
In this conversation, Mark Roberge sits down to unpack why scaling is not a milestone, but a system that must be intentionally designed and continuously recalibrated.
Loading summary
A
Welcome to the Revenue Builders Podcast, a weekly show featuring B2B sales leaders and executives. Hosted by five time CRO John McMahon and Force Management co founder John Kaplan, the show goes behind the scenes with the people who have been there, done that, and seen the results. If you enjoy our content, please subscribe, rate and review the show to help us reach more people. Revenue Builders is brought to you by Force Management. We help companies improve sales performance, executing the growth strategy at the point of sale. Find us@ForceManagement.com Enjoy today's episode.
B
We're joined by the one and only Mark Roberge. He's the founding CRO of HubSpot, senior lecturer at Harvard Business School, co founder of Stage 2 Capital, an author of a new book that's coming out February 3rd, the science of Scaling. So, hey Mark, what I love about the book is that it doesn't treat scaling as like a one time event. You know, it doesn't treat the ideal customer profile as a one time, one and done event. That doesn't treat product market fit as like one and done. And it treats scaling as like a system and a process with guiding metrics that evolved through the, you know, major stages, you know, from startup through growth and then as you put it, you know, building a mode around the business and around the product. But you start the book by talking a little bit about, you know, five failures that you've seen multiple times. And I have two seed funded businesses, you know, like premature focus on top line product market. It's not data driven. Let's go over a couple of these things real quickly.
C
Yeah. Thank you. Johnny Mac and John Cap, you guys are some of the few folks that seen him more than I have. So I appreciate that it resonated the stuff that you've seen. Appreciate you guys checking the book out early too. Yeah. So Mac, like as you said, there was never an intention to write this book. You know, it's sometimes you go out, you do speeches, you write stuff that you're seeing and sometimes it hits in a timeless way, in an abstract way. And I, I just felt like every time I went and talked about this, I had so many founders and investors come up to me and be like, oh my gosh, like I was about to stumble on that or I have stumbled on that and like it just you, it just you fall onto something. Even this retired faculty member from HBS that's been retired 20 years is like, wow, you've stumbled across something really cool. He's seen a lot in terms of frameworks, et cetera. But yeah, Mac, as you point out, really over the last 10 years I've been out of HubSpot as an operator. So I've gone from seeing one play 80 hours a week to dozens every week. And you know, just, you just end up in a contextual area where you see a lot of patterns. And for the five years between HubSpot and starting stage two, I was full time faculty at Harvard teaching the sales course. But I was invited to serve on the boards or invest or be advisors to a whole bunch of different companies. And I chose one a quarter to really dig in some like you did with me, Mac at HubSpot. And some of them, one went ipo, a couple went to billion and a bunch went bankrupt. And it was just like, hey, why? And that's where everything pulled back to was the inappropriate choice of when to scale revenue and how fast. And it's like, it's crazy that when you ask a lot of these boards, the, the time that they choose is when they get an infusion of capital and the pacing they choose is to look up the first two years of Snowflake and copy that. That's crazy, right? Or like they, they use the, the expected internal rate of return of their venture capital asset class to drive how fast a software company should scale. And it's just like I, I think that the failure rate in the ecosystem is unnecessarily high because of a lack of rigor and framework to, for these critical questions. And I don't think those answers are right. I think we should be able to look at the signals of the business to understand when and how fast. And so your question, Mac, like what are the common things? Is like first off, just like going too early without a proper analysis of what it takes to start adding reps consistently. And we talk about that and you've lived through that, both of you. Another big one was just like sometimes, oftentimes with like direct sales businesses, scaling means adding sales headcount. And they kind of bring up the spreadsheet and say, all right, we need to triple this year from 5 to 15, which means we need to add 8 salespeople. Let's do that in January. And, and it's just like, as opposed to it just being a pacing that occurs through the year and forever. Very, very common mistake. So those were like, yeah, you see.
B
This all the time. Well, a couple points that you, you touched on. One is they take money and then they just hire a whole bunch of sales reps.
C
Yes.
B
They've never done a proper ideal Customer profile, right. So and a lot of times, unfortunately, it's driven by the founder who really thinks everyone is going to buy their baby. And I've said to them, at some point everybody will buy your baby. But right now, if you want to efficiently use those dollars, we have to figure out what the ideal customer profile is and aim our arrows at that ideal customer profile to get the greatest return on those dollars. That's one. Number two, what you see all the time, which you pointed out, is they might have a six month ramp time and let's say they're on account of the fiscal year, they don't start planning till November or December.
C
Crazy.
B
To your point, they tell the CRO, hey, you had 10 reps, they did $10 million, let's say so averaged 1 million. Okay, next year you're going to do 25. Hey, how am I going to do that with 10 reps? Well, go hire 15 more.
C
Right?
B
But when you hire 15 more right now, you're not, you're not going to get all A's, you're going to get one A, you know, seven Bs and the rest are going to be Cs. And their managers take the quota and then they horse whip the guys and then people like, I can't work in this culture and give it, give it like six months, they'll have a really bad quarter and, and then the VCs want to pull the reins back on the company. It's seen that thing many times.
C
Let's work backwards of course. Like you know, it's, how many times have we, you know, I remember I just, even this year we, we just did the investment and the VC was like, okay, cool, hire eight reps next month. The company doesn't even have a sales rep yet. Like there's no, there's no like understanding of, let's even talk about like how many applications to first interviews to like reference check to offers do we want to have? Because that funnel probably would represent the quality of the hires that we're going to have. You know, and who's even doing that? We've never, we've hired two reps in a year and now we're going to hire 15 in a month.
B
By the way, we haven't done the ideal customer profile to understand which Persona you're going to sell to for which particular use case you don't know if the reps have that skill to call on that size company to that Persona.
C
Let's go to that one. Why does it happen? And I think it's like I think where it comes back again is like a little bit of confusion around. I took venture capital and the needs of a VC firm inappropriately translated down to the execution of the business where hey VCs, you got to invest in stuff that's going to exit north of a billion dollars. Like we need to see multi billion dollar tam. It's just how the asset class works. I think the confusion happens is like the entrepreneur thinks they need to prove that in the first year. They think they need to prove that they have customers in apac, EMEA and North America, SMB, mid market, strategic healthcare finance. It's just like, no, we just need to like we need to believe that it can go there. But honestly, as you go from 0 to 1 and 1 to 10, let's pick the market that's going to be the easiest to go there, that is repeatable and that's where it just goes off. And it's crazy that like I have to stand up in front of founders and say, you think that your ICP is what you put on your website and what you put in your pitch deck. Your ICP is, is who your salespeople go after and close that. That's the market that you're in. And if you're not like diligent, they're always shocked when I tell the story of halligan in year three of HubSpot ripping up the contract from Meta that we sold them because they would have destroyed us even though it would have doubled our revenue in one quarter. This is like it sends the signal to them.
B
The other thing that you see too is when they feel like they have to hire those eight reps that you described, they think they have to do what I call sprinkle the infield. They have to put one in every major city in the United States. When I say, wait a second. If you do your ideal customer profile, you might find that you could stick three or four right in the Bay Area right now.
D
Yeah.
B
And never.
C
They're all there or New York or wherever they. Whatever we're selling to you.
B
Yeah. If that's where your ideal customer profile is, that's where you stay.
D
Then the other thing is I just want to put a little plug in for the book again. Science of Scaling, it's pre order now on Amazon. It's going to launch February 3rd. But we want to have this. If you're listening right now, go out and get this book.
C
Thanks, Cap.
D
You're welcome. I want to talk about this science part because one of the things and I told you before we got on the podcast, just for the audience, I had to take some of those special. I don't know, what's it called? Gilbo. Something to get my brain synapses going because. No, I'm serious. Because what I love about you, Mark, and it's in the title, like, I remember your background. You're a technical guy. You're like, I think you started as an application, not an application engineer, but as an engineer.
C
Yeah, yeah.
D
You went into sales.
C
Y.
D
What I love about this is you brought the science and the art together. And I don't. I. I really want people to understand what, what you're getting into when you go get this book, when also you guys are talking about here, we're talking about the seller. We're beginning to talk about product market fit and ideal customer profile, but we haven't even differentiated on the type of seller. So the science gets down to it's not a seller, is not a seller, is not a seller. In this science and in this art of scaling. Could you talk a little bit about.
C
Yeah, definitely. Cap. So the way the framework works is it clearly defines three phases you need to go through to go into the growth mode. Product market fit, go to market fit, Growth and moat. Nothing crazy in there other than like, go to market fit is probably a little less known than product market fit. And I would say, like, when I do do speeches to students, founders, entrepreneur, like conferences, and I say, when are you ready to scale? A very common answer is product market fit, which is a great answer. And that's a term that probably was invented, I think, in the beginning of the century. Eric Reiss and Lean startup Steve Blank. What's crazy about that, even though it's a good answer, is if you go back and ask those hundred entrepreneurs what product market fit is, you get 100 different answers. Half of them are like a revenue line or a customer line. I have product market fit when I have a million in revenue.
D
No way.
C
The three of us can sell ICE to Eskimos. Eskimos don't need ice. There's not product market fit. We can get a million dollars of reven. It has nothing to do with revenue. It has everything to do with customer value creation.
B
Yes.
C
And in a lot of businesses, that is measured by retention in the long term. And the book proposes how to form your own leading indicator of retention. So that's how we root and measure that we have product market fit. And then you go to the next phase. If you do, which is go to market fit, which if you do product market fit, well, you're following the words of Paul Graham, the founder of Y Combinator who says do unscalable things early. Like when we're chasing product market fit, we don't need a thousand customers. We need like depending on our acv, we need like few dozen meetings, you know, probably like 50 intros, 30 meetings, you know, 10 pipeline and five customers. Like we can do that through our network. I don't want to talk about standing up an SDR program. Content marketing. Like we're doing, we're doing founder selling. Like I don't want to pay my rep on commission. I want to put them on like just in equity just like my founding engineers. You know what I mean? Like we're, we're operating, we've seen this mode before but then once we achieve it and it's like holy cow, I signed up another five customers this qu. This quarter and four out of five are successful and going to retain. We have product market fit. I agree. Now are we ready to scale? No, because we need to go to market fit. We need a playbook. We need to know how we're going to comp our reps. We know what the, what the scalable pricing model is. We need a scalable demand gen channel. Right. So there's a new set of things we need to work on now that we shouldn't have been working on before. And we measure that with unit economics. Then we're ready to scale. Now to your point, Cap, I talk about over the decades I've talked about designing a go to market system. Not thinking about sales as an art form, but trying to systematize it. And I outlined clearly the components of go to market system, one of which is the profile of the rep that you hire, which you're talking about Cap.
D
Yeah.
C
And so that's what, not only does the book outline these three stages and how you know where you are, but it also informs you on how you optimize the go to market system. For each stage, the rep you hire, the demand gen channel you choose, the way you pay your reps, the way you price your product, et cetera. And you know, honestly like I think the hiring model that, that was best initially crafted very well on the sales learning curve that you guys might remember from 20 years ago. And basically like, and I think the work is like evolved a little bit. But like that first hire, it's not about like the first like the number one rep at PTC right now would be a terrible first hire at product market fit. Hundred percent because like they, they joined the company and they're like, they went to a month of training. Medic objection. Handling battle cards. Their manager was there with them. These guys, these folks have nothing. Right? Like we, we need someone who's like half account executive, meaning like they can run a commercial process, blah, blah, blah, talk about money. But almost like half product manager where they can like, yeah, look at like 20, you know, and then communicate to the engineers. Right. That's our first higher cap, to your point. But once we get to go to market fit, now we need a process. We need a process builder. And once we go to the growth mode now we can deal with the number one rep from ptc, someone who's like a process follow their coachable. So that's, that's how the framing works.
D
Well done.
B
You had a lot in there, so unbundle some of it. And beside some. So it's so true what you said, that the first couple reps, especially if you hire somebody as a CRO, as a startup, they have to have that mentality that, hey, I'm basically half product manager and I'm half sales rep. Definitely. Number two, going back to the snowflake thing where they think that, you know, everybody thinks the thing just, you know, took off like a rocket. And I can remember standing in the hallway with Chris Degnan, we had an ideal customer profile and at the point, at the time, could only sell to tech and ad tech, but kept going down those bowling alleys making money until, yeah, we had enough revenue to fuel development to go into other bowling alleys. Right. So the ideal customer profile continued to change over time, but we weren't just sprinkling the infield with reps and then running out selling to all different types of industries. We knew where the money was. It's like the old Jesse James thing. Why do you rob banks? Well, because that's where the money is. Right. One other thing you mentioned that's a major concept in your book, the science of scaling, is that you talk about the leading indicator of retention. Can you, can you expand on that a little bit for the Definitely audience?
C
Definitely. As we discussed, the three phases I see these companies go through is the first product market fit, then go to market fit, then growth and moat. And the subtitle being using data to decide when and how fast to scale. So we need a quantitative way to know where we are. So, you know, even we read about product market fit, like the Wikipedia page, half the definitions are like, it's a feeling. You just know it when you have it like that doesn't settle with me. No, as like an investor, you know, as a, as a CRO, that's like putting my career on the line. And so as we discussed, like I think the purest definitions of product market fit is not like how much inbound demand you have, it's not how many customers you have, it's not how much revenue. It's what we talked about is the predictably creating customer value. Like when you promise something to a customer in a sales call and they buy it, that it comes to life. That's like qualitative of what it is. In a lot of businesses today, especially the ones we live in, that's best quantified by long term customer retention. Like, you know, we talk about greater than 90% customer retention is great. Greater than 100% net dollar retention is great. That's how we think about it in the ecosystem. Now the challenge is in startup world we have to move fast. We can't sign up a bunch of customers and then say, hey, let's just wait a year and see if we have product market fit. We don't have that time. And so we have to def. We have to come up with a quantified metric that we can see in say the first month of a customer's history with us that if it occurs, they will retain and if it doesn't, they probably will leave. So that allows us to like hinge this concept of product market fit that's rooted in customer value creation that ultimately is measured by customer attention. But to understand where we are with it every single month that we sign up new customers. And I encourage folks with a framework of P percent of customers do E event every tee time. Right. So you just boil it down to three variables. P, E and T. So Slack has a very famously documented version of this. 80% of customers send 2,000 team messages every month. Perfect. Imagine if the founder of Slack, you picture in the early days, there's six engineers in a room, they do their series A and the founder of Slack says our first North Star is to get to 2 million in revenue. You can start to see what starts to happen. Right. Versus they say our first North Star is to get 80% of our customers to send 2000 team messages every month. That's just the right foundation at that point. So that, that's what I propose. Mac is the lead indicator. Attention is P. Percent of customers do.
B
E event every tee time, percentage event and timing.
C
Exactly.
B
Ir love it.
D
I love this because the more companies that I'm interacting With lately and in the beginning when I heard this, I was kind of like they would start to talk to me about how much time somebody spending in the platform, how much time a customer spent. And so I was thinking about this and I'm like, yeah, but like the revenue and, and I'm guilty of it because I'm less science typically and more on the, probably on the go to market side. But I completely understand what you're saying because and this is the, this is the caveat, I think if you don't have good technology metrics, vision insight to what your customer is doing or to what that potential buyer is doing or what that journey is going to be, meaning you're not outside in, you're inside out. So we're going to do this and then we're going to feed it to the market. We're going to do this inside, then we're going to feed it to market versus we fed this. The market has responded this way. We're going to change it that way. I find that people just miss the boat. And you talk about it in your book. You talk about inside out thinking versus outside in thinking. You can't do this without understanding how customer is going to use your product.
C
Right? And like you brought up time and time in product, which is great. I mean you're getting in the right direction. And that's a tricky thing. Cap is the E. I haven't found like a universal correct E. Just like we don't have like some people run medic, some people run band, you know, some run yes. It's just like it just depends on the context. The guidance I give folks there is like first off, like I see so many people that the first stab they take on it is like they show an aggregate number across all customers. Like, here's how many transactions ran in the install base last month. Look how it's going up. It's like, yeah, no kidding it's going up. You're hiring more, you're adding more customers like you. Okay, you had 10,000 transactions last month. Did 9,990 of them come from one customer and everyone else is not using it. Right. So we have to look at it on a per customer basis so we know how many of them are in the green versus the yellow and the red. And then the other one that Mac, you're going to love because it ties it to comp. And the thing that you guys do so well is appreciating your unique differentiator and holding the reps accountable to it. So basically when you define an email event very well, it is a unique differentiator. And I can tell a very personal story from HubSpot. So the e event for us, the whole formula was if 80% of customers use five or more features in the HubSpot platform every month, then that was our LAR because we were a platform, right? We had the marketing product, the SEO product, the blogging product. We had like 25 modules and that was our differentiator against Hootsuite and Marketo and Mailchimp and like if you wanted the best in class you go there but if you wanted the all in one. So because of that the E was the number of features you were using. That's our unique differentiator as Mac always says. So that's cool. Now you instrument this philosophy of unique differentiation. Now what's even cooler is to incorporate that into the comp plan. So our LIR is when the customer uses five or more features in the platform we got them and a lot of customers can get, get through that in a week or two. Like they sign up, two weeks later they're using five or more features. We can pay the reps half of the commission when the customer signs the contract and half when the LIR is achieved. So we don't have to wait a year for retention. But we're forcing the reps to be LTV focused. And what a lot of times when you do this if you happen to have a, a product context where you have a free trial, you have reps getting customers achieving the LAR before they sign the contract to accelerate their commission payment. So there's, there's some magic that can be orchestrated here in the entire go to market system to align with the health of the business.
B
You see that a lot in the consumption model as people consumption pricing, they're booking their, their complaint says I'll pay you so much on the booking, but I'm going to pay you really at the end of the day I'm only going to pay you when they burn down the credits.
C
That yeah. So I do want to make one.
B
On creating, helping the customer, as you said, create value with the product.
C
I do want to bring up one point for your guys audience in particular is while a lot of this was written from the perspective of founding and starting a company, I have worked with hundreds of businesses that are billion dollar businesses bringing new products to market using the same exact thing. Right. Like I think you guys probably do this more in your work is hey, we're a multi, you know, you know we're a global con company trying to know, add two growth points, you know, to, to our growth or to our ebitda. And we're bringing, they're bringing new products to market all the time. And I think you both have all lived through like, okay, we're going to launch this new product like Engineering, you build it, you know, we're going to launch it at the customer meeting in April. Engineering, you build it. Marketing, change the website. Salespeople get the training going and the competitive battle cards and the pitches and then they go to the conference in April and they launch it and the reps start selling it and they miss the number by like 50% because they don't have product market fit. And I think one of my colleague, one of our stage ULP is that kind of does this at Microsoft. I love his framework. He talks about in 555005 reps, 50 reps, 500 reps. So he's like, we got a new product we're bringing to market. Great, give me five reps. Let me prove product market fit according to the science of scaling. I don't know if it's going to take a week, a quarter or a year, but we're going to do it. Once we do that, give me 50 reps, then we're going to build the go to market fit. The process that all this stuff, I don't know if that's going to take a week, a month or a year, I don't know. But then once we got that, now I got 500 reps. So like the board's like, okay, I get it, I need $1 billion of value here, but I get it that they're not going to be able to do on day one. Right. So I just want to make that point clear that this methodology is just as important to big companies bringing products to market as it is startups bringing new products to market as well.
D
Although your brain is very science oriented coupled with art oriented, I want to say that this is, this is a frame, this, this is a framework. And I like to say that wherever you are, there you go. Because we haven't yet talked about the moat, which we will. You've been alluding it, alluding to it a little bit with unique differentiation. But today's unique differentiation is, you know, is tomorrow's comparative differentiation, meaning everybody's going to try to do that. So these are kind of sliding scales and the frame. And when I think about a framework and you've done such a beautiful job in this book, you put the Framework on top of wherever you are. And then you see what's missing because it's going to change. How many companies have we all worked with that their ideal customer profile, their go to market fit and their product market fit actually begins to slide. Because something that happens in the market.
C
Every single one.
D
Yes. So whatever you came up with, what is it? Von Molke's, you know, no battle plan survives contact with the enemy type of thing. You can write the plan, but then when you're out there in the market, it's going to slide. But if you don't have a framework, you don't, you're not sliding. You're, you're, you're just washing away. Does that make sense?
C
Approaching $2 billion in revenue, I think, and they're still measuring the LIR every quarter.
D
Yes, you have to.
B
That's right, Mark. I want to go back to icp. You know, you treat the ICP in the book and again we're talking to Mark Roberge about his brand new book due out February 3rd, called the Science of Scaling. And you treat, you know, ICP as a scientific input into scaling essentially. But what I gotta understand is why don't so many companies build an ICP and then to Cap's point, the market changes, competition changes, your product is constantly changing and they don't go and change their icp. To me, it's like you're constantly going through, if you do it right, constantly changing your ICP and constantly looking to see whether or not there's real product market fit in different company use case.
D
I have a hypothesis. Can I try?
C
Mark, Go, go.
D
Because I look at this, I do this for a living for the last 20 years. And I find that if it's emotional and ICP is emotional and it's not backed by data, we real data. That's the fundamental principle that I see. Because if you have people like me, you guys are a little bit more scientific than me, that can will their way to selling things and no customer is a bad customer and we can sell something to, you know, this customer. If you don't have the data to support what actually happens from that revenue from that customer. I think that's fundamentally like you said, people are chasing revenue versus ideal states. That, that's my hypothesis and I think it's, it's supported in your book.
C
It's so hard because like the, I think a great ICP is based on ltv, not lowest cac. And that's where a lot of people get that mess.
D
Just for some of our Young listeners or some of our early people. Would you mind kind of defining what that is?
C
Yeah. So it's just like a lot of people and it. There is an intuition there, but when you really think about it, it's a broken intuition which is like, okay, what's your ideal customer profile? Some people might say it's where our inbound demand is, whoever's coming. And it's like, that's scary because that's like, that's not necessarily correlated to who you built the product for and who is going to end up being a successful customer. The next level answer is like, oh, it's wherever we have the highest close rates, which usually would correlate with a lower customer acquisition cost. Right. Cac. And again, like, only true. If that's a great customer that's going to stick around, really needs to be around lifetime value. It really needs to be that the ICP is based on segments of the market where when people buy it, they see tremendous value, they renew, they expand, and they tell their friends. And sometimes those companies are not necessarily the highest inbound demand to your website and sometimes they're not the easiest to sell, I. E. Have the lowest cost of acquisition. But like, it really needs to be around ltv. And it's a cool thing because I do talk about at the end, I had to talk, I had to write appendix about how AI will change these concepts.
D
Yeah.
C
And I still wanted to write the book because these are not like post AI irrelevant. These are principles that will need to be built into the models. And I'm very excited about how AI will guide this ICP piece. And I actually write a little bit about it. So, Mac, what you saw in there was like I took a little deeper view on the ICP and so I basically set up this green, yellow, red framework that I use with a lot of companies. And basically what you're picturing here, folks, is you have five to eight attributes of your icp. It's common to have company size, location, industry, stuff like that. So we might have five in there. Okay, so let's just work with three to keep it simple. It's like they're in the US they're between, you know, 50 mil and 500 mil. And then there we'll deal with health care and finance just to have a working model. Okay, so that, that is our green icp. Our red is like green is like, we want to sell these, we'll take them all day. If we're going to build an email campaign, we're going after we're going to build a cold calling campaign. We're going after those people. Red is, you cannot sell these people. If you sell them, we'll rev up the contract. So, like, if they're in Europe and Asia, we're not ready for it. If they're in manufacturing, we're not ready for it. If they have over 500 million in revenue, we're not ready for it. We're ripping it up. But then like, the yellow is like, we're not sure, you know, like, it's probably going to be where we're going to go next. But if someone comes up to our trade show booth and says, I've been looking for this my entire life, how do I sign up? Or if someone comes as an inbound lead and they're like, I, we. This is our number one issue this year. And they're in the yellow. Maybe they're in the tech sector. Maybe they have 40 million in revenue, not 50. So they're close. Then we can sell those people. And that allows us to create a structure where we can explore the periphery of the ICP to naturally expand. And that's what Degnan talks about@Snowflake Mac, through your guidance. When I asked him, like, when do you go upstream? He's like, you're pulled up.
B
Yes.
C
He's like, that's what happened to us is we were selling SMB mid market to the, the smaller sectors. You were talking about Mac. And then once they hit 5, 6, 7 in revenue, they had these early adopter enterprises come in.
B
Yeah.
C
And that was in the yellow for them. And they're like, okay, you know what, we'll give this a shot.
B
Right. And your point is? We're not going to have our sales reps target the red or the yellow. But if a customer comes to us and almost like, wants to buy, sure. Okay. We'll, we'll the effort.
C
And then we wake up.
D
You guys are making an assumption which, which I love. But what you just said, Johnny, is, you know, we have a red and that sellers aren't going to sell to it. But in reality, like when we go into these things and I ask a question, what's your ideal customer profile? And if I'm lucky enough to get it, then I get it. And then I look at the pipeline and I'm telling you it is, it's right there. And so how do. What advice in the red? Yeah, they're in the red.
B
It's right there. It's in the red.
D
They're right there. So, Mark, like, let's Say like in a framework example. In a framework example where you're laying it down like an X ray and wherever you are, there you go. And you see a pipeline like this, you're talking to existing customers. Let's say you didn't do this, you know, sequentially. And again you are, there you go. What do you do as a company as. As a total company, not just a dealership, you know, sales, marketing. You gotta readjust to prevent it.
C
Or if you wake up now and you realize you wake up and what do you do now?
D
Wake up.
C
Yeah, sorry. So you, you fix it at that moment. Right? So like I've never seen a company where we're starting out, we're, we're pursuing product market fit and we do our icp. We do that with every company. I've never seen a company where that initial hypothesis is right and state, even for the first year, yes. Like, it's just, we don't know. And once we start going in there, a dozen, 2,000, three dozen customers, we start to see the patterns. It's like, oh my gosh, we totally missed that. Like, it turns out that there's a critical component of the tech stack that we're not ready for. So we have to put these people into the non icp. So you'll see in the frameworks there that when we're measuring the pursuit of product market fit as measured by the lir, we push out some early customers who are no longer icp and you can just. And then, so when you're measuring the LIR and you're, you're only measuring on the current icp. So that would be my answer. Cap is love it. Nail it down now separate your install base into ICP and not. And then like on the not, like just put them over here. Maybe if you run like an account manager ratio of like, you know, one account manager to 30 accounts, you know, whatever it is, let's just run it as a one account manager to 200 over there. You know what I mean? Like just, we're not going to ignore them, but we're not going to, we're going to model pretty bad.
D
This is an important part right here. I love what you're saying because if you are down the line and this has happened to force management a few times in the 20 years of our existence, I found it very important from a leadership perspective to communicate effectively the science behind what you're doing as a company. And I have not had one company come back and like think we were jerks because we somebody was not in our isp. But when you don't talk that way because nobody, no customer wants to buy something, that is not. If you're not going to serve me as if I'm your ideal customer, nobody wants to be in that red area, right? But the sellers listen. All sales leaders listen. What most people do is they just stop calling on those accounts. They stop supporting them, or they stop. And then all of a sudden you get a bad reputation in the marketplace. You're not following up. You're so just, yeah, you need a little something.
C
It's a very difficult place to be because even, like a lot of people make so many mistakes, they try to save every customer even though we've decided they're no longer a fit. We can't afford that churn. You just got to like, wipe off the, you know, rip off the band aid. But again, you have just like a lower version. You have some people there to see what they can do. You're going to forecast a much higher churn right there. And you will wake up a year from now where if, if you listen to this and you're like, oh, damn. And you carved it up and 60% were an ICP and 40% weren't. And you put the right ratios on the 60% and all the future quarters only sell into that ICP, you will wake up a year from now and 95 will be in the ICP and 5% won't. And I get it, Cap. Like, it's tough because there's going to be a bunch of customers out there. Like, I tried it and it didn't work. And that's. There's nothing you can really do about that other than being crisp about who you're for.
D
Love that. Thank you for that.
B
The huge benefit that, that we're not talking about right now is that this isn't only about sales and selling this. When you do the ICP right, and you know, you have product market fit, it rallies the entire company from a marketing standpoint. They know what the sales people are talking about. It rallies development when they're prioritizing what features and capabilities should be in the next releases. That's a huge part of this that I don't think most people really take into account. And. And that's why it's so important to have an ICP and truly understand product market fit and where you to your point, Mark, you know, creating value for that customer.
C
There's a reason why I didn't call it the sales system. I call it the go to market. System. There's a reason why I didn't call it sales market fit. I called it go to market fit because like we, the sales team is part of the go to market team. The go to market team is how we create demand marketing SDRs, how we sell that demand, sales, how we retain that demand and expand that csm, account managers, retail Rev Ops is supporting us sales name like that's the whole system and those have to be looked at holistically. That was a massive movement in the cloud era that helped companies, you know, separate themselves.
B
Let's talk a little bit about aligning go to market with product market fit and doing.
C
Yeah, definitely. Hey Matt, can I do something really crazy for whatever you want really quick while we're talking about ICP and you guys are so passionate.
B
2-3-3 by Marco whatever you want.
C
I want to go Star Trek and talk about something that probably won't ever happen, but maybe, maybe it will. And I want your guys feedback and how AI will help us with this ICP thing. Okay, so this is what I'm kind of seeing here where if you are world class in ICP evolvement, you have this Rev Ops team that maybe once a quarter sits down and re evaluates the icp.
B
Yes.
C
And they go through various customer segments on all the things we talked about, size, industry, you know, location, blah blah blah, and hopefully analyze the LTV and see and find new hot pockets where the LTV is going up and we can recalibrate our demand generation tactics over there and maybe some areas where we've lost it a little bit and maybe we turn that off a little bit and go back to product and say, hey, what's going on here? We're, we're no longer competitive. What can we do that's like world class today. Now think about what it's going to be like when AI does that. Because humans are limited to doing it once a quarter across like a handful of attributes, AI can do it every hour on an infinite number of attributes. And so like where we might be able to go at some point is the AI is just running and looking at the pipeline and looking at the conversation. The discovery calls the demos the pipeline evolution and it's, it's evaluating that. Hey, guess what? We have sales capacity to call 17 new accounts. This week I have recalibrated the ICP and I've looked at the 129,000 companies that are in our ICP and through a variation of like strength of the firmographic as well as engagement. These are the 17 accounts to call. So the ICP, do you. The ICP becomes dynamic and fluid. Is it too Star Trekkie?
D
No, I think it's, I think it's very close to. I think it's. You know, in the old days I'd read something, I can't remember who that dude was that would read all those, those books about like what was going on in Washington D.C. and about spies and like then I find out the guy had 30 year history with all the intelligence agencies. What was it?
B
Isaac Asimov?
D
It might be. It might be. But so like my point is if you can dream it up.
B
Mark.
D
Yeah, I know it's coming because what everything that you just said people are doing individually like right now, totally in different segments. Like okay, I want to capture what's happened in the conversation. I want to capture what's happening from differentiation, I want to capture what's happening in pricing and I guarantee you that that's the world that we're going to live very shortly.
C
One of many cool workflow iterations that we're starting to see here.
D
So love it.
C
All right. Anyway, just to capitalize. Go ahead Mac.
B
Yeah, please add to that. That is you do cease. You do think you have your ICP locked and loaded and you have some really good sales reps that know how to venture into different corners of these companies. And they find, I've seen it a couple times now at different companies. They find a use case you had no idea you probably could do. And it. And it's a big time use case. The ACP is gigantic. And then you know, once that person finds that you try to duplicate it.
D
Exactly.
C
I love that. I think like, and I don't know.
B
That AI will find that but because.
C
Well, I think where very creative. Yeah. Where at will Mac is like I, I have this perception of how like the AI will influence go to market. In the first phase the humans are still selling to the human buyers but the AI is behind the scenes doing all this analysis. So I think in that case, Mac. Yeah. The creative. I think it'll be a long time for AI to be more creative, truly innovative and creative than humans. There's something because requires like past knowledge to train models and the creative process sometimes is like outside of that and that's really what you're talking about Here is such a creative rep that went out and asked the great questions and they were so consultative and they were able to apply this like existing product that wasn't intended for this but ends up solving this problem. So I still think that would Work is that rep would go out and find it, maybe their neighbor would see it too and they, she'd start closing a couple and the AI would pick up on that. Like, hey, this is a totally different use case. Now we're going to move this into our ICP and when I choose the 17 accounts next week, I'm going to have a couple sprinkled in here to see if there's some energy there. So who knows me exciting time.
D
Part of the big part about AI is that you guys are talking about is. And I kept saying the name Steve Wah, realizing that people aren't. If you've listened to the the Revenue Builders podcast, we had Steve Wah on Johnny and a huge elephant hunter where he was, he was moving icps all the time at, at companies. And I think the important thing, like I don't think we could capture what Steve Wa was thinking, but I think we can capture what Steve Watt is doing in his activities, who's he's calling on and how he's calling on them.
C
What they're talking about, what they're talking about.
D
And Mark, I believe that's coming really quick. Yeah, I think that's coming really quick.
B
Mark, Scaling. You know, you talk about bottoms up scaling, so can you talk a little bit about why you like bottoms up scaling? And yeah, the leading hit question because.
C
No, no, no.
B
Because the top down revenue stuff usually fails. But. So let's talk a little bit about.
C
Yeah, it's another part of the narrative in the boardroom, right, Is like it used to be triple, triple, double, double. Now it's like 10x 10x 10x which I'm a little skeptical of what, that, what's going on there in some of those companies. But most of the time it's kind of a perplexing question. When I ask my students, I'm like, hey, hopefully you're in this situation because a lot of them are founders and they're in the product market fit phase. And I'm like, you all aspire to one day having product market fit, getting a $6 million Series A and the VC is gonna say, how much are you gonna do in 2026? You aspire for that to happen. You're, you're at a 1.2 million. You're going to close out 2025 at 1.2 million. How much are you going to do in 2026? They stare at the, the chalkboard for five minutes and then finally someone's like, triple, triple, double, double. You know what I mean? Like, it's totally Top down. It's totally top down because again, a VC is doing their long term IRR calculation of what they need to return. And like first off, let's just remember that when a company is bought eight years from now, no one's going to ask what your growth rate was eight years ago. They're going to maybe look at the trailing two years of like growth and all this stuff. And it's so like, so we don't necessarily have to be so attached to the irr and God forbid we're off from the IRR for a quarter, we, that's fine as long as we're fixing the business and doing the right pacing. So yeah, fine, we can start with a top down as a general guide. But the bottoms up is the sales capacity that you create, the demand gen capacity that you use to feed them, right? So like a lot of people get the sales capacity, right? They're like, all right, we gotta go from like 1 to 4. So I need to add 8 salespeople according to my commission math. And that's interesting. Now again they add the eight people in Q1 and we just don't have the flywheel set up fast enough. So we have to, we have to spread those like maybe two a quarter instead of eight all at once. And then we also have to factor in attrition, then we have to factor in ramp time. So these are all in the sales cycle, right? These are all coming through basic, you know, modeling from the bottoms up. We have a whole bunch of templates on that. But then the other thing that they always miss is the demand gen capacity. It's like, oh yeah, we have two reps, they're productive. And yet no one looked at it that like the entire pipeline of those two reps came from VC introductions and that doesn't scale to eight reps, right? So like whether these reps are creating the demand themselves, we have SDRs, we have a content marketing machine, or all the above. We need to make sure that we have the bottoms up sales capacity to get to the 4 million and also the demand gen capability to feed those reps, whether they're doing themselves or some outside factors are.
B
You see that all the time though as they start to scale, they don't. The demand gen does not keep up. So the demand gen can support reps. When you go to 20, it's still the same demand gen team and they.
C
Don'T find out until like the board meeting in April to review the Q1 results. Whereas we, it was very evident back in August when We saw the meetings per rep go down, but no one looked at it.
D
Let me ask you guys a question on this topic because it's coming up a lot right now with planning in companies that I'm involved with. On demand gen company. Demand company generated leads. Individual generated leads. What advice do you have in these models, Mark? When people are like, because Johnny just said, well, we're not gonna, we don't have a demand gen that can do that. Does that change when we go to two from eight or what have you, does that start to change the type of seller, where they come from and how they're gonna do their own generated leads? What's, what's the science behind that? Because people are struggling with that.
C
So it's a great question, Cap. So bring it back to the framework. Product market fit, go to market fit, growth emote. We know where we are and we have the go to market system that's changing as we go through. Who do we hire, how do we pay them, how do we price the product, how do we generate demand?
D
Yeah.
C
And so in the product market fit phase, I like unscalable things early. Let's just use professional networks and get there. We can, we can get to enough meetings. We don't have time to build a cold calling program. When the go to market fit, that's one of the things we have to do, is we need at least one scalable channel proven so that when we look at the plan to add 8 reps next year, we have the confidence that we'll be able to feed them and not just like star everyone. And when we get to the growth emote, we usually have to go multi channel. We need multiple. I've never seen a company go to IPO exploiting just one channel. Even PLG companies, every big successful PLG company got to whatever, 10, 50, 100 miles through PLG. They eventually add salespeople. Right. To like upscale, go direct, whatever.
D
Yeah.
C
So the one very important abstract concept here that hasn't come up yet, Cap, that you're instigating with your question is what I call the inappropriate cut and paste. Meaning like, oh, what do you think we should do for the demand gen? Like, should we do a company driven. Should we do a marketing driven. Should we do it the rep driven. Let's look at what HubSpot did. Oh, they did content marketing. Why don't we do that? Yeah, but HubSpot was selling marketing software to small business VPs of marketing in 2008. You're selling cybersecurity software to you know public company CIOs in 2025.
D
Yes.
C
It's a different context. Right, so, so we have to be aware of the context as we divine every component of our system, whether it's the rep we hire, the methodology we run, the demand gen channel that we choose. And that's something that I outline is like eight commonly chosen channels and what contexts are usually compatible with them and which ones are not. All right, so, so I know that's like a it depends type thing but cap, that's just to give it real substance is first off, you can just read that material and you could also focus on companies that have scaled to the next phase that you aspire to that have a similar product and market target. That's going to be a better starting point for you to like pontificate on which channels you should try. And usually the exercise you go through is you put like a team in the room for two hours and say how, what do you think is going to work? Events like direct mail, like cold calls, throw it up on the wall and then take a step back and say, okay, we have time and money to try two of them. Which two are we going to go at? And we go and like, and we try it and we see if we can find like a low cost per meeting and a high conversion rate from meeting to pipeline and customer and all that kind of stuff. And we let the data speak and maybe we got them both wrong and we tried the next two. Right. But we're always experimenting. Let's say we hit one, we find SDRs are working on this market. So great. And that's going to support our next year growth. With the eight reps, we're still experimenting with other channels because there's the, the SDR channel probably won't scale till a billion dollars. We're probably gonna have to exploit. Maybe when we go to Europe we're gonna have to go through the channel, whatever. So that, that's kind of. I know it's like a little bit of an. It depends but I think I provided a process there, Cap, on how to get to that answer you absolutely have.
D
And Guy, I know we're doing this a lot, but pre order science of scaling because I know this is about the 10th topic we talked about that I can just hear people are looking or listening to this and going, holy smokes, I need this now. So free order.
B
Let's talk about the cause. So you're taking the book and you're donating all the proceeds to a charity. Could you talk a Little bit about the charity you're donating.
C
Yeah, this one's going to mental health, specifically McLean Hospital, who I think is the. A lot of people agree is like the global leader. They're here in Boston area on, you know, psychiatric research and supporting mental health of all stages. You know, the last book, the sales acceleration formula, I chose a charity called build.org to donate to, which was bringing entrepreneurship to high schools for the worst high schools in every city to get these kids through high school and to college. And I'm really proud that that became a bestseller and I was being able to make an impact on that organization. And that was one of the big drivers for me is like, again, I never really set out to write a book, but sometimes you go out and you speak and you stumble across a framework that becomes a little timeless and pretty abstract, whether you're selling software in the US or tractors in Brazil or pharmaceuticals in Europe. And I just happened to stumble across one. And then it happens to support our VC firm, Statue Capital, and the way we invest and support. And I also, like, have been very passionate about mental health as of late. It's been a big thing for me, for my family, and also I think as technologists, I think a lot of tech people, really, the biggest thing they love is bringing products to make the world a better place. So we never want to bring stuff out. And I think the Internet made the world a better place. But with any new tech shift, there's also some negative ramifications, and we're in a mental health crisis for a variety of reasons. And I don't know if AI is going to help or hurt that. I think it will be a little bit of both, depending on who you are. And that was it for me. And it's been a personal thing for me. I've struggled with debilitating mental health throughout my life. It probably spiked the worst. I. I was at ground zero on 9 11. And that was the first trigger for me is, you know, God bless our military folks and who have to go through this all the time. But being in a. In a moment like that, and then the PTSD that occurred six months later and, you know, the severe panic attacks, the debilitating. I could. It was very difficult to work and to work through that, and it's. We just need to talk about it more. You know, it's. It's kind of crazy to think that when you meet someone and you find out they're a cancer survivor, let's say they're applying for a job and you're like, yeah, they fought, they bought, they beat cancer five years ago. You like elevate that person in your mind.
B
Yes.
C
But if you find out that they were like had, you know, life threatening depression and were in a psychiatric hospital, unfortunately a lot of people have concerns and that those are the same things. They're, they're, they're an illness that someone worked through. And so, you know, we need to, we need to kind of raise that up. I think every generation that comes through does a better job. And I, I've been so, I, I've been blessed with some wins in my career that people think are important and they elevate you such that you can talk about these vulnerabilities and, and contribute to that cause. So that, that's where the book is going. I know this cause is very important to the two of you as well.
D
Amazing, amazing brother. That just gave me chill bumps.
B
Thanks so much for that.
D
It is like, go get this book Science of scaling now even go get it. Even if you don't like business guys.
B
On the mental health thing, it is something that people should talk about more. Like you said, I think if you're a cancer survivor, more people these days are more open to talking about it. And you know, certainly amongst guys, if you've had some sort of, you know, mental health issue or something, most guys just take that thing, stick it in a jar and stuff it down.
C
Totally. I mean that's where all the research goes is the. When someone does start to vent to you about this and it's, it's another kind of blessed position I have where because I talk public about this a decent amount, I get multiple calls a week from people. Hey, can you talk like you're the only like that? You know, I just sat down for coffee on Tuesday with someone I just talked to me about something that he hasn't talked to many people beyond his wife about. And the, the biggest, all the research points that they just want someone to listen and not try to fix it, but just like empathize with what they're feeling. And that allows someone to move through and prevents them from spiraling further into a bad spot. You know, so we'll get better about this stuff and appreciate providing the platform to talk about.
D
We should do a. We don't have enough time to get in. All three of us are card carrying members of mental health crisis and we should, we should definitely come back and, and talk about this. I just want to encourage people that are listening. Don't isolate the first thing that I know from experience. If you're struggling and you find yourself isolating and you find people reaching out to and you not reaching back, you know, there's your sign. Plenty of love out there in the world. You just got to accept it. So, buddy, I just got tremendous respect for you. Thank you for doing what you're doing.
C
Thanks, Cap. Appreciate it. Thanks for the platform to talk about. I always love jamming with you guys.
D
Amen.
B
We're talking to the one and the only Mark Roberge, author of what's going to be a best selling book, the Science of scaling throughout February 3rd. They could get it. What, Mark? On Amazon or their favorite.
C
Yeah, anywhere you want to. Barnes Noble, Amazon, wherever you want to go. It's out there now. You can pre order preorder now.
B
Thank you, Mark Roberge, for everything.
C
Thank you guys. I think it's probably our third or fourth time together in this setting and probably our hundredth time together outside, so I always enjoy jamming with you guys.
D
Yeah, same here, buddy. You crushed it. Thank you.
C
All right, thanks guys.
B
Mark. Thanks, John Kaplan. And thanks to everyone for listening to another episode of the Revenue Builders podcast.
A
Thanks for listening to today's episode. If you enjoy the content, please subscribe, rate and review the show to help us reach more people. This show is brought to you by Force Management, where we help companies improve sales performance, executing the growth strategy at the point of sale. Check out forestmanagement.com for more information.
Guest: Mark Roberge, Founding CRO of HubSpot
Hosts: John McMahon (“Mac”) & John Kaplan (“Cap”)
Episode Date: January 15, 2026
In this episode, the hosts sit down with Mark Roberge—founding CRO of HubSpot, Harvard senior lecturer, co-founder of Stage 2 Capital, and author of the forthcoming book The Science of Scaling. The discussion dives deeply into how revenue leaders, founders, and executives should approach scaling as a system, not an event—emphasizing the danger of premature scaling, the importance of data-driven decision-making, and the evolving process of defining Product-Market Fit and Ideal Customer Profile (ICP). Roberge shares both his research and first-hand experiences, providing listeners with a clear, actionable framework for sustainably growing a go-to-market engine that lasts.
Failure Patterns
"A lot of these boards, the time they choose [to scale] is when they get an infusion of capital and the pacing they choose is to look up the first two years of Snowflake and copy that. That's crazy, right? ...the failure rate in the ecosystem is unnecessarily high because of a lack of rigor and framework..."
— Mark Roberge [03:38]
Timestamps:
ICP as a Living, Data-Driven Entity
“Your ICP is who your salespeople go after and close. That's the market that you're in. And if you're not diligent...they're always shocked when I tell the story of Halligan in year three of HubSpot ripping up the contract from Meta that we sold them because they would have destroyed us, even though it would have doubled our revenue in one quarter.”
— Mark Roberge [08:24]
Timestamps:
Three Critical Phases:
Product-Market Fit is NOT Revenue
“The three of us can sell ice to Eskimos. Eskimos don't need ice. We can get a million dollars of revenue. It has nothing to do with revenue. It has everything to do with customer value creation.”
— Mark Roberge [12:44]
Leading Indicator of Retention (LIR):
“Our LIR is when the customer uses five or more features in the platform... We can pay the reps half the commission when the customer signs the contract, and half when the LIR is achieved. So we don't have to wait a year for retention, but we're forcing the reps to be LTV focused.”
— Mark Roberge [24:44]
Timestamps:
“...the number one rep at PTC right now would be a terrible first hire at product-market fit... We need someone who's like half account executive, half product manager...”
— Mark Roberge [15:08]
Timestamps:
Timestamps:
“The bottoms up is the sales capacity that you create, the demand gen capacity that you use to feed them...We have to factor in attrition, then we have to factor in ramp time...and the demand gen capacity.”
— Mark Roberge [46:33]
Timestamps:
Timestamps:
“AI can do it every hour on an infinite number of attributes... The ICP becomes dynamic and fluid. Is it too Star Trekkie?”
— Mark Roberge [41:21]
Timestamps:
“When you're measuring the LIR, you're only measuring on the current ICP... Nail it down now, separate your install base into ICP and not. … We can’t afford that churn. You just have to rip off the Band-Aid.”
— Mark Roberge [36:10–38:28]
Timestamps:
“We just need to talk about it more... It's crazy to think that when you find out someone’s a cancer survivor, you elevate that person…but if it’s life-threatening depression, unfortunately a lot of people have concerns. They're the same things.”
— Mark Roberge [56:54]
Timestamps:
This summary captures the essential frameworks, actionable insights, and frameworks from the episode, as well as the energy and candid tone of the hosts and guest. Whether you’re a founder, sales leader, or executive, Roberge’s ideas offer a robust guide for scaling sensibly and sustainably.