I feel so privileged and honored to be able to introduce our next speaker, and it's Mark Roberge. Mark is the co-founder of Stage 2 Capital, the industry's leading go-to-market limited partner venture capital firm, which I'm sure Mark will talk a little bit more about. He's also a senior lecturer at Harvard Business School and was the founding CRO at HubSpot and really kind of revolutionized how SaaS revenue leaders were using metrics to inform their decision-making. So with that, I'm going to give the stage to Mark. Thank you so much, Mark, for being the speaker today. Yeah, you bet, Ray. Thank you for having me. Hey, everyone. Excited to spend 25 minutes or so together. So for today's speech, The Science of Scaling, I need to go back about 10 years. We had completed the IPO at HubSpot, and I had been there since day one, so it was really tiring. It was exciting, but tiring, and I needed a break. And so I set out to take one. Serendipitously, I was called and approached by Harvard Business School right around that time to join the faculty, and they were looking for a build-out of sales curriculum for the MBA program. And I thought, wow, what a remarkable way to rest more, obviously, than running a global sales team, but not having to be on a plane all the time, all that kind of stuff, but also being able to give back, being able to give back to the ecosystem, to have some time to reflect and think, and do it in a structured way at a very honorable place to do it. And so I did that for about five years before we started Stage 2 Capital. I can tell you about that a little later. Not only did I teach some brilliant people, but I also spent – I chose a different company, a different startup every quarter to help. I joined their advisory board or their board, or I invested in them, whatever it was, and I spent about a day a week throughout that quarter. And it was largely around when they were professionalizing their first sales organization. And I did that over five years and a couple dozen folks, and as is the case in startups, some of them failed, some of them flatlined, some of them had great success, companies like Asana and Drift and VTS and Salsify and others. And it was the first time in my career where I wasn't 80 hours a week on one company, but I was between my students and these advisory gigs, I was touching 100 startups a year. And it gave me a lot of sort of data points to reflect on where people stalled out, where failure originated. And those reflections kept coming back to two strategic questions that I found boards, founders, senior leadership teams were not being as thoughtful as they could be around. They were making bad calls. And those were when to scale revenue and how fast. If you think about that, when should we scale? I know a lot of you are in the room, have been at startups, you've been asked to come in and run revenue teams, you are founders. Such a critical question is when is the moment when you're going to go from, hey, we're building our product, we're building our minimal viable product, we're recruiting design partners, we have initial customers, we're doing founder selling, we're trying to prove that it works. And now we're ready, meaning we're ready to commit to an annual revenue line. We're going to go from half a million to 2 million next year or something. And we're ready to add salespeople. We're going to add salespeople now. That's a key decision. And I think there's some pretty soft answers out there. Almost everybody says product-market fit and I think that's a great answer. From my research, product-market fit originates back to Eric Ries in the early part of the century, Steve Case and like Eric's work on the lean startup and that was a groundbreaking moment for entrepreneurship. But what's funny is even though that's such an important term and breakout framework for us, when you ask 100 entrepreneurs what product-market fit is, you get 100 different answers. So that's a little silly that the answer to this critical question is vaguely defined. I would say half the people talk about product-market fit as a customer acquisition or revenue acquisition milestone like a million in revenue. And I'm sorry for being blunt, but I fiercely disagree. I mean, I think someone could sell ice to Eskimos and get 2 million in revenue because they're great at sales. It doesn't mean Eskimos need ice. It just means you have market message fit or you're a shark of a seller. But the better definitions of product-market fit have a lot more to do with creating value consistently for your customers. Those are where the definitions, the great definitions I think need to reside. Now you've got a really good one out there with Sean Ellis who says, if you were to survey your customers, how disappointed would you be if your product didn't exist? Not disappointed, disappointed, or very disappointed. And he wants to see 40% or more of very disappointed. Now we're talking. It's quantified, there's a clear line, and it's rooted in at least the value perception of the customer. I just think I'm not quite there with it because surveys are just known for false positives. There's a thing in product management called the mom test where people don't like to hurt people's feelings and they'll want to tell them what they want to hear, even though they might not honestly be feeling that. So for me, when I want to say like, when are you ready to scale? I think the best, yeah, product-market fit is great, but I like to define product-market fit. I like to root it in customer retention because I don't think there's a lot of like mom test in there. I don't think there's a lot of people like, yeah, you know, I don't really like your product, but I'll renew anyway. I don't think there's a lot of that happening. I think that's like a defining moment that's extremely quantifiable is someone bought your product and they decided to almost like re-buy it, like renew it, whether it's a month later or a year later or whatever. Now the problem with that is it's such a lagging indicator. I don't really honestly advise people to use annual contracts in the early part of a startup, a lot of people do, and even if you don't, like many folks will take many, many quarters to decide whether or not they're renewing or they're going to cancel. And that could be a significant lagging indicator to what we're really trying to get at is, is your product delivering on the value that you promised and you envisioned? And furthermore, like we just don't have that time in startup, in the startup world. We don't have like six months, nine months, a year to wait. We need to know as early as possible. And so I advocate the creation of a leading indicator of retention. This is a metric that if it, if this behavior, this customer action or behavior were observed in say the first month of a customer's tenure with you, that they're going to be with you forever. And if it isn't, they're probably going to cancel like pretty quickly. So I feel like that's an important number. And that's like, it's not something I can say like 80% of the time, this is about like daily active usage. I can't, there's no universal answer on these. And I find this to be like an opportunity, a moment for true creativity in entrepreneurship. Something that founders actually love is to not have the exact blueprint laid out for you. I'll try to codify it a little bit for you where it's P% of customers do e-event every t time. So it's kind of tries to boil this definition down to three metrics, three variables. And I can give you some industry examples that are relatively well documented. So Slack, 70% of customers send 2,000 team messages every 30 days. I like that a lot. I just imagine the founder of Slack standing up in front of the company in the first year, whatever, he had a dozen employees and saying, our first North Star is to get to a million in revenue versus the founder of Slack standing up in front of those employees saying, our first North Star is to get 70% of our customers to send 2,000 team messages through our platform every month. That's extremely different guidance. I think that yields a very different company. And I like what the second one, the company that that yields. Dropbox, 85% of customers back up their device every day. HubSpot, 80% of customers use five or more features in the platform every month. Okay, so just some examples there. So I'll give you some guidance on each number. The E is the most difficult one and the most important. The P and the T are not as... P is, I don't know, definitely more than 50%. 75, I think is great. Ultimately, if you want to benchmark it to Wall Street, they're going to want to see 90% annual customer retention and over 100% net dollar retention. So it's probably got to be pretty high on the P here. The T is usually a month. You know, if you're something super PLG, quick value, like a Dropbox, you can go to a day or a week. And if you're something like huge lift, enterprise, change management, like a workday HR system that you're talking quarter or more. But in most cases, you're about a month. And the E, that's the tougher one, right? So a couple of guidance. Objective, meaning it's not customer success manager says they're happy. It's not objective. Instrumentable, we can't have humans going around figuring this stuff out over time. Because HubSpot is a billion-dollar revenue company, they still track this. So it's like, you've got to be able to instrument it into your operations to codify. Obviously, align with customer success and value creation. Like, what the heck are you trying to do and what behavior would exhibit that? Ideally, the unique aspect, right? HubSpot was the all-in-one player relative to all the other point solutions. So the fact that theirs was around the breadth of feature usage made, that was aligned with their unique value prop. And you can do multiple. With Slack, it could have been send 2,000 team messages and have more than five users or send 2,000 team messages or have more than five users. Just be careful because there's something beautiful. I almost prefer not in the beginning because you don't understand how beautiful it is when a 24-year-old customer success manager joins your company and all their other past employees were like, your job is to have 100% net dollar retention. And they're like, okay, how do I do that? And you're saying to them, your job is to get your customers to send 2,000 team messages every month. And it's like, wow, that's super clear. But all of a sudden when you're like, it's to send 2,000 team messages or have five or more users or maybe have the Salesforce integration, then you're like, okay, now you're diluting their efforts. You're complicating the vision of what they think they should be doing. You can do it maybe over time, but just tread carefully. And then measurement-wise, I don't like it when people just give me the number. It's like, here's how many customers are using it. Like we had 72% did it last. Yeah, well, that doesn't tell me a ton. Like, yeah, I can do that month over month. But if you signed up a bunch of new customers, obviously that number is going to go down. You're going to penalize yourself. If you're sucking at sales right now, that number is probably going to go up. If you churned a bunch of people out, then that number is going to go up naturally. So it's just like, it doesn't tell the story. And I want to tell the story in the month the customer is acquired. This tells me a lot more. And it tells me if I'm truly getting better. So if this was Slack numbers, it's not. But if it was like they acquired 24 customers in January, after a month, 3% had sent 2,000 team messages. After six months, 39% had sent 2,000 team messages. All of those 24, that's not good. You're Slack. You're like an organizational collaboration platform. And less than half the people or even half the customer are even hitting that number. But through a bunch of changes around onboarding, around selling, around who they're selling to, around the product, by September, they signed up 50 customers and within two months, 68% had hit the number. So product market fit occurred right around here. This is just a good way to look at it. And I can also see if it's degrading or if it's flipping back up. Like this one's degrading a little bit. And this one degraded and then flipped back up. That's a good sign that it's stabilizing around the mid to high 70s. Okay. And then like if you do encompass this, like do not spend two months with your leadership team like trying to figure this out. Like it's don't fall into over analysis, especially if you're small. Just remember that 90% of your peers and boards are choosing revenue and they're going to build a lesser business because of it. If you're just choosing any number that's a metric that's based on customer value, you're way ahead of the game. And we'll be able to analyze it like a little bit later, right? So let's say Slack had chosen the number of team messages that are sent. And if they analyze it and they're like, okay, well, it's been a year now. And a year ago, we acquired 68 customers. So all 68 of these customers are more than a year old. And 82% of them are still here. 55 of them sent 2000 team messages every month. They hit it out, the leading to get attention I call LIR. And 93% of those are here still. And 13 did not send 2000 team messages every month. And only 39% are still here. So boom. We learned something beautiful about our business. This does correlate with attention versus look at the same numbers. Like again, 68 customers acquired last year. 82% are still around. 55 had hit the lead indicator attention every month. 13 had not. But in this case, the ones that did hit it, 84% are around. And the ones that didn't hit it, 77% are around. That's not a big enough delta. And so we have to come up with something stronger. Maybe you have to choose the number of users. Maybe it's more about having more, 10 or more users. But it doesn't matter. Like we can, we have all the user logs. Like we can do this analysis in a couple of days. We don't have to wait another year. And also we didn't build a better, a worse business. Like we were focused on customer value, even though that particular value point didn't correlate super strongly with retention. We'll figure it out this weekend. We'll figure it out this weekend. We'll iterate on it. So just don't go into analysis paralysis. And if you're like a big company, if you're scaling up now, like you have user logs, you can actually do the analysis yourself today and figure it out. We were at 10 million, I think, at HubSpot when we finally figured out to do this. So we ran 50 different permutations and it was the breadth of the feature usage that came out. So that's my programmatic answer right there on what, you know, you should scale when you have product market fit. Product market fit's not about revenue, it's about retention. But retention takes a long time to figure out. So you have to determine a lead indicator retention. And I gave you some guidance with some variables on how to figure that out and measure it. Okay. But even then you're not ready to scale. Like in the last couple of years, we've introduced the term go-to market fit. And because at product market fit, I didn't say anything about like profitability. I didn't say anything about unit economics. At product market fit, you know, all you've proven at this point is that if you were to sign up 20 customers last month, the majority of them would realize the value that you promised them. They would hit the lead indicator retention. And Paul Graham, the founder of Y Combinator, is famously quoted as saying, do unscalable things early. Beautiful. Do that. David Cancel, the founder of Drift, during the product market fit phase, was like onboarding himself, often in person, the first customers with Drift that were paying them 50 bucks a month. That's not scalable. They're not ready to scale. But that's beautiful behavior at the product market fit time. Do everything you can to hit that lead indicator retention achievement. Very few entrepreneurs do, even the best. But once you do, you're not ready to scale. You have to prove that you can do that. You can sign up those 20 customers and deliver on the value you promised them with a profitable business. Profitably. And we don't talk about that in terms of gap accounting profitability because, especially in startups, we're going after big, high growth opportunities, which often entails heavy investment in engineering and product that is going to yield an unprofitable business, at least now, while we build out the big vision. But the actual scalability and profitability of the costs that grow with revenue acquisition need to work. And we call that unit economics. It isolates the cost needed to scale revenue from the overall cost of the business, like the finance suite and the team that's building the product and the office space that might not necessarily scale as revenue scales. So, we have to have both of these. And similar to retention, unit economics does take a little while to surface. If you're waiting for the quarter to end and the finance team to pull the numbers together and tell you what your unit economics are, you're running your business many, many months after when it's actually occurring. Those are all measures of what happened four or five years ago. Those are all measures of what happened four or five, six months ago. So, you will have to extract back the lead indicators of unit economics, just like retention. Now, that is easier. It's more codifiable. There's a more clear framework for that than retention because it's really simply algebraic. So, I'm going to choose a very common measure of unit economics, LTV to CAC ratio, lifetime value to customer acquisition cost ratio, which the industry wants to see three or more. Now, there's payback period. You can read about that one. They want 12 months or less. There's the magic number. They want one or more. There's the burn ratio. There's many different ones. I like payback probably the most. You can measure all of them. I mean, if you codify, it doesn't take any human effort each month to do it, as long as it's automated. But if you just have this target, theoretically speaking, when a 27-year-old account executive joins your company and they're like, hey, what's my job? What should I do? It really is getting LTV to CAC greater than three, but they wouldn't know what to do. So, we naturally translate that down to quotas and lead flows and sales cycles. That's all we're doing is we're doing the algebra. And so, as an example, a lifetime value is the average contract value. How much do they spend you, the annual contract value every year? How much do they pay you every year times the gross margin divided by the annual churn? And the CAC is at the customer level, how much of that customer acquisition cost was allocated toward getting their attention through demand and how much of it was towards selling the demand into a customer. And for the demand CAC, that's your cost per lead divided by the close rate, SQL sales qualified lead. So, if it costs you $100 to generate a lead and you close all of them, then the demand side CAC is $100. But if it costs you $100 a lead and you close 10% of them, the demand side CAC on a customer is $1,000. But then there's the CAC to sell the lead. So, that's your cost per account executive divided by the number of customers that they close every period or every month. And then the number of customers they close per month is how many leads they get or generate themselves or whatever times the close rate. So, this is an example of taking our goal of Uniteconomics LTV2CAC greater than three and extracting it back to more consumable, more understandable, more measurable, more real-time metrics, like how much the customer's paying us, how much did it cost to generate the lead, how much we paid our reps, how many customers they close in every month. And so, we can really assemble a business model. Like you're not going to get it right, especially if you're early and you're about to professionalize your team. This is not going to be right. It's just like through a combination of talking to advisors and looking at businesses that have a very similar price point or similar customer target, you can start to assemble a formula that yields good Uniteconomics. And then just instrument it, right? Like the red line is what you came up with and the blue line was what is produced every month. So now, while your peers are waiting for the quarter to end and waiting for the three or four weeks for the finance team to pull together the numbers and reconcile things and package up the board deck. So, now it's like finally April 15th and we're first getting visibility into January versus you have visibility in January. As long as the blue line stays above the red, you're going to hit the Uniteconomics target and you're going to be ready to scale. Now, the red lines might move around. Like this is a guess, right? Your cost per lead may be higher but your close rate might be higher too. So, it still works. But you're understanding this stuff in real time. So, that's the answer. When are you ready to scale? Not when you have a million in revenue or 30 customers or like when it feels like there's a pull in the market. Not at 18 months because that's when Snowflake decided to do it. You're ready to scale when you have product market fit and go to market fit. And we have very precise leading indicator measures using your numbers as to whether those things are in place. And I haven't said this yet but those are not approached at the same time but in parallel. Don't waste your time on go-to-market fit if you don't have product market fit because you could be optimizing the go-to-market on the run product and market combination. Yeah, you can have a couple wraps if you want in the product market fit phase. You can have quotas if you want. Just don't be dedicating resources that could be helping you get to product market fit sooner to go to market fit if you don't have product market fit now. They're done sequentially. Give it a shot. Now that was the first question. We're more than well more than halfway through. This one's a little quicker now that we've established that framework. Now we know precisely when to scale based on our numbers but how fast? This is the other deadly pothole. I mean, I go to so many board meetings where it's like, or I like talk to entrepreneurs and they're like, yeah, we did a million in revenue last year and we're gonna scale to four this year. It's good. That's fast. How are you gonna do it? And they're always like, we hired 15 salespeople at the beginning of the year in January. And I'm like, okay, how many salespeople did you have last year? And they're like, two. But they all do that. It works on Excel. But there's no analysis on like how many qualified interviews does it take to get to a good qualified hire? Who's doing all those interviews? How many managers do you actually need to coach and onboard 15 reps? Did you do any demand generation analysis? Like where did your demand gen come from last year? And is that gonna scale eight X? Like you're gonna fire 15 salespeople in a year. And that pretty much always happens. And they're also like choosing the numbers based on what Snowflake and Atlassian and HubSpot did in their year one, two, three, not what they're capable of doing. Like how capable are you of achieving that pace? You have to earn that. So scaling, as you enter the scale mode, this is not about like lump sum hiring of massive amounts of reps right after a series B or A round. It's not about massive amount of hiring at the beginning of a fiscal year because the board said you have to triple your revenue. It's about establishing a pace of let's say two reps a quarter. We're gonna bring on two reps a quarter. And we'll do that for two or three quarters. And we'll determine if we're going too fast or too slow. So how do you know if you're going too fast or too slow? Well, all your peers are deciding at the board meeting, which is like six weeks after the end of the quarter. So they're like, again, they're managing their business like five months late. You're managing your business by the day because you're leading indicators of attention and leading indicators of unit economics. That becomes your speedometer. So if you're adding two reps every quarter and all of a sudden these numbers start to get compromised, you're going too fast. But you know four months before your competition. But if we're adding two reps every quarter and it's fine, then go to four reps a quarter. And do that for two or three quarters. Then go to eight reps a quarter and do that for two or three quarters. And congrats, you're a unicorn. And you did it in a very scientific way based on your capabilities, not what some other unicorn 15 years ago did. Okay, so that's the answer to like when you're ready to scale and how fast. Using your data. The other thing that this framework does to finish up is now that you've got these clear phases determined, the actual optimal go-to-market decisions and the design of the go-to-market system changes as you progress through these stages. So I already kind of commented a little bit on the product market fit phase is this is like do unscalable things early. It's about finding early adopters. It's about throwing everyone in the kitchen sink at winning and onboarding these deals. You do not want a salesperson from salesforce.com that joined there four years ago. Like they had a huge playbook and training program and like leads and territories and you don't have any of that. You need someone who's like gonna talk to 12 customers in the week and summarize the feedback and talk to the engineers and iterate, iterate and change really fast. They're almost like half product manager, half account executive. I don't even need them to be on a quota or a commission plan. There's part of the team. We get some stock. Let's go. Let's figure this out. We don't have to talk about cold calling scripts or like inbound marketing campaigns. We got to get like 50 meetings. We can do that through our network. It doesn't scale. Who cares? That's the point. Don't worry about optimal pricing. Price for commitment. We're going to charge $50,000 a year for this. We're looking for 10 customers to be able to show value. We're going to give them a 90% discount for the first year, $5,000. Who cares? I don't care. As long as they're committed, let's get them in on and let's get their leading caterer attentions lighting up. That changes when we go to market fit. We need at least one scalable demand gen to go. We need to prove cold calling. We need to prove content marketing. We need to prove events. We need to prove like AE sourcing their own stuff, whatever. We need a playbook. If we're going to start adding people, we need to onboard them. We need someone who's going to build the playbook, different type of rep or manager. And we absolutely need to get the pricing right and the comp plan right. That's going to drive our un-economics. It's going to drive our contract value. It's going to drive our CAC, everything. And then a whole nother stuff comes up as we go into the growth mode. We become multi-channel. We have content marketing going inbound and we have cold calling. We might be selling to enterprise and mid-market. We have a very vibrant, large marketing team that needs to be aligned with sales, that needs to be aligned with customer success. There's a whole new set of situations. Okay, so I hope you appreciated that. If you want more on it, as Ray had mentioned, I published this work under Stage 2 Capital. If you don't know about us, we are the first VC firm that's running back by 500 CROs, CMOs, CCOs, thought leaders, Ray's a part of our network. And basically like the CRO, CMO, whatever, of Snowflake, of LinkedIn, of Salesforce, of SAP, of Atlassian, there's a whole bunch of them. And our objective is to operate like a VC firm, but we're bringing all that knowledge to the entrepreneurs that we partner with to help them with these things. And so we also publish a lot of content at stage2.capital. You can download the Science of Scaling ebook. It's about 45 pages if you want to read the detail. But we publish stuff on like setting up your annual plan, how do you pay your reps, like how to create a sales playbook, et cetera. So that's all I'll say on that if you want to check that out. Ray, what questions you got? Mark, wow, thank you so much. And I don't mean to co-op your leading indicator, but I'm going to call it my pet metric. Performance, percentage, event, excuse me, and then timing. But, you know, Kyle Poyar is also speaking at SaaS Metrics Palooza, and he talks about activation and activation rate. To me, your leading indicator for product market fit is really a great framework for activation. Do you agree with that? Yeah, I think activation is part of it. Like Kyle is one of the best in the world at product-led growth. And product-led growth is one business model, one sales motion, which I'm extremely passionate about. And I've done a lot of collaboration with Kyle on it. Usually when you're in, I would like to say that this is a little more abstract to almost whether your sales-led growth or your enterprise. I've even played with this outside of software in general, with even apparel companies, all that kind of stuff. So yeah, I think that's part of our job at places like Harvard is to come up with stuff that is abstract to many different businesses. And the product-led growth arena, you know, like, dao wao ma is pretty freaking good. Like, it's almost like you just take that, so that's daily active user, weekly active user, monthly active user. And so I'd almost report all three of them. And you want to look at them. And this came from Brian Belfort, who ran growth for us at HubSpot and is now the founder of the Reforge School, which teaches growth, where a lot of PLG people go. And he kind of just shows that exact same chart I was showing on like weekly or monthly customer cohorts and how the wao, the free user activation changes. The free user, you know, active account, how often are they set? Like, you often find like, okay, I acquire 100 customers in my PLG product and only like 60 activate. So I've got a 60% activation rate. And then after two weeks, 40 are still active. So 40% are active. And that goes down to 35. And that goes down to 30. You're not ready to scale if that goes to zero, if you acquire 100 users and then after three months, they're at zero. It has to flatline somewhere. And obviously the higher it flatlines and the sooner it flatlines, the better. And that's what Kyle's getting at. And it's, I see it Ray as an implementation of some of these concepts in a very specific business model and go-to-market motion of PLG. Good explanation. And the last question, it's not even a question, it's just a compliment. The fact that you are so focused on the cohort analysis, those customers I brought in January versus February versus March or this year versus last year. Do you find that the majority of companies aren't going to that level of cohort or segment-based analysis, Mark? Definitely. Definitely. I'm actually just publishing and this is the Q3 of 2023, depending on when you're listening to this, but I'm publishing something on this quarter on what the sales and marketing section of your board decks should look like, in my opinion. And there's especially the seed in series A level, it's just not there. And I rarely see this cohort stuff, Ray. I rarely even see an attempt at anything to do with user engagement. It's like, this is how many customers we signed up and here's how much revenue. That's the sales deck. It's like, that tells me nothing, especially at that stage. So yeah, Ray, I keep talking about this and thank you for creating platforms, Ray, for me to be additive to what you've done here because we still have a lot to learn as an ecosystem, in my opinion. And a lot of our audience today, CFOs, CEOs and go-to-market leaders, but for the CFOs out there who maybe they just raised 15 million in series A and they're thinking about how do I accelerate my growth right now? Understanding, do I really have go-to-market fit to determine how quickly I accelerate? This is such a valuable framework, Mark. I cannot thank you enough for being at this year's SaaS Metrics Palooza. Thank you, Ray. Thanks for organizing everything. Okay. All right, everyone. Our next session will start in just a few minutes.