that, right? And I knew, that there was a better way to provide insights from research. And I found my cofounder at Market Muse had taken one of those workflows and automated it, the workflow researching topics. So truly telling the story about what it means to be an expert on a topic. And when you turn that and you integrate that information into a research process that speaks the language of the editorial leadership, that speaks the language of the subject matter expert, because you’re not giving them a word because of some esoteric data point.
You’re saying, “Hey, if you’re going to cover this topic comprehensively, you want to make sure that we have pages about this, sections about that–the language of the user” and that became Market Muse was the solution that brings those teams together by providing insights that are going to be awesome, exciting amplifiers for editorial leaders, writers, but also have the same benefits that SEO want–traffic conversions leads engagement. So everyone’s happy and everyone’s working together towards the same data-driven goal for the first time.
So we’ve created an objective measure of content quality and comprehensiveness. So because, hey Pam, I can look at your site and I can say “you are obviously the expert on sales enablement, but here are a few things you haven’t really covered comprehensively. And you’re like, “Oh yeah, those are obvious that I should write about, but I just shot you a spreadsheet with words in it it’s like, “yeah, I’ll get to that next year.” That’s the difference?
Pam Didner: I hear you. So I actually, for listeners, I have seen the demo is actually pretty impressive in terms of taking SEO, taking the initial preparation that you need to do to determine a topic and also create a draft and then move your content along the production stage, if you will, not necessarily like when I say production and it’s not video production, but like creating that draft to get to the finish line.
Jeff Coyle: Yeah. So if you are the person who’s coming up with, what should we create or what should we update today that’s going to have the biggest impact on a business? or let’s say you have a business goal: I want to own this topic by the end of the year. Market Muse can tell you what articles need to be built or what articles need to be updated. And then we follow through with the content briefs.
So we actually build the content briefs at scale. We have teams that get thousands of content groups and they basically provide an outline or a single source of truth for the writer. So when the writer produces, they can focus on being creative. They can focus on being the expert that you’re paying them for, not doing stuff that they don’t want to do.
Pam Didner: Like SEO, like doing a lot of research and finding the title. What works whatnot.
Jeff Coyle: Yeah, exactly. So then when they give you that draft, it’s highly likely, because you built the brief, you ordered the content with the brief, they’re going to give you something that’s going to have fewer feedback cycles. And it’s likely going to meet the guardrails, the guidelines that you’ve set. That’s what a brief does. It creates a point of reference.
We automated all of those processes, so you can say, “okay, the best article I can write today, Pam, is on how do you make the best product marketing materials for events. I just made that up. Maybe that’s not a good idea that it is, and we will build the outline and you can write it and you know that it’s going to be successful.
Pam Didner: Awesome. Love it. Love it. And one of the key product features is to use artificial intelligence, AI, to build content planning, optimized content. Like you said, such as your blog, white paper, your thought leadership pieces.
So, can you elaborate a little bit how Market Muse uses AI and how does that help, uh, the content marketers?
Jeff Coyle: Sure thing. So we built our own natural language processing platform effectively. One of the components is in a branch of AI called topic modelling, in natural language processing. And we are a machine learning think tank. We have our own data science lab in Montreal called the M4 Lab. Those are spending nearly all of their time on back in engineering on this platform that can assess effectively what it means to be an expert on a topic.
We tell the story that if you were to cover this topic comprehensively and truly be an expert, how would you do it on one page? How would you do it in one cluster of content? How would you do it across the entire site? And then we’re able to overlay that information into common workflows. So we can take that model and map it to your page and say, “Hey Pam, here’s a, here’s a page-level audit of thing you wrote. You had a few blind spots.”
Pam Didner: Okay, we’ll fix it.
Jeff Coyle: Or we can go up to the competitive dynamic and go, “Oh, across the 20 competitors who you’re researching, here’s the stuff they did and here’s the stuff that Market Muse says is very relevant and no one’s done. And there’s your area of differentiation. So we don’t just tell you how to do everything like everybody else. We tell you how to differentiate yourself from them. That’s very special. That’s unique to Market Muse.
At the site level or the network level if you’ve got a lot of sites (we work with a lot of large publishers) we give you basically an on-demand content inventory and audit and all the value that goes along with that. And it’s all personalized. So we’re the only in-market application that can tell you how hard it will be for you to rank for a topic. Not just how hard it is, cause that’s not valuable to me.
If I have a brand new website and I go write the best review of the new iPhone ever. I have no chance of getting any traffic from the market, but I’ve seen that – I go write that article and I win. Market Muse answers the question of why behind that dynamic. So Pam, if you go write sales enablement tactics for the first year BDRs, it’s going to rank.
Pam Didner: Your platform or your tools can actually identify some of the gaps.
Jeff Coyle: Areas of opportunity as well.
Pam Didner: That’s what I call it a niche, but you say “areas for opportunities.” I like that term much better.
Jeff Coyle: And the gaps and the gaps, too. And the gap could be a variant. So you covered this with a great long-form thought leadership piece, but you missed a few details or it could be vice versa. You wrote a, something that answered one question and you never spoke to the big picture.
So we tell a story of what that true user intent profile is and everywhere you’ve done it everywhere you haven’t. Where are you the most important pages to lean into for links, uh, to point new pages to, what are your, um, your risk areas? So we, this is a big area for us. It’s a risk.
You have a page out there it’s doing well, you don’t even know why, and you can need to support it. You might have a page that’s out there–I call it the arm, a hand without an arm it’s just out there. It just gets a bunch of traffic and you have never supported, you never built the foundation. You’ve gotta build that foundation.
Pam Didner: Yeah to amplify and expand that. So it sounds like that you are using artificial intelligence to do kind of like a very holistic assessment, identify the areas for improvements, identify the gaps that you need to do you look at a site level at a page level, or even at specific content. Is that correct?
Jeff Coyle: Any level of specificity, even networks we can say like we tell our publisher – if you’re going to write an article about this TV show, right? And you’re going to write about the actors on the TV show. You put that over here, and if you’re going to write about ways to stream it, you know, put that over on the other side, right? So we can even get into that level of detail. We set guidelines for content. We set guardrails for intent. We build a topic model. The next step is obviously to start to build out drafts.
Pam Didner: It’s a virtual editorial assistant, in a way.
Jeff Coyle: Absolutely. Giving you the inspiration so that you know, if I focus on this and really put my energy towards it and I write it, great production value, high quality, a wonderful narrative that meets these guidelines, it’s going to be so successful. And now we’re building out drafts that give you inspiration form of completely original drafts.
Pam Didner: Yes. Yes. With that being said, you kind of touched the topics I want to ask you next. Obviously, the latest buzz in the field of AI is GPT3. And I read an article or a blog post you guys wrote about expose GPT, in terms of what it can or can’t do.
And the GPT3 is this very powerful language model that uses machine learning to generate human-like text. Are you leveraging GPT3 at all? Or you are doing stuff on your own?
Jeff Coyle: All of our natural language processing and natural language generation is all proprietary, built by us. All the models built by us. GPT3 is one of the most amazing innovations in artificial intelligence, of all time!
Pam Didner: Yeah. Up to date. Yeah. There may be something that’s coming in the next year, but as of now, yes, it’s quite amazing.
Jeff Coyle: It’s, it’s amazing. And it, the amazing scale that it works at in the applications out of it are significant.
Pam Didner: I 100% agree with you. I address that a little bit at my general session at Content Marketing World 2020. And I specifically talk about content pretty much can be written for you if you use artificial intelligence.
Jeff Coyle: Yeah. And so that’s the thing. It can be written. But is it telling the story you want it to tell?
Pam Didner: Yes. Does that meet your objective? Does that resonate with your audience? You know, does that bring a hook? Yes. I hear you.
Jeff Coyle: It’s a general-purpose language model. And so it’s not a solution. It is an API. It requires, you know, now in its current form, a lot of programming skills. It lacks structure, it degrades over time. It is, you know, not considering your workflows. It suffers from degradation.
GPT2 suffered from degradation, a severe rate, right? It was, it was “Pam, Pam, Pam, Pam” and makes some really kooky stuff. GPT3 is shockingly better. And there you can begin to work with it, but it is still subject to that degradation in long-form content and short-form content.
Pam Didner: I do agree with you. I think having a machine generate say a draft, just like you said, is a draft that humans still needs to take over and to do conduct the quality checks. Still need to edit as you see fit. You still need to make sure that you read it, that aligned with your objectives and what you want to accomplish.
Jeff Coyle: Yeah, exactly. And that’s where, so the solution Market Muse First Draft versus of GPT3. First of all, you know, like you were, you mentioned, I believe that you know, the Guardian post. That is an output of eight generations and it’s curated by editors.
Pam Didner: They provided, they brief GPT3, they provide specific instruction, you know what they want GPT3s to write at the same time, they also provided the opening paragraph. So they provide the kind of like in-depth instruction for the machine. So the machine can process and determine how they want to write it.
Jeff Coyle: Yeah, what I’ll tell you, though, is if you saw all eight of those generations back to back, you wouldn’t have been all that impressed. The outcome’s amazing. It’s curation, right? You had to do work on it. So the human brain now has never experienced the outcome of generation at this level. Washington Post–these are the problems that content marketing people need to solve, or like Washington Post solved with Helio.
Pam Didner: Right. I was, and they did that back in 2016 and 2017.
Jeff Coyle: Right. But the thing about the problems, right? The problem was: I can’t cover the election.
Pam Didner: On the local level, right?
Jeff Coyle: Right. I cover 8%, but now Heliograph lets us solve all of it. That’s solving a business problem. You’ve got platforms out there that can do data-driven template content. It’s not this right. But GPT3 isn’t doing either. It’s not producing content that is meeting specifications. It’s not producing content that is connecting with your goals as a content marketer. A template might even do a better job than that, but you’re not hitting targets for tone. It’s not training it, training on your content either. So Market Muse First Draft’s value is we can train it on your blog. It can write like Pam.
Pam Didner: Yeah, understood.
Jeff Coyle: It can write on spec.
Pam Didner: There’s another question I would like to ask you. Can you tell us what features your teams are working on in the next six to eight months that you see is additional features that you want to add? I understand that’s a product roadmap tend to be confidential and, uh, but if there’s anything that you can share, that would be fantastic.
Jeff Coyle: Yeah. So the big things that we will be, first of all, Marketing Muse First Draft, it’s in private alpha right now. It launches into public beta, most likely, Q4 it’s coming out. That’s a big splash for us. Also for the first time Market Muse is putting new offerings for at a couple of lower price points.
Jeff Coyle: I know. I know.
Pam Didner: Jeff I remember I was talking to you guys last year and you will tell me like, “Oh, we focus on enterprises and our average price is this.” And I was like “I thank you so much. It’s very nice talking to you” and that quiet walk away. (laughs) I love your product. I still went and did a demo for me. I was like, “Oh my God, this is amazing.”
Jeff Coyle: Yeah, it’d be, so that’s going to be a big deal. So we have those things. That’s all happening in October. So many people who, you know, we have some demand and we didn’t have a place for a solopreneur, or a small team, even, you know, a startup or a small agency. Now easy decisions are there for you. We’re also upgrading a lot of our current clients based on that. That’s exciting. So those are huge things.
Pam Didner: What are two pricing points? Can you say that again? 79 and …
Jeff Coyle: $79 a month, $179 and $499 a month are our self-service agreements and then we have our premium product line, which will remain at those price points.
One of the exciting things, there are some options for going annual, which makes us even more attractive. And so some of those features that I’ll mention: we’re going to have the most accurate measure for potential upside with content and topics.
Pam Didner: So you can actually forecast how the content will perform?
Jeff Coyle: Yeah, we’re going to have to get where to basically say you are here. Here are the estimated traffic and value that this page has. And the value is if you had to buy that traffic. We’re going to have some customization for getting even more specific into the actual return on investment, right in the application.
And you’re able to see what Market Muse thinks your potential upside is right now. And then what’s your upside is if you were the best. So we don’t think you maybe might not think you can be the best, yet. But this is where you could get with an easy, quick win. So we’re basically trying to get into predicting your likelihood of success because then that tells the story of when you using it you can justify the expense very easily.
Pam Didner: Yeah, exactly. So you actually quantify that ROI?
Jeff Coyle: Direct quantification of value. We already do it for our clients with things like QBRs–quarterly business reviews–through custom reporting. But we need that in the self-service offering, right? It has to be automated. And so we’ve really focused on that.
You know, it’s so fun to go in and look at your content and go, “Oh, wow. Like, if I had to go buy the same traffic that this page generated, it would cost me 10 grand a month.” Holy cow, right? So all of that is those are our big, our big moves. We’re really stretching the market down market as well as releasing innovation for the generation side.
Pam Didner: Very, very good. Jeff is so wonderful, so wonderful to have you on my podcast and the show. You know, I hope to see you soon, hopefully at Content Marketing World or like in a physical event.
Jeff Coyle: Yeah, I did my Content Marketing World presentation. I think we’re going to be one and two in the best. I’m the second-best one, you obviously the best one. (Pam laughs) We’ll see how that goes. Yeah, Content Marketing World going to be amazing. The lineup, I’m so excited to just veg out and watch all the recordings, frankly.
Pam Didner: Very, very good. I’m going to see you at the virtual event.
Jeff Coyle: Thank you so much, Dan. We really appreciate it.
Pam Didner: I hope you enjoy the podcast. I’d appreciate that you leave a five-star review on Apple Podcasts or my website, pamdidner.com. If you think this is useful, please share with your colleagues and friends. I also love to hear from my listeners. So join my Facebook community, Build Marketing Skills to Get Ahead.
I’d answer any marketing questions you may have. I mean, any marketing questions you ask, I answer. All right? See you there. Have a great week.
Enjoy the podcast? Subscribe to the show on your favorite podcast platform or leave a 5-star review and subscribe to Apple Podcasts.
If you prefer watching a video, I also have a YouTube Channel, check it out and subscribe.