196. Your North Star Metric Is Wrong (and It’s Killing Your Retention) ft. Mark Roberge (Stage 2 Capital)

I don’t have product market fit, but I’m just gonna learn, iterate and fix the products so that they don’t churn. And I’m like, well, how will you know of when you have product market fit? I’m like, when they don’t churn. And I’m like, exactly. It’s not about half a million in revenue or a hundred customers or 500 inbound leads. It’s about retention, which basically says, like, you delivered the value you promised. And that’s rarely an answer.

Josh Schachter [Host]:
You’re listening to Unchurned, brought to you by the Gainsight Podcast network. Someone hands you $8 million Monday morning, you hire 50 reps. That was the plan. Mark Robert has spent 12 years watching companies die this way. The founding CRO of HubSpot, now a partner at Stage 2 Capital and Faculty at the Harvard Business School. His new book, the Science of Scaling, says, the signal that tells you when to scale isn’t sitting in your pipeline, it’s hiding in your retention data. And most founders are reading it wrong. I’m Josh Schachter.

Josh Schachter [Host]:
This is Unchurned. Subscribe to our substack@ Unchurned.Gainsight.com where we go deep on every episode. Like how one post sales team at cloudbeds built over 150 AI agents. That story and more at unchurned.gainsight.com hey, everybody, and welcome to this special edition of Unturned. I’m your host, Josh Schacter, senior vice president of strategy and go to market development at Gainsight. This is a special episode because I’m speaking to a very special thought leader in sales, venture capitalist, the founding CRO of HubSpot, Mr. Mark Roberge. Mark, thank you for joining.

Mark Roberge:
Hey, thanks, Josh. Thanks for having me on.

Josh Schachter [Host]:
I’m a big fan of your work because you are one of those VCs out there that actually knows what he’s talking about in a really, really, like, in a really awesome way. I mean, there’s lots of V, but your. Yours kind of carries a lot of weight, so.

Mark Roberge:
All right, let me asterisk it though, Josh. Appreciate it.

Josh Schachter [Host]:
My nose too brown right now.

Mark Roberge:
VC knows what they’re talking about. I think perhaps there are some situations where if you are an awesome investor, perhaps you do a little imposter syndrome in the boardroom when you’re advising people on setting up their first CS team. Okay, so like maybe from your side as an operator comes off that way. I just think overall like there should be far fewer operators turned VCs that are leading investments and far fewer finance only investor, only VCs that stand in the boardroom and advise people on operational strategy.

Josh Schachter [Host]:
Do VCs know what customer success is?

Mark Roberge:
Yes, because they saw it at another portfolio company that didn’t. That did it well. But I think oftentimes there’s a mistake around the inappropriate cut and paste that just because gainsight ran their CS team in this way doesn’t necessarily mean a PLG fintech AI first company in 2026 should do it the exact same way.

Josh Schachter [Host]:
Yeah, their pattern recognition is probably a little bit less than, than all those sales folks like yourself. Okay, so. So you’ve written a couple of books. The most recent, the Science of Scaling. I’m holding it up here for all those that are watching here on YouTube. Science of scaling Using Data to Decide when and How Fast to Scale Revenue by Marco Berge. It’s a wonderful read. It’s actually funny because you know, my startup was, was part of part of the Stage 2 capital program.

Josh Schachter [Host]:
Your V recognize a lot of this from the curriculum that you were educating us founders on. So it’s really cool now to have it in book form before we get into the actual content. And by the way, for those that are listening this episode, we’ll bring it back to CS and post sales when we can. But I wanted to have Mark on the show because he knows what he’s talking about. He’s got a lot of great energy, he’s got a lot of great material. It’s going to be a little bit more based on scaling on the earlier stage companies and his thoughts around go to Market fit and Product Market fit. So just a little caveat there, but stay tuned because I think we’re gonna get into some really valuable stuff before we get into that stuff. The proceeds, all the proceeds of your book are going to the McLean Hospital, which is a leader in mental healthcare.

Josh Schachter [Host]:
That’s cool. Mark, tell me a little bit about the origin of doing that.

Mark Roberge:
Yeah, I mean with every both books I wrote, one of my big motivators, as you’re kind of pushing the early mornings and the late nights is a cause and also as you do the, the cross country tour, it just wakes you up in the morning even more. And yeah, this mental health has always been a really important theme in my life and we still have a stigma associated with it. We’ve made a lot of ground in the last generation, but we still have a lot more. I have been a primary caregiver on multiple occasions. I have also been a patient, and I can stand here confidently and say that. And because I’ve been blessed with certain resume wins that society values, and not everyone has that luxury. You know, I think when we’re interviewing a candidate and find out that they survived cancer 10 years ago, we probably elevate our perception of them. But if we find out that they had a severe mental health setback, we may have some concerns.

Mark Roberge:
And it’s just they’re both diseases, oftentimes genetic, oftentimes outside of the control of the patient. And, you know, that was part of the drive. The other one is completely separate. As we enter in this AI era, many folks have pointed out to me the disconnect of how much capital, resources and talent are going to building AI and how few resources are going into helping society prepare for the impact. And I think, like, as a technical community, is this where.

Josh Schachter [Host]:
Is this where. Insert Universal Basic Income. Is this where we’re going, Mark?

Mark Roberge:
Yeah, it doesn’t. UBI doesn’t work because there’s not like a raw number we have in our head. It’s just how we have right now. The way we’re wired is the relativity to our neighbor. And so UBI doesn’t solve that. Yeah, but, like, it’s a little bit. It’s a little bit of that is like, as a tech community, we need to balance that better. We can’t delegate it to Washington.

Mark Roberge:
We can’t delegate it to the economists. They’re just not close enough to it. I think, like, just like in the Internet era, we will come out a more evolved society, but there will be some scars along the way, and there still are. The scars are, like, piling up to be even worse in the AI era. And we just need to be proactive as a tech community. And everyone just needs to do their little thing. And right now, my little thing is this work and the donation, because I think mental health will be a piece of it. And I’ll do more later.

Mark Roberge:
And I just hope everyone else can find a way to balance the building with the thinking about society.

Josh Schachter [Host]:
Thank you. Thank you for standing up and setting that example. And Mark also wanted to take the time on this episode to announce that he’s writing a check for $10 million to the institute as well. Right.

Mark Roberge:
Via the book sales.

Josh Schachter [Host]:
Via the book sales. Via the book sales, yes. Via the first half a million sold.

Mark Roberge:
You’re awesome, man. Yeah, there you go.

Josh Schachter [Host]:
Oh no, you are. Okay, let’s talk about the book. Let’s talk about helping. Helping. And I’m going to use the word startups loosely because like even gainsight you could consider is like startup. Any tech company these days wants to be a startup. And yeah.

Mark Roberge:
And to your point and like the connection, like as you think about this, this work is relevant whether you’re two people in a garage bringing a product to a market or whether you’re IBM bringing a product to market this quarter. Like I think the same issues occur that cause unnecessary failure of a new product in the revenue generation. It’s the exact same playbook. So we can talk about it in either way and hopefully that helps the audience connect it regardless of their current position.

Josh Schachter [Host]:
I didn’t get very far into the book by the time I read the sentence here. Like I read the book but like I got into immediately you have this thing that’s that kind of as a, I would say stop me in my tracks that we scale haphazardly, not scientifically. Unpack that a little bit for me. Sure.

Mark Roberge:
I mean just think about and just like two backdrops. One is I didn’t intend to write this book. I left HubSpot 12 years ago after the IPO after a 9 year run. I was invited to the faculty at Harvard Business School and invited to countless number of boards as an investor or independent. And that was such a joy for five or six years before the founding of stage two. To have all that pattern recognition of helping these young founders both in the boardroom and the classroom and now in the VC suite. Hundreds, thousands and blessed with this 12 year journey of pattern recognition around why some become IPO and some fail. And those reflections led back to the decision on when to scale and how fast.

Mark Roberge:
And so that’s what inspired the work is like holy cow. How many classes in college are there that teach us how to account for and revenue our and recognize our revenue through Pure Accounting 101 rigorous frameworks. And how many classes are there that teach us the frameworks on scaling doesn’t exist. No wonder most of my reflections on unnecessary failure came back to this choice. Because right now if I could just generalize the playbook, it’s like someone gave me $8 million and I just hired 50 reps the next week and that was my plan. Yeah, this is kind of embarrassing.

Josh Schachter [Host]:
There’s no plan. There’s no plan. So you lecture at Harvard Business School. You have served as A senior advisor to Boston Consulting to BCG, you know, advising Fortune 100 companies on what they’re doing along these lines. What have you seen some of those larger companies are actually getting it right when it comes to scaling?

Mark Roberge:
Yeah, I would say probably the, the biggest success I got was from a, a gentleman that was an LP at Stage two that did this at Microsoft with his pattern that he called 55500. And what that means is the number of reps that he adds into it as he achieves various milestones which like parallel pretty closely a big company implementation of the science of scaling methodology. And so usually if like we talk about this in a big company framework, what happens is big companies are going to roll out a new product. They’re overly confident that the first version of the product will just hit and in parallel we’ll develop because they’re a

Josh Schachter [Host]:
big company, they’ve had success before. And so that, that’s a pretty good.

Mark Roberge:
We know our market, we know what they need. It’s not hard to build it right. So they’ll build it and in parallel update the marketing material, website, whatever and in parallel train all the salespeople on it. And it falls flat for a variety of reasons. All the way from the product never had product market fit and the company never figured out how to acquire the customers and the company never figured out how to onboard them in your arena and extract. Help the customer extract the value. That is just a learning mechanism.

Josh Schachter [Host]:
Yeah.

Mark Roberge:
And that’s part of what the product market fit then go to market fit then scale up Basic methodology of the science of scaling book outlines. And that’s essentially what this gentleman at Microsoft did very well is you feel pressure when you’re at like when you’re at Microsoft. Like yeah, we’re going to bring this new product to market. But I’m going to have, I’m only going to, I need like a quarter or two to like get it out, have some design partners and they’re like, we’re at Microsoft. Unless this is a trillion dollar revenue, it doesn’t move the needle. So that’s what he does. He’s like, I’m doing 550, 500. I’m going to have five reps on this team until we have product market fit.

Mark Roberge:
And we’re going to use Mark’s model of defining product market fit in the long term by retention and the short term on the lead indicator retention which is P percent customers do E event every tee time and we can unpack that. And once I have that and it Might take a week, it might take a month, it might take two quarters, I don’t know. And then once I have that, I’m going to work on Go to Market Fit, which is I know I can create value with this product. I just need to know that I can acquire the customers and onboard and serve the customers profitably. And I’m going to use 50 reps during that journey and implement medic and figure out design partners and how to like cross sell it and what the comp on the reps is and what the right price model is, et cetera. And then once I have that and, and call and call it success on Go to Market Fit, then I’m going to grow and I’m have 500 reps on this and you’ll have your trillion dollar business line. So hopefully that like helps you see the pothole.

Josh Schachter [Host]:
Does that fly? Does that, does that fly when you’re presenting that to, to the gm Definitely at Microsoft or even definitely helps everyone

Mark Roberge:
align because like what’s happening is.

Josh Schachter [Host]:
No, I listen, I agree that it’s the right thing to do. It takes a lot of strength and courage and execution to be able to present that I think to a large company.

Mark Roberge:
No, yeah, it’s just like it helps connect it with the framework because otherwise it’s this debate of like the product leader being like I just can’t deliver it that fast and know it’s going to work and the sales leader being like I just can’t train my team that well enough and still hit the short term quarterly numbers. And, and also the, the GM being like it’s too bad, right? Like the street needs it, right. And so this creates the dialogue because like the GM hears this and be like oh great, they’re going to print, they’re going to find product market fit in one week, go to market fit in two weeks and I’ll have the numbers. But the GM agrees that they were not going to go to 50 and then 500 unless these things occur. And of course Josh, like there’s some like blurriness to this, right? Like if they get 91% of the way there, we’ll call success and go to the next stage and continue to measure the instrumentation on the prior stage. But at least this gives us like a narrative and it’s the same thing in an early stage board is the VC saying I need a 5x this year and the CEOs like do it. And the CRO is like no way. Like there’s I, I would need to double close rates and Triple demand gen in two days.

Mark Roberge:
It’s just not going to happen. And this gives you a narrative to, like, help see the path to it, but do it in a way that’s not going to cause unnecessary bankruptcy.

Josh Schachter [Host]:
Yeah, yeah, no, I get that. I get that. So in your first chapter or in the introduction, you talk about haphazard scaling. In the, in the first chapter, you ask, you know, title of the chapter is is product marketing? Or, excuse me, is product market fit a feeling, a vibe? Now, I’ve always described it, quite honestly. It’s like, oh, you just kind of feel it. You know, your customers are, you know, dare I say, happy. They’re, you know, sales are seemingly going smoothly. Marc Andreessen, you know, he actually is a little bit fuzzy in his own definition of product market fit cores.

Josh Schachter [Host]:
Correct me on that. Is product market fit a feeling?

Mark Roberge:
Right. I feel like if you’re going to just use it as, like, it’s a good company, that’s fine. But I just think it, like, really diminishes the utilization of the term product market fit, because I do think it’s a great answer to the first step of preparing for massive scale is to have product market fit. And it, you know, it was invented by, I think, Eric Reese and a lean startup 20 years ago, which had an enormous contribution on the startup ecosystem, moving from building stuff in labs and having techs without homes to building products hand in hand with customers in an agile manner. That was such a breakthrough for us, totally. But it fell short. Whether I don’t think it was Eric’s fault, I think it was the ecosystem’s fault. For us to universally agree on what product market fit is from a quantitative definition.

Mark Roberge:
And I think if you’re going to use product market fit as step one milestone to prepare for scale, calling it a feeling is similar to calling profit a feeling. Like it shouldn’t be arguable. And when I do double click and challenge people to quantify it, most of the answers I get are half a million in revenue, a hundred customers, or a thousand inbound leads. And you can define it however you want. I just think all those, all three of those are dangerous because you can have all three of those and yet everyone’s churning off your product. And so most product leaders and founders would be like, yeah, I’d agree with that. If I had half a million in revenue or a hundred customers or a thousand inbound leads every month and everybody using my product churned, I’d agree. I don’t have product market fit, but I’m Just gonna learn, iterate and fix the products so that they don’t churn.

Mark Roberge:
And I’m like, well, how will you know when you have product market fit? I’m like, when they don’t churn. And I’m like, exactly. It’s not about half a million in revenue or a hundred customers or 500 inbound leads. It’s about retention, which basically says, like, you delivered the value you promised. Yeah, and that’s rarely an answer.

Josh Schachter [Host]:
Tell me about the leading indicator of retention. When you were lecturing when I was part of your cohort at Stage three Capital, it was something that really was awakening to me. I’d love for you to share it with our listeners too.

Mark Roberge:
Yeah. So, like, now that we’ve got that intuition of like, okay, I believe that I do agree there are certain things I need to have before I start adding 50 reps a month. And I do agree that product market fit’s probably a good thing to have. And I do agree that it’s more about retention than it is customer acquisition. But when I acquire 20 customers this month, I won’t know the retention for at least a year. I can’t just sit around and wait.

Josh Schachter [Host]:
Right.

Mark Roberge:
So hence the invention of the leading indicator of retention, which is something we can measure in the first month of a customer’s tenure with you and your product that if it occurs, they’re very likely to stay with you forever. And if it doesn’t, they’re very likely to churn. Hence the leading indicator of retention. And I go so far to simplify it down to three variables. P percent of customers do E event every T time. Right. So Slack. 70% of customers send 2000T messages every month.

Mark Roberge:
Harvey, 70% of lawyers process one document every day. What’s the gain? Gain size. As the master of this, trying to think like, is it like product breadth usage? What do you.

Josh Schachter [Host]:
Well, we have a feature called Journey orchestration. Journey Orchestrator, Jo, we call it internally. And we know that the retention when folks use that is, you know, goes through the roof. Yeah.

Mark Roberge:
85% of customers exhibit Jo every month.

Josh Schachter [Host]:
Perfect. Yeah.

Mark Roberge:
Right. So. So that would be the lead indicator. That’s the first slide I want to see. For every pre seed startup, that’s the first slide I want to see. If I’m coaching IBM, I’m bringing a new product to, it tells me whether or not the product’s actually delivering the value.

Josh Schachter [Host]:
Isn’t the canonical example like Mark Zuckerberg and Facebook, if you get seven friends and you were like, totally, there’s One

Mark Roberge:
there’s one deficiency to it. But yeah, that’s the inspiration. As we entered the Internet era, the consumer tech arena was probably the driver thought leadership on these types of lead. You know start with value where the famous Facebook like get seven friends in a week and then eventually the PLG community with say Dropbox early on eventually notion and Slack adopted this as like an aha moment movements very paralleled stuff. It’s just, it also applies whether you’re doing tractors in Brazil or you know, you know, marketing consultancies in Europe. The. The only deficiency in a B2B context and what we’re talking about here with the get seven friends in a week is it’s not ongoing, it’s just a one time moment which does. It wouldn’t help you protect against someone trying it and then stopping.

Mark Roberge:
You have to detect that.

Josh Schachter [Host]:
Got it. I see. So in the customer success world, every

Mark Roberge:
tee time, not by tea time. So not like Slack sends 2,000 tea messages within a month. It’s Slack sends 2,000 team messages every month.

Josh Schachter [Host]:
There’s something there that’s subtle but profound because in CS we talk about time to value as being such an important metric. I mean if you just onboard and your folks and get them quick time to value to demonstrate that value like your goal in the path is set. I like this better because it’s saying well you have to give them that recurring time to value.

Mark Roberge:
You have to continue and they’re not mutually exclusive like another underlying. If you get into the nuances of LIR leading the can of retention design, one other lens of it is there is this maturity model that you can go through where if you are starting out you if you want to just spoon feed it and baby step it, start out with a setup one like X percent of customers just have it running, that’s it. And then most people do, I mean with security software that’s all you need, just get it running. And then most people graduate to what I call an engagement lir which is the ones we’ve been talking about. You know, send 2,000 team messages, process one document. Some people do graduate to an outcomes ROI one which in you know, in Slack it would probably be an Employee Productivity 1. In HubSpot it would be a lead lift every month. In Gainsight it would probably be ultimately a retention outcome harder to get to and have attribution.

Mark Roberge:
But that’s an option.

Josh Schachter [Host]:
Well that’s also the agentic day. Right? That’s the, that’s where we’re going with

Mark Roberge:
these, what we hope for.

Josh Schachter [Host]:
Yeah.

Mark Roberge:
What we hope for.

Josh Schachter [Host]:
Well, let’s actually get into the lira lir design a little bit, the setup of it. So yeah, you know, when, when I was running update AI and conversational call intelligence for customer success and, and I forget we did use lir and I, but I forget now what the, what the.

Mark Roberge:
Well, let’s, let’s go back and work on it.

Josh Schachter [Host]:
Yeah, let’s make it up.

Mark Roberge:
What’s the value it creates? What was the value it created again?

Josh Schachter [Host]:
I mean, at the core you can get into some of the other stuff, but at the core it was meeting notes, you know, really well synthesized meeting notes with, with insights for, you know, about your customers. So, you know, we would need to have X number of users using, you know, generating their meeting notes and, and consuming those meeting notes per month.

Mark Roberge:
Perfect. Yeah, so we unpacked that into this complexity for a second. Because you’re asking about like, what are the best practices on build, on building this out, right?

Josh Schachter [Host]:
Yeah.

Mark Roberge:
Okay,

Josh Schachter [Host]:
you’re probably gonna go there. But what I’m asking is like, it’s almost a little bit of like the cold start problem of, okay, how do I know if it’s like 70%, 80% of my users? How do I know if it’s, if it’s four times a month or if it’s ten times a month? You know, I’m sure we can. If we’re a large organization, we get our quant people on it. Right. But like, like, how do I get to that?

Mark Roberge:
Okay, so like zoom in. Way out. Yeah, don’t worry about it too much in the beginning because you have to remember that 95% of your founder peers are like setting the first north star of the company on a million in revenue or customers. And that’s just really bad. As long as you are rooting it on something regarding value, you’re just so far ahead of everybody. So just like, just for, for starters, lock yourselves in the room with the decision makers, give yourself two hours with a stopwatch, and then at the end, as a founder, just frigging make the call. All right? Just don’t, don’t spend too much time on it. If you aren’t on value, you’re better off.

Mark Roberge:
Now in the medium term, you’ll be able to know because a year from now, you’ll look at those 20 customers you acquired in Q1, five of which churned, 15 of which stuck around, and you can analyze whether the ones that stuck around exhibited massive lir adoption and the ones that churned did not. Statistically. And if you found that, then you have great confidence that you’re on the right path. If you don’t, if they’re the same, then you probably have a year’s worth of user logs to run. 50 permutations with AI to figure out what actually does matter.

Josh Schachter [Host]:
You’re basically running a regression analysis.

Mark Roberge:
Yeah. So maybe with Slack it’s like we’ve been running all year thinking it was 2000 team messages. Turns out it’s not. It actually is 10 users and we just use the user logs at that moment. So in the beginning, like just pick something correlated value, knowing that in a year you’ll be able to run the analysis to tune it. Now the P percent thing, like should we do, should we call success at product market fit when half our customers hit it? When 80% customer, when all of them. That depends on the blitz scale risk of your category. Because like if you are competing in say like you’re, you’re in a customer success, I woo huge bliss scale risk.

Mark Roberge:
There’s 15 companies that raised 100 million. Like you can’t wait till 95% of customers hit the LIR. You got to like, I wouldn’t do it if 20% hit it. But maybe we say 40 or 50% and let’s go and kind of build the plane while we’re, while we’re, you know, while we’re scaling it and just continue. We’re going to continue to measure, I mean HubSpot does 2 billion a year right now and still measures it. So we’re not going to stop measuring it because we’ll, we’ll lose product market fit at some point and I’ll help us know. But I’m going to like graduate sooner versus one of our portfolio companies, Arrow Cloud, is bringing airports into the cloud era. Turns out they haven’t gone yet.

Mark Roberge:
And their only competition, these two incumbents that are worth $20 billion, founded in the 70s, all the contracts are five years. There’s no competition and the tolerance for bugs is really low. Like if your CS software has a bug, it’s not great, but people aren’t going to die.

Josh Schachter [Host]:
Right.

Mark Roberge:
But if there’s a bug in your Airport software, that’s not good. So they’re going to choose a P that’s pretty high.

Josh Schachter [Host]:
Yeah, yeah.

Mark Roberge:
So it depends on the blitz scale risk of your category.

Josh Schachter [Host]:
I wonder if you can take this framework and apply it to any function really that you’re leading. I think, you know, I lead our customer marketing against. I’m thinking, can I do this for entering in new advocates into our marketing program to our customer. Marketing program.

Mark Roberge:
That’s cool.

Josh Schachter [Host]:
Cool thought.

Mark Roberge:
I never thought about that.

Josh Schachter [Host]:
Yeah, right.

Mark Roberge:
Something’s cool and rigorous when you can find broad applicability at the same time. It’s also the downfall of academia. Sometimes it’s extracted so far that it like lose relevance to everyone.

Josh Schachter [Host]:
Yeah, it’s fundamental.

Mark Roberge:
Fired from my faculty post, but let’s bleep that one out. Joe.

Josh Schachter [Host]:
Yeah, no, that’s okay. No, I think people will enjoy it. Yeah, no. Okay, so, all right, so. So then let’s assume that you’ve got your, your product market fit now. You’ve got your, your go to market fit and this is the next, you know, section of the book.

Mark Roberge:
Yeah, let’s unpack the why there. Okay, so why not boot do both at once? And just to be clear, product market fit, all we’ve essentially proven is that if we acquire another 20 customers this quarter, most of them are going to see val. Most of them will see value that we promised them. That’s what we’ve proven. We haven’t proven that we can acquire onboard and serve them profitably. Okay, Now I like to sequence them because I love Paul Graham, founder of Y Combinator’s advice of do unscalable things early. I think it’s brilliant. I think a lot of Y Combinator companies have followed that and it’s a beautiful form of entrepreneurship.

Mark Roberge:
And let me bring that to life within the science of scaling framework using one of the companies that I advised early and a good buddy, David Cancel at Drift, who ultimately sold his company for a billion dollars in the first sequence of Drift. When they were pursuing product market fit, David was flying to customer headquarters to onboard them on Drift. When the customer’s paying him $50 a month. That is not scalable, that is not profitable. But it is a beautiful, beautiful exhibition of Paul Graham’s do unscalable things early because even the best of the best, from jobs to musk to cancel, cannot guarantee product market fit and sometimes don’t find it because you’re an inventor and you don’t always invent something that creates the value that you hope. So throw all resources and the kitchen sink at that problem. Do not worry about optimizing price. Go to market demand gen, sales rep comp.

Mark Roberge:
All of that doesn’t matter at this point. It’s all about creating customer value. But once we have deemed success, as we just discussed with the lir, we can’t scale until we know we can do that stuff profitably. We need a couple reps. We need a sales process. We need a scalable demand gen channel. We need to get the quotas right, we need to get the comp right, we need to get the pricing model right. Don’t work on any of that during product market fit, but once you have it, we need that.

Mark Roberge:
Okay. And it might take a week. It’s at my month. It might take two quarters, who knows? But it is measured by unit economics. Right. So when we say by the way,

Josh Schachter [Host]:
by the way, Mark, as you’re talking about this, is this. If I’m raising, let’s say I’m a founder and I’m raising my, my, my, my pre. Seed or seed and you can correct me on the proper stage, is this the way that I’m structuring my pitch deck in some way? It’s, hey, we’ve got this idea. We think it can be a deck of corn. This is, we’re raising, this is the first phase of what we’re doing. We’re finding our product market fit. You know, it may not be as extreme as David flying out for a $50 a month renewal, but, you know, this is where we’re going to start the journey. Is that, is that a compelling narrative as a founder?

Mark Roberge:
Yeah, it depends on like, for the most part, that’s what I’d prefer. But like, you know, let me try to play a devil’s advocate. If I am Brett Taylor, CEO of Sierra, chairman of the board of OpenAI, and I sell Sierra for, you know, $50 billion to OpenAI, and I go to the beach for six months, spend some time with the kids and start another company. Maybe he doesn’t do that. He does it all at once.

Josh Schachter [Host]:
Yeah. Because someone hands me about the average man here or woman.

Mark Roberge:
Yeah, yeah. I mean, for the most part you want to do that because all the risks, you know, and, and like, I want to just make sure we understand this. This is not about going slower or waiting. This is about going at the right pace because half the people I meet wait too long and go too slow. I have a number of examples in our portfolio. And half the people I meet go too early and too fast and it creates unnecessary bankruptcy. Right. So, yes, I would agree with you.

Mark Roberge:
Josh is like, it’s good to have the board align with this and there’s just different implementations of this of like, how much, how quickly we go at that lower p to take a risk because we have to, because we have blitzscale pressure. And unfortunately, a lot of VCs today think that first movers win all the time. And history doesn’t tell that and think that blitzscaling is the only way to create a huge win. But look at Klaviyo, look at Zoom Info, look at like a lot of companies, Dropbox that didn’t do that and ultimately won. So it’s not like the only way. In fact like it leads to more failure than it does success. But that’s the thing. VCs don’t care because they need one into every 10 or one every 20 to hit.

Josh Schachter [Host]:
Right, Right. Okay, so, so we’re on to Go to Market Fit. Now finding that we’ve addressed that, that product Market Fit, if you want to do it well, is not really a vibe. And so although these days maybe it can be a vibe coded application. But how do you know when you have Go to Market Fit?

Mark Roberge:
Yeah, so that’s where we were is like we understand qualitatively that all this is doing at the Go to Market Fit phase is we know that we will create value for the customers. We just need to know that we can acquire them and serve them profitably so we have a business. And again that’s what we said. This is where the price models, pricing model makes sense. This is where we need to know how much are we paying our reps, how the quotas work. We need to know that we have a scalable demand gen channel, what the CAC is going to be. We need a sales process, all that kind of stuff. Right.

Mark Roberge:
So again, it could take a week, a month, a quarter, two quarters, whatever. And we, we’re going to measure it by unit economics that like profitability, we’re not going to use GAAP accounting profitability because that has a lot of like noise in it around like our office space and all these things that aren’t necessarily going to grow with customers. But unit economics nicely isolates the costs to the revenue and the customers. So pick your poison. If you’re like Economics Microeconomics 101, it’s marginal revenue is greater than marginal cost in SaaS these days. We’ve redefined that as payback periods and LTB, CACs, whatever. And you can extract those outputs algebraically back to the key inputs of like cost per lead and conversion and sales cycle and CAC and churn and all that kind of stuff. All right, so, so that’s all we’ve done is I can summarize, given that backdrop is we are ready to scale when we have product Market Fit followed by go to Market fit.

Mark Roberge:
We do those sequentially because it’s really hard. There’s there’s very few investors or entrepreneurs in the world that find product market fit on all of their ideas. So throw all your resources at it in the beginning and then graduate to go to market fit if you find it. Furthermore, if you do them at the same time, you run the risk of optimizing your go to market on the run product market combination and remove the opportunity to pivot and be agile. We are going to measure product market fit and go to market fit by retention and unit economics in the long term and the lead indicators of retention and the lead indicators of unit economics in the short term so that we can get a view every single week where we are on this journey. That’s when we’re ready to scale.

Josh Schachter [Host]:
And going back to your idea of you don’t have to slow down for all of this. I mean you’re becoming more hyper focused on each phase. So you theoretically you could actually could hasten the process for you because you’re not distracted. I mean, but you need to go quickly through each phase, even if it is in isolation.

Mark Roberge:
Yeah, I mean you talked about like the speech you saw in our LP base on Monday of the CRO of lovable Ryan Meadows. He’s one of my early hires at HubSpot and he’s an LP at Stage two. And like they’re, they’re a massive scale right now with very few reps. And they’re measuring this now because it’s like, it’s just like why you’d measure burn. Like I think we’d all agree that there’s a burn number that would raise a red flag that we might need to like slow down for a week and fix this thing. I think we might agree that burning $2 billion this month probably isn’t advisable. And so there is a similar metric on the LIR and the lead indicator unit economics that if we see it we should go faster and if we see it in the red zone, we should probably stop for a week and fix it.

Josh Schachter [Host]:
You know, the LIR I can very easily see translate to a larger company to a level lovable that’s still early but growing at light speed, to a gain site that still has a startup mindset but is more mature, PE backed and all of our PE backed peers go to market that, that science feels maybe a little bit more difficult to replicate. If you’re a larger company, let’s say you are a gainsayed or a private equity portco there, there, I don’t know. That’s just kind of like my, my suspicion. It’s harder.

Mark Roberge:
Yeah, I can bring it to life for you.

Josh Schachter [Host]:
Yeah, bring it to life.

Mark Roberge:
It’s not. I, I don’t know if it is that hard because essentially most people are doing it because they are reporting reporting the unit economics at their board meeting. Okay so let’s just, let’s just assume we’re going to be a. We’re going to do payback period like 12 months or less and NDR, we want NDR over a hundred percent. The I like that’s kind of probably my favorite is using those two together. Okay so we’re reporting those already. The only like up leveling I would do is can we just algebraically extract payback period back to number of new opportunities per rep per month, the close rate, the sales cycle, the cost to generate the opportunity, the cost to close the opportunity and the acv. It’s really just basic algebra.

Mark Roberge:
And can we just like at the board meeting look at the actual unit economics outputs of Q1 which is looking at the in the mirror as well as those inputs for Q2 to see where unit economics are likely to play out? Right. So you’re just extracting things back to their first principles inputs and essentially managing your business one or two quarters ahead of your peers.

Josh Schachter [Host]:
Yeah. Okay, that makes sense. The non fixed cost is basically you’re.

Mark Roberge:
You’re trying to. Exactly.

Josh Schachter [Host]:
To get into those. Yeah. Going back to early stage, when’s the right time for. I’m assuming that you’re, you’re an advocate of founder led sales at earlier stage companies. When’s the right time to. Or what are the signals for you to know when it’s time to move on beyond founder led to hire somebody?

Mark Roberge:
It does vary quite a bit and the, the variance is largely correlated to the abilities of the founder. So obviously if I have a founder who comes from sales it’s going to be delayed quite a bit and I want the founder to stay close. If I have the best AI developer in the world, I want to bring on someone right away because I do not want to use that person spending half their week selling. Right. So that, that’s usually the correlation shooting right down the middle. I do like the founders to try to get to product market fit on their own if they are going to bring on a seller at that point. I talk about this a lot in the book of like your design of the go to market system depends on what stage you’re at. So if you’re going to bring a seller in at the pursuit of product market fit phase, your most important output of that role is not the revenue, it’s the feedback.

Mark Roberge:
So just make sure that you hire someone. Like the, the number one rep at workday right now would be the worst hire. Like, they’re just like, they’re just a machine. Like, you need to find someone that’s like half PM, half AE. Like they’re going to talk to 20 customers and then, like, sense. Make the

Josh Schachter [Host]:
third CFCs too.

Mark Roberge:
The whole. It’s a whole athlete.

Josh Schachter [Host]:
Yeah, yeah, it’s a whole good, good,

Mark Roberge:
good pick, you know, good fix. Like it’s a whole athlete, right? Who, who’s not going to be your top rep in the growth phase, right? So that’s who you’ll hire. And then once you move to go to Market Fit, you need to build a process. So again, if you’re a founder and you’ve done it, do it and hire two reps and try to teach them. But if you haven’t done that before, if you need to build like GPTT or ga, you know, or medic or whatever, you need to find someone that can do that. So that’s usually the timings of it.

Josh Schachter [Host]:
We haven’t spoken too much about AI. You wrote this book, what probably like you started probably writing it like, what, 15 months ago?

Mark Roberge:
Two years ago. Two years ago, writing it.

Josh Schachter [Host]:
You already had all the content, dude. What.

Mark Roberge:
I know what took you. So I had to. So what happened was I signed Wiley had write a first refusal with me from the prior book. So I signed with them in May of 2024 and then I started writing it and it was due in May of 2025. And I just had some personal stuff come up that I had to, like, do some dad stuff that I had. I kind of scrambled to get it by that May 2025. And I started putting the finer suit and touches on it then. And you’re absolutely right.

Mark Roberge:
I’m like, damn it. I have to, like, talk about how AI is going to influence this stuff. So I wrote this appendix about it. And fortunately it is very much coming to light in that direction. And it basically just talks about, like, first off, you know, as we build AI, it’s really important that we have first principles in mind. And I do find a lot of AI builders today are younger. You know, they’re like in their 20s. And they’re kind of building the models based on how sales is running, like at this moment without reflecting, like how it has evolved and what the first principles are, because that’s the real opportunity.

Mark Roberge:
Just the stuff we’re talking about today, like, I just don’t think when we have agents buying from agents that it’s going to change the definition of product market fit being about value creation and the lir. Yeah, like, I just think that’s like similar to how you measure profits. Right. So just I hope that like some of the models in the book will create. Will. Will serve as a, a blueprint for some of the future AI agents and model design. And then I also talk about four phases of the AI revolution and go to market which are essentially like the re. The extraction of human involvement further and further from today’s point of work.

Mark Roberge:
So the first phase is just completely eliminating the admin work and just increasing selling time, which I can kind of circle back to. And we’re kind of, I think we’re having that right now. And then the second wave is the agents are sellers and then a third age, the agents are buyers. And I think in the fourth age you really start to lose the functional design of orgs, you know, product engineering, finance, sales, marketing, which are designed that way because of human limitations. And you start to see this as like, it’s more like everyone’s a GM of their own business line and they’re kind of athletes. You’re starting to see that come to life a little bit.

Josh Schachter [Host]:
What are you telling your, your, your Harvard Business School students?

Mark Roberge:
Yeah, there’s, there’s a couple of things. First off, like at the board level, I show up and they’re like, oh, dude, our like sales team is like, so AI native now. It’s crazy. And I’m like, how do you know? So I want, I’m putting pressure on people to quantify that. And I, I’m, I, I believe the quantification is selling time and rep to manager ratio. I think that’s the opportunity for us in 2026 is selling time is the percentage of a week that a rep spends with a customer or prospect. It has historically been best in class in the industry, around 25 to 30% because we spend so much time prepping for calls and generating meetings and in pipeline reviews and updating CRMs. And a lot of that can be streamlined by AI.

Mark Roberge:
And so I think we’re starting to see evidence that that can get to like 75% for the best teams. I would like to see more teams measure it and get to 75%. And then on the other side, the traditional rep to manager ratio in the industry has been about 7 to 1. And managers hire, coach, hold people accountable, it’s a process, et cetera. AI can do most of that. Especially the coaching today better than most managers. So I do think we have an opportunity to drive the rep to manager ratio to at least double and actually accelerate enablement and productivity. So I think those are the two inputs that I want to look at is you drove selling time to 75% you drove your rep to manager ratio to 15 to 1 and PPR for your tenured reps doubled.

Mark Roberge:
So if Josh and mary were for 7 years producing 250k a quarter once you implement this they’re producing at least 500 maybe 750 a quarter. That’s what AI enabled in 2026 means in sales.

Josh Schachter [Host]:
Lots of stuff to review there. We’ll leave it at that. Mark as always I learned so much from you. Such a pleasure to have you on the program again. Thank you so much and best of luck with the book it is the Science of Scaling by Mark Roberge. Again thanks for being on the program Mark.

Mark Roberge:
Yeah thanks for the platform Josh. Appreciate it. Always good to see you.