[02:31] So for the folks who might not be familiar with you or your company, let’s start with the easy stuff. Who are you and what do you do?
- Yeah, so we’re MarketMuse. We use AI to help content marketers, writers, content creators, optimize their content through AI and make better planning decisions around the content. So we help you figure out if you want to cover a topic comprehensively in your web content using text. So when we say content, we mean long-form text, typically. How do you do that? What articles should you write? And how should you write those articles to answer each topic or each question comprehensively so that when people search for things related to your business, they find you, and they educate themselves, and you become a thought leader or an authority on that topic.
[03:28] And when did you launch the company?
- I started working on it as a science project in 2014, 2015. At the end of 2015, my co-founder, Jeff Coyle, joined me. That’s when we became an actual commercial enterprise. So from that point on, it’s been about four and a half years.
[03:50] Are you guys funded or bootstrapped?
- We raised money. We’ve raised over a number of different rounds and both equity and debt, some hybrid instruments, but we’ve raised about 10 million total.
[04:05] So, your ideal customer is who?
- We basically sell to three different types of customers. The most prominent are content marketers, content strategists, director of content, or marketing manager who does content in his or her day job. That’s one base. Another type of customer are SEOs. So there are people whose job it is to improve just the lead generation and the search performance of a site and we sell to those both directly and through channels as agency partnerships. And the third type is a managing editor or publisher, so we work with very large and even very small niche focus publishers to build their authority and hub guide and editorial team.
[04:55] Okay. And going back to when this was a science project for you, what made you pick this?
- Yes, you know, a little bit of a meandering story, but I came from kind of dual backgrounds: business and technology. I worked in a management consulting firm, where we built a data and analytics practice back quite some time ago, I guess 13 years ago. And I just fell in love with data, analytics, information, you know—the ability to take a bunch of data and get insights from it.
Then I went to a venture fund, OpenView Venture Partners, where I was the first associate looking at big data, and machine learning, and marketing automation on this side. And at the time, this is 2011, there were a lot of database companies. So there were a lot of ways you could store large amounts of data. But I was really interested in building a solution that solves an actual business problem or somehow makes our lives and our professional lives better.
And so then, well, I worked at another startup for two years where I kind of had time to flesh that thesis on a bit more. And then I actually went and started—or tried to start—a healthcare analytics company because I wanted to help people. But some of my mentors sat me down and said, “You know, healthcare is really hard. Don’t do that for your first one. Regulation, HIPAA, the sales model’s challenging, you name it—it’s gonna be rough. Why don’t you do something different first and then come back?” And I said, “Okay, well, what’s another thing that would help society?” Well content, because content is marketing. So, you know, it generates leads and drives revenue. So it’s a little bit easier from the commercialization aspect than healthcare. But on the other side is education. Content is, by definition, it’s information value. You’re learning things as you progress. And I thought that was really cool. And then I started focusing on that, plus it was a pain point that I had seen at both the startup and at OpenView. And so I just started digging into that, but I wasn’t quite sure exactly how to even articulate what problem I was working on.
And then I also wanted to build an AI technology just mostly by myself with one advisor and a couple of part-time, and interns, and so on. So that took a while. I had to relearn how to code, you know, learn a lot of different languages, so on so forth. And then it took after two years of the kind of science project phase. There was enough of it together where we could actually make a buck.
[07:31] So for someone like me, who has written plenty of content, both for myself and I had an agency at one point where we like created a lot of content for them, I don’t have a solid understanding of how AI is going to make content creation easier. So for the lay folks in the audience, self included, can you just walk us through what you’ve built actually makes it easier for a content producer, easier and faster, I assume, for content producers to make content?
- Yeah, absolutely. We automate or semi-automate the research involved in building a content strategy. And when I say content strategy, I actually mean two things. One is what articles should I write on my site that I don’t have today? And what existing articles should I improve or flesh out more? Or repurpose or optimize? That’s one piece of this strategy.
The other is given a particular goal, I want to answer question x. How do I do that in this article, and then what article do I link this article to? And how do I kind of connect them up? All of that falls under the umbrella of content planning essentially. Even if you’re just writing one article, you want to plan it. And so what we do is we do the part that machines are really good at, which is reading a bunch of content very quickly.
So humans have great synthesis abilities, you know, abilities to have a point of view, ability to tell a story. Humans are great at that. Humans are not great at reading 100,000 articles in 90 seconds. But computers are really good at that. Right? So that’s sort of the two-sided nature of this, and it means very kind of practical when you, if you’re going to write an article, how do you write it? You’re going to sit down, you know, a blank sheet of paper, you’re going to use whatever domain expertise you have in your head, you might interview people, you might Google around, and read five to 10 articles just come up on Google on the topic. That third piece of it, we accelerate that. So instead of five to 10 articles, we’ll download a whole bunch of them until we have statistical significance for that particular query. And then we use a branch of science called “topic modeling,” and we build what we call a “knowledge graph,” which is basically just to distill down the essence of those 100,000 articles. And we give that to the user in an outline.
And then the user can basically see a little bit like a Carfax report for your article. So you can see, you know, based on what you’ve written here, things you should cover and have covered here, things you should cover and did not cover here, things that are optional. But they could help you differentiate, and then you make better decisions on your own. It’s a little different from a self-driving car because there’s only one way to drive a car correctly. There’s only one way to get from point A to point B in an optimal fashion. Content is not like that. There are thousands of ways to write about a topic. So the research gives you kind of a base and a leg up and an independent kind of verification. But it’s still up to the writer to flesh it out, tell the story, you know, inject your expertise.
[10:59] So once the software has done its job, the writer is then presented with, “Here’s what you should write about. Here’s the outline for what you should write about.” But they still have to write it, correct?
- That’s right. Absolutely right. You know, machines, they’re already starting to get to a point where they can generate pieces of the content, but machines generate facts for the most part. There are of course, you know, ever since we had the AlphaGo beat the gold player years ago, machines can also do some more advanced things and they’re getting better and better. But as it relates to content, basically you get the building blocks, and then as a human, you assemble it using your knowledge. And I guess one easy way to think about it is, you know, I’m not a doctor. I might have a perspective on some medical condition, but if you go to an average doctor, they’ll have a more informed perspective. And if you go to the best doctor in the world on that condition, they’ll have the most informed perspective.
[12:00] It makes a lot of sense because I think a huge challenge for a lot of people is “What should I write about? What should the article cover?” If someone gives me a basically, “Hey, Trent, fill in the blanks with your knowledge,” that’s a whole lot easier than just sitting down in WordPress and going, “Hmmm, what should I write about today?” So I can see how that would be hugely valuable.
- Yeah, absolutely. It doesn’t fully eliminate the FOMO. But it helps a lot with, you know, here are 50 things, where we show 50 things. I mean, in reality, we could show 5,000 things, but the human brain can’t do that kind of analysis. So we should have the top 50 things, here are the top 50 things for this particular question or topic that you might want to consider. And here’s sort of what the landscape looks like, and that’s enough for the human to really dig in.
We can also do a little bit more. So if you have a completely blank page, we can suggest how you would write an article net new. The other use case that’s really interesting, which we use a lot, is just optimizing something that’s existing. So you take the best article you have on a topic, run it through the machine. And if it tells you a couple of things that are valid points that you didn’t occur to or you may have forgotten about, but those are things that make sense to you, then the machines did a good job at predicting how an expert would write about that topic.
[13:29] So how many customers do you have now? And what’s your year-over-year growth rate then?
- Yeah, you know, great question. A little bit hard to answer at the moment because we’ve just rolled out a completely new self-service model, but I’ll tell you what I can tell you. So we have over 200 companies that are paying customers. Some of them are quite large, you know, large publishers, and so on. Some of them are small businesses or small–medium businesses. Some of them are venture-backed or PE backed where the VC gives you 10, 20 million, says, “Hey, you know, go grab all the market share on this particular topic” and as they need to do that and content as well.
Some are agencies where they pay varying amounts based on how all that is structured. And that can get a bit complex, but we have over 200 of those. Where it gets a little more murky is, two things happened at the same time for us. One is in February, we launched a self-service product called MarketMuse Pro. Before, you had to talk to sales and go through a whole proposal process and so on, so we were not able to really effectively service smaller companies. We could really only effectively service larger companies.
[14:47] Your price point was quite high on the original version, right?
- That’s right. Yeah, the price point was fairly inaccurately priced for the particular segment of companies, but it left out a lot of other segments. And that was never our intent. It was just sort of the way we built the platform kind of got us there. And then we started servicing startups and SMBs, then we got to that price point. So we were labeled more of an enterprise platform. And then we always had an intent to go back, and we’re finally able to go back with a product that is $499 a month, but right now it’s free for three months. And so we launched that product and it’s a month-to-month subscription, so you don’t really risk a whole lot.
And then COVID happened, and in order to kind of help the community as well, we just made it free for three months. It wasn’t free originally. There’s a trial experience where you can run a couple of queries in the system, and then you have to put in a credit card, but then we made that credit card step also free for the next three months. And it was originally meant as a way to really just add value to people who, you know, frankly, whose businesses have suffered, who had a harder time doing lead gen, and all that. But it’s turned out to be a really effective way to keep our ear on the ground and better hear our customers’ requests, and concerns, and challenges with the platform, and opportunities, and so on.
And so there are over 5,000 people signed up in various states of that trial and pro kind of onboarding flow. And we’re in May right now, so we only had three months of data. So we’re just trying to figure out kind of how that will eventually shake out, but we’re just glad that it’s gotten into a lot of people’s hands who otherwise would have not been prevented from using the market.
[15:51] So I want to ask you how you attract customers. And there’s that first tranche of customers, the 200 that you mentioned. I’d like to know, at least at a high level, how you attracted those customers. And then you’ve got these 4,000 or 5,000 people who’ve just signed up. How did that happen? So let’s do the first ones first. Was it a lot of content marketing that had those people find you? Or was it outreach? What did that look like?
- The first couple of users are pure magic. And then from a user 5 to 15 is a little bit less magic but still fairly magical. Then you get to 50, and then it becomes a bit more of a repeatable process.
That first chunk was challenging because, well, one, we ourselves were describing a concept that there wasn’t enough terminology at the time. So for example, topical authority. We coined that, if I remember correctly. We coined that way back in 2016, or something like that. We were trying to explain to marketers why they should look at topics and not just keywords. Things like that were challenging for even us to articulate.
The other challenges were very practical. Because we were semi-bootstrapped or mostly bootstrapped, we didn’t have enough budget to bring on a full-time writer. And I looked around at some writing networks and so on, but, at the end of the day, every time I had $1, I’d rather put it into product engineering or sales and so on. So to answer your question, the very first thing we did, some of it were dumb luck, and some of it were, well, most of it was dumb luck, let’s be honest.
One of the things that we did was influencer marketing. So we partnered with two influencers, and we just met them by kind of working on this problem for years at that point. We just kind of met these two influencers. One was Neil Patel, and one was Brian Dean. And they’re both kind of SEO content marketing folks. And we did data journalism. So for Neil, we took 400 articles from his site, and built our content outlines, and showed him how to make those 400 better, or just gave him that data for free. And his team made the changes, and they saw a doubling or tripling in traffic within a year. So it was a huge return for Neil. And so we were able to publish a case study and then Neil referred to us for a period of time and so on. I think that helped.
And the other one is Brian Dean. We did a more comprehensive analysis of something over a million URLs. And statistically, he was able to show that the thing that we were doing, in terms of improving the quality of content, was the one of the top three most impactful things it could do to improve your ranking in Google, obviously, if you implement it correctly. And so we were able to show that our content scoring methodology using this outline was predictive. And then that brought in a lot of interest. So the data journalism piece was really important. And we just got that just through luck by just meeting these folks, and kind of networking, and friends of friends connected us, and found our way there.
The other thing that happened that was also kind of dumb luck, was, in the SEO space, there’s a company called Moz. And the founder of Moz has since left but he is this guy, Rand Fishkin. And I just reached out to Rand because I reach out to folks all the time. I still do every day as a founder. And I just thought he’s just a really cool guy and actually a kind of informal mentor of mine, Dug Song, who built a big company called Duo Security in Michigan, in Ann Arbor, where I’m from, Dug was like, “Hey, you should talk to Rand sometime. He’s smart.” And I’m like, “Oh, yeah.” So I reached out to Rand, and Rand was like, “Oh my gosh, I love this idea. Let’s try to do something.” And we weren’t able to do something with Moz because of other things that came in the way. But Rand himself just got the idea. He had already been thinking about this idea for some time. And the fact that there was a platform that could actually do this, he just started talking about this topical authority, and content clusters, and a lot of these concepts, and then also helped drive just top of the funnel kind of awareness and visibility.
And then we came into that market with a product where you could actually verify that it works because given any site, and any article, you can just plug it in. And like I said, if the machine tells you a couple things that are good ideas, you’re like, “Well, yeah, that’s pretty cool. That works.” So those were the things that helped us.
And then at some point, we were actually able to hire a content writer, which really helped because we had a lot of very embarrassing investor presentations. We were embarrassed a couple of times because early-stage investors said, “If you believe in content so much, why don’t you have anything on your blog? You just have, like, three articles.” We were like, “We love it, but it’s just so expensive.” It’s really hard writing content. And every time, I mean, this is probably something I could have done better. But every time I would have a minute, I would just either work on the product or I tried to sell it directly. And so . . .
[22:28] Oh, I know the feeling. I am a good writer, but I don’t write anything for Flusters because I’d rather, again, same thing, I’d rather do outreach, or direct sales, or find an influencer or something, so I feel your pain.
- It’s a tough problem. I used to work at the writing center at the community college that I went to back when I was like 15. And I watched [unintelligible], and I love writing, but it’s that kind of value equation. But on the other hand, even your founder has to kind of get that founder magic into the content. So if I were to go back and do it again, I’d probably just go ahead and pull the trigger on that outsource writing network as somebody for five hours a week or whatever. Write an article, I add a little fluff—you know, not fluff, I’m sorry—a little magic to it. It could be fluff, could be incredible insights, and just kind of keep that engine going.
But tactically we got so much inbound interest from the influencer channel that we just had the three articles. And then we’ve got funding; we hired a writer. And that writer, a content strategist actually, basically just went out and over the years he’s knocked out over 200 articles about what we do. They’re highly technically sophisticated from an SEO perspective. There are a lot of technical terms in there and even I tried to read every one that has come out, but I learn something every time I read it. So it’s turned into kind of like a publicly facing knowledge base that helps us as a company, our employees, but also helps just the public at large learn about these advanced concepts.
[24:16] So I’m assuming then you’re now in the cadence of producing content on your own blog on a regular basis.
- So we raised our first money in Q1 of 2018. And we use that money to hire this writer and an editor. Since then, we’ve been publishing on a regular basis and we don’t have to have awkward conversations on that anymore.
[24:39] Okay. So in addition to your content marketing and your influencer marketing, is there any other strategies or tactics that have worked very well for you in terms of lead generation and customer acquisition?
- We use partnerships extensively, or at least we try to. There isn’t any sort of main KPI against it. It’s just sort of the right thing to do is to figure out who else is in our space and in what ways can we partner, and there are different types of partnerships. But as part of that, we ended up basically doing a lot of cold marketing. So we have a lot of webinars. We have at least one webinar a month. Some months we have two webinars. My co-founder, Jeff Coyle, has done an incredible job there and the team supporting him has been great.
Recently, we launched a Slack channel, so we have the content strategy collective. We have a couple of hundred people in there and that just launched a few weeks ago. And already a couple hundred people in there talking about it, you know, very advanced content strategy concepts. We’ve tried a number of different things, email nurturing campaigns, kind of this standard meat and potatoes stuff, you know, education flows. We’re launching an academy, where people can not only train up but also certify themselves through a learning management system platform. I’m probably forgetting a couple of things but . . .
[26:11] But let’s dive down into the partnerships a little bit more and get down in the weeds a bit. Who else is in our space? And how can we partner with them? First of all, let’s start with an example. Give us an example of one of your most successful partnerships.
- You know, great question. I would say there are two types of partnerships. I mean, there are more, but the two main types that we’ve found are one is just a very light-touch referral partnership. So for example, we have an agency. It’s actually an individual, although he does business as an agency, and he just personally, years ago, just really started liking what we do. And it didn’t make sense to do any kind of formal commercial agreement, but he’ll talk about us as kind of a new cutting edge solution to his customer base and people he needs for conferences. And then when he finds one with an actual content project where it’s big enough for our sales team to want it again, he’ll send that over and make that referral. And then we’ll pay 10% commission on year one, and 5% on year two. The reason we do that is because you want the customer to be successful so that they renew. And once they’re past the first renewal, the risk that it’s a bad fit or that we did provide service (and) all that falls, we need that referral partner to help support that relationship through the first renewal so if something does go wrong or something goes right, you know, that the person who made the introduction is still involved in tier two. So we pay that.
And the key to that structure has been very easy. It doesn’t cost anything. They just send a little referral form on DocuSign. It tells you, “Hey, all you have to do is send over a qualified lead. And a qualified lead is somebody that will close within six months.” They have a legitimate project. So you’re not throwing over random things; you’re throwing over something that has a reasonable chance to close. And then if it closes, you know, and that’s been really easy, because the entry point is almost zero. They just have to click this sign and nobody gives anything away. And then you see the engagement in that channel.
The other types of partnerships we have are more in-depth. I guess the most successful would be an agency partnership, where you have 10, 20, 40, 50 people who are basically using our software to deliver their service. So they’re doing content strategies. They’ve been building these strategies by hand. Now they use the content planning side of MarketMuse to build the strategies. They use content optimization et cetera with their in-house writers and freelance writers. So they’re using it to deliver better service to their existing customer base. And they’re also using MarketMuse as a way to pitch new prospects or new clients and also to upsell existing clients.
So if you have 20 clients you represent, and five of them are really strong in content, you can basically have MarketMuse come in and say, “Hey, what additional services could we offer these five to make them even stronger?” And we can add in a second layer of that, and we’re really good at finding those with our upsells, with our software, and with our sales team. So that then becomes a revenue generator for the agency.
Over time, we want to have even like a proposal generator, where you basically just pointed out a site and it builds a kind of a proposal you can pitch in. We’re not quite there yet. But the goal is how do we make agencies, you know . . . well, first of all, content is expensive. That’s the reason also founders have a hard time writing it. It’s just high-value and high-cost. So how can we lower the cost? And then secondly, how can we help you gain revenue?
[30:14] Are you guys in talks with HubSpot about a partnership yet?
- Yeah, HubSpot is interesting. It’s funny because I actually was just emailing Brad Coffey from HubSpot this morning at 7 a.m. Not that we do that a lot, but the timing is funny. So, HubSpot is unique. One of the reasons that I pick content is because at the time I was living in Boston, HubSpot was growing rapidly. They evangelized inbound marketing as a term, content marketing. There were another headwind or you know, kind of macro, or reason, or factor for that industry. But we haven’t ever actually partnered with them.
It’s almost like a meeting of the minds when we talk to them. They’re really proud, and they’ve done a good job of their content. And we’re really proud of what we’ve built. And it’s just a complex kind of connection. But they’re certainly open to it. So HubSpot is building a platform business, and we would be a component in their platform. We just don’t have that product for them yet.
[31:29] So, in terms of cold email outreach, have you ever experimented with that? Like, “Hey, let’s identify customer segments. Let’s build a list of them, and let’s just send them messages into the inbox.”
- You know, for us, I have definitely done outbound prospecting now for companies. So I believe in outbound prospecting, I do it every day. I do it for fun. It is fun. After you get past the first couple months of, you know, kind of hesitation as I do because I’m a little introverted. You know, it’s extremely value-added. But programmatically, we have not done that because, well, our business doesn’t require it. I’ll put it this way. We haven’t had to do outbound because we sell to content folks, and they’re all inbound-oriented and they prefer inbound.
That said, I absolutely do. Sometimes we do cold outreach through LinkedIn, Sales Navigator, ZoomInfo, that kind of thing. So we did cold-prospecting. I get cold introductions from our investor network, or our community, or I asked for cold introductions. I’m doing a couple today. But programmatically, we haven’t done an email blast. I’ll give you an example. We don’t even have outbound BDRs. So every time we hire an outbound BDR, a few months later, we end up moving them to something that we see as more higher value-added. So for this particular industry at least, we haven’t seen a lot of value, or rather we just haven’t really doubled down.
The other other thing that we haven’t done is paid marketing. We just haven’t really done paid. We did a couple of quick experiments here and there just to kind of mess around with it. But for this business, we haven’t done paid. What we do that I forgot to mention are conferences. So every year we go to six to eight conferences. And this year, they’re virtual.
[33:42] Yeah, say, “Not this year.”
- Yeah, right. But you know, we just went to a conference, a Serious Decision Summit a few weeks ago. And we had 180 people stop by our virtual booth. And conferences, you know, there’s just a type of interaction. I mean, we’re all feeling it right now in this new kind of fully remote world. There are types of conversations that are better done in either face to face, or kind of in a longer format, or in a deeper interaction. And conferences provide that kind of higher quality, higher depth, high-touch conversation, which is also valuable.
[34:21] Yeah, the email question I was asking, I was coming at it because you mentioned you form partnerships with agencies. So wouldn’t it be great to have, you know, another hundred agencies or another ten agencies like the agencies that you have? And I thought, well, if you can’t get HubSpot to refer you into the agencies, undoubtedly, at some point, they might find your content and find their way inbound. But why not just build a list of agencies and be doing cold email to them?
- Well, you know, Trent I forgot to mention a part of my early strategy. This is not for lead gen, but this is for a product validation of the idea. The very first thing I did before I wrote any code was, I actually did go to the HubSpot partner, HubSpot agency portal, or whatever it’s called. And they have a couple thousand agencies on there. And I just downloaded a subset of that that did content, and I just called them. And I just called them and pitched them. So, you know, “My name is Aki. I’m working on this thing to optimize your content. Would you be interested in a demo?” And about 25% of the agency said yes. And then we got to the point where one agency partner was basically on the first call. He was, like, asking for pricing. He’s like, “I want to negotiate a volume discount. You’re really small. We should get a good discount. We’ll develop it together.” So that was a clear signal that there’s demand, and then I stopped all that. And then I started writing code. So that was the very first thing I did.
[35:57] Okay. All right. Let’s move on. So a lot of SaaS companies, mine included, look at pricing. Do we have the right pricing? How should we change our pricing? How do we improve our pricing? Because there’s endless discussions around SaaS pricing in every corner of the Internet you could ever go to that have anything to do with SaaS. So coming up with your pricing plan, how did that happen? Are there any takeaways there potentially for other SaaS founders with respect to the pricing model?
- Yeah, early days pricing was very, very much magic. And it was very random, I would say. So random inputs. So for example, the very first user that I had, he just used it for a while. And I had actually, the HubSpot founders, I think, had a kind of thing about, you know, when you’re doing early product validation, don’t give away your product for free. You know, charge 10 bucks for it because you’ll get better feedback on 10 bucks a month than on free. So you know, adopting that principle, I charge 20 bucks because, you know, inflation or whatever. So the first cut using the system was paying 20 bucks. And then at one point, after a few months, I asked him, like, “Would you pay 50 bucks?” He’s like, “Sure, put it on my corporate card.” And I went 50 bucks. And then a few months later, “Would you do 100 bucks?” He’s like, “Yeah, okay, I’m getting that much value. Fine.”
And then I went to a meeting and I pitched a series B-funded company, and they were quite interested. So on the way back from the meeting, I stopped by at Starbucks and changed the price into 500 bucks so that they can negotiate me down to $350. That was it. That was the early pricing model. That was the input, and then the market kind of support kind of ended there. Or I should say the market kind of rallied around that price point. So that kind of locked in that. And we tried to do more kind of quantitative data-driven ways. But at the end of the day, we’re like, “Look, it works. It’s not broken. Let’s just leave it and revisit it.”
Then years later, in 2018, we did an actual pricing exercise where we did, essentially, like a conjoint analysis, but we basically broke down all the value pieces. Because at the end of 2018, we launched a new platform, which is the whole content planning, content inventory platform. So we have just a single-page app called Content Analyzer that was 500 bucks a month, and we moved to this whole platform.
So the platform had all different types of value. So we did that basic analysis and the product team actually did an offside. And my co-founder, Jeff, who’s brilliant at pricing and packaging, he led this exercise where we basically took a little post-it notes, and he would say, “Okay, here are the five factors. Now imagine you have 20 of these, and two of these units, and 10 of those units. How much would you pay for that? Okay, imagine you have 100 of these units, nothing of these units, nothing of that, how much would you pay for that?” And we basically got a bunch of data and tried to find a middle ground.
And that exercise helped us, and we found something. And then we realized that we actually had a huge blind spot, because the pricing that came out, as I mentioned, was more like enterprise or mid-market pricing. And we left the SMBs behind when we moved away from $500 a month. So then we were like, “Well, actually, we do need to have like a $500 month-ish kind of thing. And maybe even less than $500 a month. And so in early 2020, we launched that product. And that product was . . . It was challenging to build when we built it. I mean, it took a couple months, but by the time we got to launching and designing, it wasn’t complete rocket science because we had this broad platform for two years. And we could say, “Okay, what are the main pieces of this that are easy for people to start with? Bam, bam, bam, let’s wire it together. Let’s add some education. Now you got a self-service.” And we launched it.
What I’ve learned from that, even just over the last few months, is giving people a really easy way to get in, you know, whether it’s free for a few months, or 10 bucks, or whatever. Giving them a really easy way to get in and start engaging is worth so much. And it’s kind of ironic that I say this because the fund I used to work at OpenView, I think they’ve pioneered or focused on the topic of product-led growth, which is what Calendly has, and all these folks. So we always knew we have to have it, it just took a while to build it. But now that we built it, we do see the benefit of that because anyone can try out MarketMuse. They can kind of self-qualify, educate themselves, get value, and it doesn’t really cost them anything. And if their engagement, you know, soars then we know that we could show you some more advanced things if you like the first set of things, we’ll show you the next set of things are going to be really powerful.
[41:19] So a couple of things we’re going to talk about in the future. I want to ask you about churn. I want to ask you about your customer onboarding. I want to ask you about revenue expansion. And then we’re also going to talk about some people in the process, so we’re going to cover some more good stuff yet before we finish up. Has churn ever been an issue for you guys? Like what’s the highest your churn has ever been? And then what is it now? And where I’m going with this is did your changes to pricing have a positive effect on reducing churn? Or was it other stuff?
- Yeah, that is probably the hardest part of— well, there’d been a lot of hard parts quite honestly, and it’s been a lot of hard work and a lot of determination by a lot of people to get us here. But yeah, churn has been absolutely an issue. And you know, I have that many VC backgrounds. I spend a lot of time talking to VCs. And VCs, of course, arguably focus more on churn than on new revenue because part of it is getting a customer, but it’s even more important what happens a year or two down the road. Will they stay with you? Do they expand? And that is probably the most accurate or predictive signal of a big software success is if you have great retention and expansion.
So we’ve philosophically focused on it, and we had years where we had a 10% gross churn, which is great. And we had a net negative churn, which is great. What happened though, which I think—we’ve kind of beaten ourselves up about it—but I think now that we shouldn’t beat ourselves up about it anymore is when you have an inbound-led funnel, you end up selling to a lot of different types of companies. The good thing about outbound is you can focus exactly on your ideal customer profile, your ICP, and just hit those. Inbound, you know, that data is very unclear. Now, that can teach you a lot of things. But you can also bring on a lot of customers where later you learn, “Oh, if this factor is x, then they’re a great customer. If that factor is less than x, then they’re horrible customers.” You only learn that stuff later. And it’s like, you look at yourself in the mirror, and you’re like, “Wow, I’m serving customers. I should have never sold, and what do I do now?” We didn’t know, but we need to treat everyone respectfully and with value along the whole way, even if we took them on against what we should have done. So as a result, we’re now seeing low double, low to moderate double-digit churn. I think our board would say moderate to high.
But the other thing about what we do is we’re selling a new product to a new market. So for example, I very much look up to Gainsight who we just signed today as our CS platform. I guess I could talk about it or whatever. But you know, we look at Gainsight because years ago, they started evangelizing the customer success automation, and so on. But also just customer success as a practice, and they’re not the only ones, but they’re one of the thought leaders, certainly. We’re in a similar boat in content optimization, but the challenge is because nobody really knows exactly what you’re looking for. You only have kind of hypotheses. And so we’ve basically used that VC money to scale the sales model that has decreased the amount of founder led sales, which is good, but has increased the amount of bad fits through the various objective factors that we’ve only realized later. And as a result, we’re kind of going through churn right now.
And then on top of that, there’s COVID churn, which is just a very tough situation for people professionally, personally. And so we are trying to be incredibly respectful to that. So we’ve sort of switched to this mode where we want to add value, we want to be supportive. If we’re not the best partner for any reason, we do want to tell you that as soon as soon as possible. So we don’t waste your time or our team’s time. And we want to deliver value, but sometimes we have to just reject folks or screen them better because what they want is either not what we do or we can’t help them enough to make a justified cost. So we’re kind of like working our way through that. But the one thing that has really helped just is getting all of those, that population of 5,000 plus folks, playing around with it. And that has helped us pore enough data through to really help us calibrate that model.
[46:09] How about revenue expansion. Are there things you’re doing to cause that?
- There are now. So the last few years we’ve done that idiosyncratically. So either my co-founder, or me or, or a salesperson would just see an opportunity and push, push, push until we got some expansion. Now our VP of sales, Marcos, has actually pioneered that account management approach. And now we have two account managers. And so they’re looking at both retention but also expansion. And we also installed a Pendo. In our case, for usage tracking.
So we started looking at product usage tracking about a year ago, which is arguably a little late but you know, these platforms take time to wire up, and they’re a little expensive too. But it was a milestone when we did that. So you can look at product usage, we can look at ROI, or value. So we build some proprietary stuff to basically say if you pointed up this website, how much money can they make if they use MarketMuse to optimize some amount of their content. And then we can basically prove out the business case. So we’re building that or we built that. But it’s not public yet, but we use it internally.
And when we actually looked at our CS health, we came up with a quantitative CS health score in February of this year. I mean, we’re in 2020. We had been in business for like four years at this point. But we just came up with a quantifiable health score because we had subjective health scores, but now we have a data-driven one. And we took 12 different types of inputs, like, do they pay their bills and on time? Are they using the platform? What is the quality of our relationship with this particular company, and so on and so forth. What are the demographic things like revenue, and so on? And who is the decision maker? And who is the day-to-day, and who are the day-to-day, and how are they approaching it? And so we baked that into a model and then that spits out a number. And then if the number is too low, then we have to do some remediation.
[48:25] So the product uses tracking, do you use a third party app for that? Or was that just homegrown code?
- The use of tracking is mostly Pendo. But we’ve had to wire things mostly . . .
[48:37] Mostly what?
- Oh, I’m sorry, Pendo. P-E-N-D-O. I think at this point they raised 200 million. They’re out of like North Carolina, Raleigh, I think. Really, really good product, really good platform. Gainsight, since I’ve plugged them once, I’ll plug them again; they also have a competing offering called PX. But Pendo is a little more focused on that specific thing. So we use Pendo. Pendo has been awesome. It’s kind of like Google Analytics for your product.
[49:12] Yeah. Yeah. Which would be very valuable. All right, let’s look at the questions. So now, before we finish up, and I know we’re rapidly running out of time . . . People in the process, so let’s see, we have more questions that we can get to. Let’s talk about attracting talent. First of all, how quickly are you hiring these days? Are you only needing to hire every now and again? Or are you adding people like crazy?
- We are somewhere in the middle. We have four positions that are unfilled. So we’re hiring four positions. We’ve done both. We were just coming from a period where we didn’t hire anyone. And we’ve even had, at different times, reductions in force. We had a reduction in force last year; that was very unfortunate and painful. But it was unfortunately the right thing for everyone to keep the company going. But right now we’re kind of slowly hiring. We wish we could hire faster. But there’s just a certain pace too, you know. There was a time when we hired very quickly, but it was too fast.
[50:26] Let’s talk about, do you have systems or processes in place for recruitment and hiring?
- Kind of. So we use an applicant tracking system. In our case, we use Greenhouse. The early hires on the business side that I found were through, a lot of them were through angel.co or AngelList. So we found a lot of great entrepreneurs through there. We also had people just kind of inbound through our industry through word of mouth referrals, “Hey, I think what you’re doing is cool, etc.” So that’s also, you know, when that happens, it’s great. We’ve had an employee referral program, so we pay employees dollars if they refer someone. And that’s not why they refer, but we do reward it.
We also do, for engineering talent, we post ads. We post ads in Stack Exchange. And then the people who answered the ads go through an automated coding test, we use Codility. And then the people who hit a certain score, then get an interview. So we hire engineers all around the world. Usually, our time zone, or something close to it, but not always.
Yeah, there are some other tools that we’ve looked at, it’s kind of accelerating hiring and so on. But we ended up doing so few at a time that we haven’t had to really scale it up tremendously.
[51:53] Okay. And how about in terms of documenting your business processes for all of your repeatable processes—things like content marketing, hiring, onboarding new customers, if you are running ad campaigns, that would be one. But you get the idea. There are certain things in your business that happen over and over and over again. How, if at all, are you documenting those processes?
- Yes, we’re starting to get to the point where processes and SOPs—standard operating protocols—are becoming more and more important as we scale. We’re at 42 people now. So that’s where we’re at. We use Confluence, it’s kind of our internal wiki. We’ve looked at other things, and we’ve just stuck with Confluence. It’s pretty good for what it is. Confluence has a feature called templates. So we’ll use the templates to make these processes. It’s not the best solution. It’s not the be-all and end-all. But that’s where we’re at right now.
For onboarding employees, which we found to be incredibly important, we use Greenhouse onboarding. And our onboarding flow has something like 50 something steps. And I would even argue, you know, that’s not enough. There are simply weeks and weeks and weeks of onboarding at minimum to get somebody, you know, with all the things that they need to know that are our best practices.
We try to document everything. Every time we change a process, we try to change the documentation to keep it one-to-one sync. We try to document everything in Slack. We try to have agendas for every call. We do employee satisfaction measurement through Lattice. We use KPIs or OKR through Lattice HQ. We do employee reviews. We try to be document everything. We try not to put on live meetings until we have the agenda outlined, have discussions in Slack, etc., etc.
And then we have some remote culture stuff like we use Donut which pairs people up in Slack. And then they have kind of serendipitous conversations. We’re about to try twine.nyc for the same purpose. Our board member Lauren started it, but it’s like a way for people to connect. We use Zoom obviously, as you know most of the world now.
And yeah, we just try to talk to or learn from remote companies. And every time we hear a new tool—one of our five company values is experimentation—so we just try to run more and more experiments all the time. And then some of them don’t pan out or don’t have the expected result, and some of them have a 10x result.
[54:42] You should keep listening to the podcast then because I asked him almost every founder I have on to talk about the process.
[54:50] It’s something I really enjoy.
- In fact, I just reached out to one of our VCs, because we have 12 VCs, and I’m like, “I’d like to download my stack.” I use Blissfully for analyzing our SaaS tools. Like, “Here’s all the SaaS tools that we use, can we take this stack and compare it to your other portfolio companies’ stacks? And maybe I’ll learn something.” So if you have a stack, or anyone has a stack, I’d love to just compare because the tools are like half of the job.