#46: Lunch With Mark Roberge
Dave: What's up people? All right, we got a little special Seeking Wisdom for you today.
KK: It's been too long crosstalk-
Dave: Why don't you tee this one up? Tell us about Mark.
KK: This episode is a fireside chat with Mark Roberge, old friend. I've known Mark for a long time. Mark was the VP of sales at HubSpot from pretty much five employees through IPO. And he wrote a book, great book, one of the best books I've seen out there from a tactical standpoint called Sales Formula Acceleration.
Dave: Yeah. So, something that, yeah. Sales Acceleration Formula-
KK: Sales Acceleration Formula. Yeah, slipping. So you should read that book crosstalk-
Dave: Well, you crosstalk to remember a book title, it's okay.
KK: No. I've read it though. So we had a fireside chat here at Drift. I gave a great presentation in front of the team, taught us a bunch of his stories about different stages of selling, and we recorded it because we want to share because that's how we do it at Drift with all of our Seeking Wisdom peoples.
Dave: Yeah. So here it is. Here's the episode with Mark. Just pretend you're here with us at Drift hanging out, eating a sandwich and then-
KK: Get cozy.
Dave: Yeah, get cozy.
KK: Mark is an advisor to Drift and I want to welcome him. Actually, Dave usually does the introductions.
Dave: Okay. crosstalk good.
Mark Roberge: Let's start from the beginning crosstalk-
KK: Start from the beginning? Okay.
Mark Roberge: Because I think you missed it.
KK: Okay, so let's start from the beginning. Mark, thank you for coming. Mark is an advisor to Drift and we worked together at HubSpot where he ran sales and lots of other things. And we invited mark in to give us a talk about the lessons that he's learned in selling to mid- market and to individuals through a PQL model and just the lessons he's learned from going to Xero to IPO.
Mark Roberge: Great.
Mark Roberge: Great to be here, Dave.
KK: I'll facilitate.
Dave: Nice. We need that help.
Dave: So, I think the number one thing that'd be helpful to start is just, we're at this stage where it's basically been I don't know, seven, eight months since we have been in market with this current product. Bunch of customers in early stage. You joined HubSpot basically from Xero, right. But maybe fast forward to year one after being at HubSpot and what your model was and how much of it was actually scientific versus how you guys were selling every day.
Mark Roberge: Yeah, so I have a particular framework that I advise SAS companies on that we did not do well at HubSpot during this stage, but I think it can really simplify and create a really interesting lens through which organizations can look at their growth and it's customer success, then unit economics, and then growth. And I think it's largely driven from this subscription revolution that SAS started, but now we buy razors on a subscription, everything's a subscription these days where the days of an 18 month sale cycle and jamming a product for multi- million dollars into a company that just sits on the shelf and isn't used are over. And so it's really easy to engage with a product today, and it's real easy to leave a product today. And it doesn't take long for modern organizations to realize that churn is the key metric when it comes to their revenue funnel. And there are too many companies, especially at your stage right now who they're like, " Oh, this is awesome. We've got 400 customers. We're doing X revenue, let's just go." Literally it sounds crazy, but more times than not, when inaudible the stage, they're like, " Yeah, we're about to hire 27 representatives next quarter." And they have three today and it's like, " Dog, I know it's going to happen," because every time I've seen that a year later, they're laying off 27 representatives. Because they just, they just went too fast and didn't understand the underlying economics of the business. And so that's what I look at right now is I'm like, " Let's first talk about customer success, and you've got 400 customers now. When you put one customer into your product, what percent of them love it? What percent are never like, 'Never take this away.'" And I don't even care how much it costs you to put one person to the product could cost you$ 20,000 and you sell it for 5, 000. I don't care. I just want you to put 100 people in there and I want to see 80%, 90% pick a number using it every week, and if you were to survey them, do an MPS survey, whatever, they're like, " This is an awesome product." That is step number one. And once you have that, instrument it so you can watch it closely. Don't rely on, " Yeah, the way we measure customer success is after a year we see if they turn or not," that's a lagging indicator. We need to understand something within the first 30, 60, 90 days that we have confidence will correlate with longterm customer success so that as we continue to scale, we can keep an eye on it. So, that's step number one. Step number two-
KK: I have a question about that stage actually. So, how do you coach the companies that you advise. Or even giving us a lesson. It's easy to say oh, this company is growing, the reason people want to add more sales representatives is because it means you can close more deals, you can grow faster. How do you balance that expectation with like you were talking about churn. How do you manage, " Hey, we might grow slower, but we're going to get this right?"
Mark Roberge: I think Dave is awesome at this. It's the same story of how do you manage an investor where you want to go through a proper product market fit journey before you actually scale. And I think some investors are coming along to that way of thinking. And David has done a phenomenal job of setting expectations that we're going to grow in a healthy way. And I think it's education to the investors that this is necessary. And I think some of them have been burned enough that they'll hopefully be on board with that. And it's not like you can just stop growing. It's just I would never want to do something strategically that is going to get in the way of us finding great customer success. You can do other stuff. You can experiment with Facebook paid marketing, you can do content marketing, you can hire a couple representatives in a quarter to do other things in parallel, as long as they don't sacrifice how quickly you're getting to customer success. That's my opinion. So then customer success then you and economics. So at this point we're comfortable with the retention and churn of the business. And now I need to pay attention to that it costs me$20, 000 to put someone into the product where I'm only generating 5, 000. Now it matters. I don't care before. I think it's Paul Graham says" do unscalable things early." Great. But now it's like I need to pay attention that the economics are healthy. And usually in SAS, people talk about payback period of less than 12 months and LTV to CAC ratio of greater than three. This is where the stuff starts to pay attention, and we got to really look at our sales execution, our demand generation execution, et cetera. If we look at that and instrument it and we're comfortable, now let's go after growth. And in a business like this, it's like let's hire a representative a month and let's see how it goes. And if that goes well for a couple of months, hire two a month, and keep a close eye on our customer success cohorts to make sure they're not going down as we expand our target market that go on after and keep a close eye on our unit economics of payback and LTV to CAC and all that stuff to make sure that we're properly ramping and bringing on new people. So that's the framework I look at. And gosh, to this day, I still see so many organizations jumping to the growth stage and 100% of the time I see that it burns them. Not all the time, but the ones I see, it burns them in the end.
KK: You said you guys learned this at HubSpot. Is this is crosstalk-
Mark Roberge: crosstalk.
KK: Tell that story a little bit. I think the best way when you tell stories like this is to say, " Here's how it screwed us, and here's what happened."
Mark Roberge: So, I'll bring us back to August of 2007. We had 10 people, maybe less at HubSpot and maybe two or three sales and 100 customers. And so we just needed customers. We were talking about this inbound marketing thing doesn't work, we just need customers. So I wrote a very hunting oriented communication plan, which is basically for every dollar of MRR you burn in pay you$2. You sign a$ 500 a month deal, you make$ 1000 commission check.
KK: KK is over there like, " Yeah."
Mark Roberge: Awesome. Hunting plan. In fact, accelerator. Your quote is 10 deals, the 11th one I'll pay you twice as much, right? As you sign it, it's like, " Yeah, this is great." And slight protection on the churn, they have to pay for four months, otherwise we call back. So if they cancel up to two months, we call the whole commission back. If they pay for five or six months, it's yours.
KK: Okay. Can I ask a question about that? Would you actually come to me and you take them? You hold that money or it comes out my future paycheck?
Mark Roberge: Year, it comes out of the future. It's hard. There's a debate as to how you manage these clawbacks. And I don't have a strong opinion on that one, but I do think that when it comes to sales commissions, the reward or penalty for good or bad behavior has to be felt close to the good or bad behavior. So that's why we did it this way. So this is a beautiful hunting plan. And we went from 100 to 100 customers in seven months. We went from maybe 200,000 ARR to three million in six or seven months.
KK: What was the feeling like? Was everybody like, " This crazy."
Mark Roberge: Series B was a piece of cake, everyone was chatting about us, huge momentum in the market, churn went from 4% a month to 6% a month to 8% a month.
KK: So over the course of a year you're basically losing all your customers.
Mark Roberge: Yeah. Monthly. So again, huge hunting order plan. This is the beginning of SAS. So no one even understood this stuff, but it's not a sustainable business. And so we're living through this and we're making this mistake. We're certainly not following my customer success. Then you're not going to have excellent growth, because during this period, I'm adding one representative a month, we went from two representatives to 10 representatives during that period, something like that. So what was surprising was we thought that the churn was a cause of either the product or the post- sale experience. So I actually looked at, I think we had maybe half a dozen customer success people at the time. So I measured the churn per customer success person thinking, " Hey, maybe one of these people has it figured out, we can just figure out what they're doing and replicate it." It turned out the difference in churn was nothing compared to all of them. Then I rerun the analysis on the dozen or so representatives we had, huge difference. Some of the salespeople had really low churn and some had just crazy churn. And in fact that was inversely correlated to their revenue production. So I was like, " Geez, this is a sales issue. If this is all driven based on the customers our salespeople are choosing to go after and the expectations they're setting." In fact what was funny was the churn rate in month one of a customer was really low, month two low, three, four low, five is when it spiked. So I was like, " Wow, do representatives really work their comp plan?" I randomly set the clawback at four months and it's just-
KK: So you're best friends for four months and then the second month is over you get paid?
Mark Roberge: Yeah, exactly. Yeah.
KK: "See you Mark."
Mark Roberge: So, that said to me okay, number one, this is a pre- sales issue; number two, the comp plan will probably fix it. So, that's when I stack ranked the representative, I shared the data we found and said, " Here's the top rankings, not from revenue production, but from LTV and churn retention." And I said, " I'm going to start comping you based on this from here on out." And we did, we changed the plan to basically instead of getting $2 and the dollar, the top quartile got$ 4 on the dollar. So those that were selling good deals, we doubled their commission. And then there was a mix, and then the bottom quartile got their commission got cut in half. And we went to rapid training on proper selling and churn dropped by 60 or 70% in six months. So we lived through this pain and from then on out when we launched new businesses, whether it was enterprise, e- commerce, partner program, the sales business, I always put pressure on the directors that we assigned the business with this framework of customer success economics and growth because of that lesson learned.
KK: So looking back, so you think of a company closer to our stage or people that are going to listen and watch this later, you had 100 customers and went to 1000 in six, seven months, that's your biggest takeaway is to apply the same things that you had to apply 1000 customers and learn the hard way to day one, basically?
Mark Roberge: I wouldn't have hired the 10 representatives during that time. We didn't even understand churn or customer success at the time. So I would obsess over that, get to whatever you deem as scale. If it's an enterprise product, if it's going to be dozens of customers, if it's more of a transactional SMB product, it's going to be hundreds of customers and get comfort around customer success, then moved to economics, then move to growth and see if you can't work on your overall capitalization structure and the expectations with your investors, that this is your journey. You're probably not going to get through each one of those phases in three months. Hopefully they don't take three years each, but we want to be on the same team here as to how we're going to navigate this-
KK: inaudible. If you did it your way could have been 100 to 300 to 500 versus 100 to 1000.
Dave: But actually, you keep the microphone, because I want to ask you. So you talk about customer success in the early days from a sales perspective, that's what you guys had to fix, right. But from DC, from your perspective, so how much of churn is based on, is it sales? Is it the product? And what's the relationship between say the VP of sales and the head of product at an earlier stage company to figure out are we building the right features? Is this working the right way versus are we selling to the wrong people?
Mark Roberge: Yeah. Well I think the simple answer is it's all, right?
Dave: It's both?
Mark Roberge: It's both, it's all, it's like there is no distinction and it's how are we selling it? How are we marketing it? How are we talking about it? How are we setting expectations? How are we onboarding? What does the product do? What is it missing? It's all those things. I think depending on the day would be some random, different prioritization of those things coming in. And then tomorrow it will be a different rank and next day it'll be different. It's constant changing that way.
Dave: Let me ask a follow up on that though, because just as my own education, because I feel like what I just talked about isn't talked about a lot and a lot of entrepreneurs mess it up.
Mark Roberge: You Just told everybody to grow slower. Nobody wants to hear that.
Dave: Yeah, in a way. But product market fit has been a term around for a decade. And I like as technical entrepreneurs, we've grown a lot from the perspective of product market fit. So, that's the disconnect I have is... I guess people are always asking, what is product market fit? How do you know when you get there? And if we've grown as an entrepreneurial community around that, is it different than the first phase of my framework of customer success and what's the disconnect?
Mark Roberge: Good question. So product market fit, at least probably a decade since it's been popularized and basically Steve blank talks about this customer discovery process and how do you get to a point where you basically have found something that people are willing to pay for and enough people are willing to pay for it, and then that process is repeatable. And then you've hit this idea of product market fit which Marc Andressen first used those terms publicly. But I think one good thing that you bring up is that people look at it as a static point in time, but it's not, it's exactly what we were just talking about. It's this continuous thing, product market fit is never finished. It's like you have a product market fit for a context. And that context is a certain time and place in the market. So competitive dynamics there, what's available. Plus what's trending from a trend standpoint. You have a certain product market fit for a certain cohort of users that expect to pay you a certain amount of money. And as soon as you start to change some of those dimensions, the market changes, competitors change, the price that you want to charge changes or your persona changes, because you want to go up market, then you're reevaluating product market fit almost everyday.
Dave: I think that that puts things together for me a little bit with this is maybe although we've grown in product market fit understanding, we haven't appreciated how dynamic it is. And even we had Brian Balfour in this last week at HBS, awesome growth marketer. And he said the same thing around the students asked him he's big on defining personas and doing a lot of customer interviews up front to understand your marketing channels and your growth marketing strategy. And they were like, " How often? Do you just do this once and then you're done?" He's like, " No, it's continual." Because you can bring in crossing the chasm a little bit. Your first set of customers, the early adopters, great. Once you move to the early and majority, a lot can change. And he says redo the persona analysis, and you're saying you've got to reevaluate product market fit. And I'm saying, you got to make sure customer success is still strong. So that might be the pattern is these are not static points. These are continual inspections that will evolve.
Mark Roberge: Yes, three- dimensional thing. I think the crossing, the chasm visual is good, which is the bell curve. So you're following the bell curve, but then the bell curve has to be applied against different times and points in the market, different sets of competitors, different markets that you're moving up and down throughout, and all of a sudden you have I don't know how many, but any number of bell curves that you're chasing at once, that you're trying to figure out.
Dave: So, the other thing that we want to talk about is what's changed in the world of sales since 2007 and even just whatever the last two, three years. The biggest things that you're seeing, the things that are coming up and what that means for companies like us or people that are listening and how they're actually going to go out and start selling.
KK: So I think one important thing that's important here is that Mark is one of the few people I know in the world probably that have gone through this MQL playbook and has gone through the PQL playbook. And so people have touched different parts of that-
Dave: Explain that more for people that might not know what that is.
KK: So I'll let the expert explain it, but he's going from this world of... It's basically how much control does the buyer have starting with the MQL playbook was highly effective in a world where buyers had more control, but not as much control as we see today and so that you could go through a marketing qualified lead process and the sales person had more power. Now we're moving to a PQL era where it's product qualified lead, where you're selling to people who may have more knowledge than the sales person, because they are active users of the product. So how do you sell them that crosstalk?
Dave: What is a good example of PQL model today?
KK: Slack is probably the one that everyone knows, Evernote, Trello, a lot of consumery type things look this, but it's also in the enterprise as well. New Relic is another one, DocuSign. There's tons of stuff all over up and down the stack. And then MQL playbooks, there are lots of those examples as well. Still there, but there's a shift, it seems to me.
Mark Roberge: Yeah, I totally agree. I think the foundation for all of my observations that would probably fit into your what's changed recently is driven by buyer empowerment. It's something that's been talked about by a lot of different people. We used to as a buyer, if we wanted to research a product or buy something, we had to deal with a salesperson. We had to go up to the trade show booth, we had to call them up, we had to take the meeting just to figure out what it does. That has obviously changed. We can be in our bunny slippers on a Saturday night on Google and figure out what the product does, what the competitors do, how much they are. We can use it to your point, try the product and sometimes buy it without a salesperson.
Dave: Can I tell you a quick story about this? Earlier this week I was trying to buy something for Drift, signed up, already put my credit card in as part of the onboarding process-
KK: SAS product.
Dave: SAS product, yeah, paid them$ 20 a month, realized we needed more seats. So I went to find how do I upgrade on their website? I was already a customer, I had to fill out a form. 48 hours later, I got a call from their sales representative like, " Hey, Dave saw, you're trying to upgrade your account. How can I help you?" I already canceled and I hopped on KK's account, and so now we're just going to free ride on that.
Mark Roberge: So that's a perfect example. I would say that's the optimization of that new age sales process. But that's the biggest thing that I think we've only scraped the surface on how this changes sales, to your point, of appreciating how today's empowered buyer who doesn't really think they need salespeople and how that impacts the salespeople we hire, how we train them, how we compensate them, how we generate demand. So one of the things that I look at is there was a great article I even this week about if you think you want to hire that aggressive sales person, that flies in the face of today's modern buyer preferences. And there's been some data and some individual hiring managers who've had more success with people who are curious and coachable and do a good job of our job is... If all we're going to do is pitch what a prospect can read on our website, that adds zero value. And that's really what a lot of selling was. Now it has to be understanding the unique context of that buyer and translating the generic information that's available online to that context. That's really hard for a buyer to do. You were buying this SAS product in some random space. It's probably the first time you've looked into that space. Salespeople help people like you multiple times a day. They are an expert at understanding your unique needs.
Dave: And all I wanted do is ask somebody, do I really need this if I'm going to do this and just had faith that they would suggest the right plan, to be honest?
Mark Roberge: As opposed to, " Hey, let me show you our 10 slide deck that our CEO or our VP of sales put together."
Mark Roberge: And then you're like... You're just sitting there trying to figure out how does it match? And I'd criticize the buyers a little bit too here because a lot of salespeople try to get on with the buyers like, " Hey dude." Enough with the questions, just show me what you got. No, when I walk into as a buyer, I'm just like, " This is my issue. What else do you want to know about it?" That's how we should buy to help the salesperson out help us." So it changes how we compensate, how we hire these people, how we train them. Without the HubSpot, there's a big thing around get the salespeople to walk in the buyer's shoes for them to execute what I just talked about. They have to understand more than just what the product does. They've got to understand the life of the buyer. So our salespeople always went through a month of writing a blog, building a social media following, doing email campaigns, creating landing pages, all for a mini business they made up. And they could relate to that small business owner or that marketer when they were skeptical of whether blogging would work for them, how hard it was to rank on Google, they went through it. And so that's an example of how we need to train... And to David's point, it changes how we generate demand as well, because the better companies are not contradicting or fighting this buyer empowerment, they're aligning themselves with it. There's a lot less friction for a buyer to engage with your product. So, let them do that. Feed into that and then teach your representatives how to sell or engage with someone who's already working with the product i. e a PQL as opposed to the cold contact needs or the ebook downloads or whatever.
KK: Are you looking at different things in this model? Like in the MQL model you might say, " Oh, Mark is a VP of sales. His company has a thousand employees, they're in this industry. This makes him a good fit." Whereas in the PQL model, you might look at more, " Oh, Mark has done this many things and invited this many people and uploaded this many contacts." It doesn't really matter who you are personally?
Mark Roberge: It changes some huge folklore or best practices in sales. An example, call high. Call high. That's what everyone says in sales. Why would you ever deal with an end user who doesn't have budget? Call high, call the CEO. That is a challenge today. And we've experimented with it and proven it, at least in context I've worked in that it's not optimal. You call the CEO, the VP of sales, and you can say the stuff that resonates with them, which is way different than the end user, but they have to convince the end user to adopt it. And just look at CIOs today. They've gone from a world where they sat next to the CEO, just trying to keep up with what technology they could buy to align with her strategic objectives, to a world where they're just trying to keep up with what their employees are using and tighten it up, make it secure or whatever. And so who has more power in that? The CEO or the fact that one third of your company is already using this product? So we've had a lot of success and I've had a lot of sales with different companies on starting low, winning engagement, trying to figure out how that first user is successful, how it goes a little viral through the organization and then using that as your competitive advantage to win the power. Just a quick example, compare Dropbox to Box. Box comes in with their really fancy suites and the big story that the CIO wants to hear about geographic redundancy and all this fancy enterprise stuff. And Dropbox comes in and it's like 30, 000 of your employees are already using our software. So, which one do you want to be there?
KK: So, the thing that you started with this is compensation plans, that's the selling strategy that works, but that probably only works like something DC says all the time is aligning incentives. If the sales representatives' incentives are aligned to start and grow, then that's different. What is the new compensation plan?
Mark Roberge: That's the second big one outside of call high, what this challenge is the compensation plans are traditionally designed to maximize the first revenue from a company. You're a salesperson, your job is to open the door. That's the hardest thing. Get our foot in the door with this company, sell a big deal, and that's what a compensation plan is aligned to do that. That compensation plan is perfectly in conflict with this new buyer journey. The whole opportunity is to reduce the friction to engage and to grow the account over time, the LTV. But we're awarding our salespeople to maximize the first purchase and don't care about anything after. So it's like that's what we did with the sales products at HubSpot is I told the guys, " This is what I'm going to do. I'm going to comp you 20% more for all the revenue we get after the first sale." And they were like, " Dude, that's so messed up. You know what I'm going to do, if I have a 50 seat on the table, I'm only going to cause one of them, one seat this month, and then I'm going to close the other 49 next month to work the compensation." And I was like, " Great," because there's no way that they're going to upgrade because I know that sales representative could sell the 50 seat deal, but then what if it doesn't work out? That's a huge churn versus if they sell the one seat and then they try to sell the other 49 a month later. If it's not going well, no one's upgrading it. And if it is going well, they'll upgrade it, and it's a sticky account.
KK: The Roberto's at every sales compensation discussion is like, " You know what I'm going to do in that case, I'm going to do this crazy thing. I'm just going to sell$ 3 million deals and ignore everyone else." And it's like awesome, sell$ 3 million.
Mark Roberge: When you're designing compensation, one of the most important exercises is you put together your corporate strategy, you figure out how to align your representatives with it, and then you take a step and be if I were to cheat this plan, how would I cheat it? It's a very, very important exercise to do. All the news on Wells Fargo in the last two months, that just wasn't done well with their plan and it cost the company a lot. So you've got to go through that and manage against it, because there's always a loophole in every plan.
KK: All right. So maybe we'll do questions in a second or something.
Uncle DC: I have a question.
KK: Go ahead. Uncle DC, please.
Uncle DC: Okay. I have a question. So you've gone through the MQL era, into the new PQL era, but in both cases you had sales representatives. How do you now look at this world where you hear rumors of whether it's Slack or Atlassian or other companies that are moving claim that they don't have salespeople or other people who are saying they're moving to salesperson free world. How do you think about that?
Mark Roberge: I'm not defensive about it. If my field goes away, I have other things I can do. But I am skeptical. I've good friends with Jay over at Atlassian and spend time over there, and I think it's a branding thing, honestly, with a lot of these companies because I'm like, " Who's that person over there on the phone talking about contracts and fees?" " Oh, that's customer support."
Dave: It's an advocate.
Mark Roberge: Yeah, why is there one guy in a suit in your company? I think it's noble, and I think it's great branding, but I do question whether we rename it sales, maybe that's the problem. I think if you ran an experiment of just completely humanness, just allow people to figure it out versus introducing a person. Let's not even call them a sales person, but it's just someone that reaches out when someone gets stuck to understand why they're stuck and to help them, and then to close the deal, to convert it to revenue. I think in a lot of those companies that experiment would yield that throwing humans at certain situations would be a positive ROI thing. So I think it's great branding. I think it's noble to move the field to where we need to move it. But I also think that ladder context exists.
KK: Cool. Anybody got questions? Yeah.
Speaker 5: So, when you were at HubSpot and you were starting to understand the importance of inaudible success, what did you find to be the most crucial inaudible and what was the most crucial inaudible.
Mark Roberge: So you want to repeat?
Mark Roberge: So at HubSpot, when we first figured launch customer success, what were the crucial things that we learned about it? I think this one still exists as an issue with some of these organizations that are moving out is we focused our efforts on the people who were about to churn. We hired six, 10 people and was just like, okay, here is an identification that these folks are going to churn and to spend a lot of effort there. And after some experimentation, we realized that that was lost because it was too late. And we found tremendous value from instead implementing a proactive engagement strategy with our customer success team. So, having that three or six month check- in, having things as opposed to within a week of usage drop offs, or two weeks of usage drop offs to engage with them. That served to be a much more a higher ROI value from our customer success folks. So again, it's just natural. They're like, " Oh, customer success, the churns are crazy right now. Let's just figure out the people who are at the biggest risk and focus on them." I would challenge that and say instead just think about a proactive model and you almost have to look at those people as a sunk cost in a way.
KK: Any other questions?
Speaker 6: One of the things that we care a lot about here is inaudible your approach to testing different things especially as relates to inaudible.
Mark Roberge: Sure. Yeah. So experimentation is an enormous entrepreneurial philosophy that the great companies have these days, especially in this whole big data explosion, we've had so much data at our fingertips, especially in the go- to- market funnel, that it's a shame not to experiment with this stuff. Just, we have a lot to learn from the basics of the scientific method that it's like, " Oh yeah, well, of course," but people just don't do it. Number one, be really precise around what success means. Quantify it. We've got to run this experiment, what is success and failure and define that in advance so that we just know and don't lose sight of it. Number two, be really precise around not contaminating the experiment, like control groups, et cetera. So we actually do learn. And then number three, I like to have an experiment funnel in a way. You can't run with 20, 30 people in a room, you can't run 50 experiments at a time and you probably have 50 ideas right now. So which ones do you choose to do? Be aware of the speed by which learning comes out? If I told you a lot of experiment that would create a 10% lift potential in your sales funnel and would take a year to learn, that's not as appealing as something that could literally double productivity and we'll know in two months. So you've got to look at it from the lens of how rapid the learning can come.
Speaker 7: I have a question. So one thing that you touched upon there, but we didn't go into was your approach to experimentation, specifically sales and marketing experiments. And I think that was an interesting discipline that I had not seen before in working with you. And if you could just talk about how you evaluated each of those experiments, how you thought about experiments and investing in those, that'd be great.
Mark Roberge: Yeah. You'd be surprised. One of the philosophies I teach at Harvard actually is sales compared to any other function is arguably the most... Success and failure is the most quantifiable. It's really hard to look across an engineering team and be like that's our best engineer by 7%. It's hard to do. With lines of code, adoption of your features. In sales probably you can get damn close. That is our best sales person by 7%. And so, it's a shame to not leverage that context in everything you do in sense. Hiring is a great example around experimentation. Every sales context in my opinion, is very different. And the ideal hire for each one, the ideal hire here is different than the ideal hire at HubSpot, which is different than the ideal hire at Slack, which is different than the ideal hire agenda. They're just different contexts. And so, that's an area of experimentation is if you can be precise around the characteristics of the people you're evaluating and quantify how you measure those things. And over time start drawing correlations between your metrics, quantifiable evaluation of someone and their quantifiable success or failure, you start to build a hiring formula, and that same model can be applied to training, giving people tests, after training to see if that correlates, it can be a higher applied to coaching and the progress they make on connect rates and opportunity conversion. So, I think to have that experimentation culture, the underlying piece for me is sales, success and failure is so much more quantifiable than any other discipline.
Speaker 7: So we talk a lot about interviewing and hiring, especially on the podcast. But there's been a few people that I've seen interview as many people as you have, who most of them come out a little shaken up about your questions. So tell me about how you approach hiring, because I think-
Mark Roberge: They're shaken up about my advice? It doesn't quite crosstalk-
KK: No, about the tough questions that you asked in the interview process. So tell me how about how you approach the interview process.
Mark Roberge: So, the biggest one for me is, okay, I'll do the whole interview real quick. The interview starts when I meet you in the hallway, and it's not like I'm going to cut you off at this point, but it's a huge opportunity for you to win, which is, do you know anything about me? You should research who you're interviewing with, and know a little bit about them. And also, is it me breaking the ice or are you breaking the ice and asking questions? That's a huge opportunity around their preparation and their curiosity, right.
KK: That cuts 80% in my experience.
Mark Roberge: It's not a showstopper, but in an outside environment, it would be close to a showstopper for me and inside I can deal with it because it's not as much as the context that they're selling in. Then I break the ice and like, " Why are you interested in this company?" " Where do you want to be in five years?" All that kind of stuff, which is interesting to make sure we're aligned with where they want to go. Then I ask them a little bit about their prior success. Like, " Hey, you're at Oracle and you're an account executive, where do you rank?" " Is that based on bookings or revenue?" And then the biggest thing for me is the role- play. And I'm amazed at how few sales hirers actually do role- Plays. It tells me it's so much, and I do the role- play on our environment, right? So, " Hey, you're going to be a sales person at Drift, I'm going to be a VP of sales at a security company, and I just entered into a free trial. Go." So, I watch how they do. And if they do the show up and throw up and basically spend five minutes telling me everything they memorized on our website, it's not over, but they have a lot of work to do on their natural curiosity and consultative selling model. Versus if they ask me great questions about why I entered the trial and what I'm trying to accomplish, and then that's really great. And then I try to school them on domain knowledge that they hopefully would have learned. In the HubSpot environment, I would just, "How do you rank in Google? How does that work?" And I keep asking questions until they're stumped. And the longer you go, the better you've done in terms of your research, and then I can see how they react when they're stumped. If they can remain confident and hold their own, or even just say, " I'm not sure about that, let me check." That's great. But if they fall off their chair and" Oh, I'm failing," That's tough. And then I stop the role play, and this is my most favorite part. I'll just say, " Okay, how do you think you did?"
KK: Have him self assess.
Mark Roberge: Now I'm testing their coachability. And if they're like, " Oh, I was awesome." I'm not that psyched about that. And if they're-
KK: crosstalk how few people will pass this.
Mark Roberge: But if they're like, " This is what I did well, this is what I would do better," that's great. And then I coach them. I say, " Hey, listen, I always give one piece of positive feedback and one piece of need- for- improvement because I don't want them to lose their confidence. And then I coach them and I see are they glassy- eyed or are they taking notes, and then I have them redo the role play. So, the role plays so critical to me, and they all fail the role- play the second time, because they're just like, stuff's going through their head, they're in a high pressure environment, but the effort they make and how they're receptive is a big tell for me. And also gosh, sometimes I actually see improvement in the second role- play. That's pretty cool, man.
KK: Pretty cool. Yeah.
Mark Roberge: If you spent 10 minutes coaching a salesperson on a critical motion that will lead to success and they improved, what is it going to be after a day, a week, a month with that person? That's pretty cool, yeah?
KK: One definition there. What did you mean by glassy eyed?
Mark Roberge: You're coaching them on a need for improvement, and they're just asleep. They're just not absorbing it.
KK: That's awesome. We talk about that a lot, but most people don't get it. They're like, "What does that mean?" Just look at their eyes. They're asleep.
Mark Roberge: Exactly.
Mark Roberge: Exactly. How receptive they are.
KK: That is awesome. Any other questions?
KK: All right.
Kevin: So, inaudible over time inaudible what you guys did at HubSpot for sales products? What do you generally see works best? Is that customer success role that does the upgrades? Is that a sales role? Every company does differently. Have you seen inaudible?
KK: The question was from Kevin. He asked, when you go into PQL model where you bring someone in for a low price point, let's say 50 bucks a month, and you want to grow that over time. Who's best suited to handle the growth of that account. Is it the initial sales representative? Is it customer success? Is that some combo of the two?
Mark Roberge: It's highly circumstantial and I think it's usually a combo. I will tell you that the mistake most people make is they put through the compensation plan or the rules, they put those two people at odds with each other? So, if they upgrade in the first 60 days, the sales person will get the credit, but after that, the customer success person will get the credit. That's putting them at odds with each other. The customer success person is like, " Maybe I'll wait to call them." And the sales representative is like, " Geez, I hope they..." And God forbid, they upgrade in the 61st day, the sales person is "Oh, that sucked. I worked so hard. Took me a year to get that account. And then this guy made all this money on it." So there's ways to figure out the financial engineering of the business to, it's not double compensating, but it's sharing the upsell on that. And then in a way you can almost let them figure it out. I really like the models of not having a marketing team and a sales team and a customer success team, but having these smaller cross- functional pods of one or two BDRs, a couple account executives and one or two CSMs that are working together and all the accounts are shared and you can have a blueprint around what your different roles are. But as long as they're sitting together, they have a relationship and their compensation is set up to be aligned, they'll figure it out on their own. And usually it's like if the expansion requires what we perceive as complex sales tactics, worming through the organizations, politics, getting to champions, getting to power, these are typically things that CSMs might struggle with. And that's where the sales person wants to be more involved versus if they're trivial and it happens more naturally and it's really just around proving success and then asking for expansion. Then those hunting skills are probably better spent on new customer acquisition and the CSM can handle it. But that's the big thing is making sure you set up your organization design and compensation structure so that those people are playing on the same team.
KK: Awesome. Luke, you had a question?
Luke: Yeah. I was wondering how these things like freemium mini rates and all this stuff, something that sales people aren't even thinking about, all the marketing or make you do an outbound sales stuff, then you're talking to someone it's like this person makes sense in a freemium, maybe I can plant the seed to replay her. How do you form a call structure on that?
Mark Roberge: Yeah so-
KK: Yeah, so longer question from Luke has to do about freemium, and how does it relate to the sales representative, and do they think about the freemium at all or do they consider that in their process?
Mark Roberge: I think it's a big overlap with some of the questions we talked about around it's a perfect representation of lower buyer friction. Freemium is the extreme of that, of just allowing people to permanently forever use this stuff for free, and once they hit certain thresholds to upgrade, we call it adopt- before- buy philosophy. Usually the salesperson was, " Okay, you're all excited. Now you have to buy it, then I'll let you use it." And it reverses that. And I think it should be embraced by sales teams. I think sometimes when things aren't set up properly, when sales representatives aren't trained properly, when they're not compensated properly, it can contradict and you'll see that when salespeople try that, " Can we just hide the free trial button?" Or they're doing sales training on how to convince people that they don't need a trial, they shouldn't start with the freemium product. The system is not working there. So I think it's a lot of the stuff we talked about, which is to compensate the representatives more for the follow- on success and revenue expansion than for the first deal. If you do that, they're going to love freemium and they'll learn it. And they'll learn it the same way than go back to the old cold calling world. You started to develop a sense of what were the perfect fit companies and the perfect fit roles that a cold call is worthwhile to, and you started to develop a sense of who wasn't worthwhile because of budget or authority or whatever. That same lens can be applied to freemium, but in a slightly different way, it's like you start to figure out which of these users are a really good fit in aggregate, and that are worth my time to help them embrace the product. And it's almost like we've replaced the whole open the door, do a discovery, a qualifying call, get that meeting, set up the demonstration, talk to the CEO and close it. With emotion that's more like talk to the end user, get them to love the product, get them to tell their neighbor, get them to love the product, and then go to the CEO with, " Hey, your team loves this." But they have to have that same qualification lens of if you're talking to a part- time business in a garage, you probably won't invest that time. But if you're talking to a hot SAS company that's doubling their employee base every quarter, then you probably want to engage with those folks.
KK: Awesome. Last question. Will.
Will: When you talk about figuring out, originally you said the number of people who would love your product, that's 80-90% of whatever the number is. How do you quantify that? Do you need a certain number of customers inaudible and you come up thresholds or how do you make that more actionable other than looking at the customer or inaudible feels good?
Mark Roberge: Yeah, I think it's a tough question early. I think you've got to do a little of this early, but in the more long- term, it comes back down to what an acceptable churn is for your business model. In a lot of cases, eventually you've got to get to what they call negative revenue churn in a SAS business. So, you don't need to be there even in the$ 50 million journey, I would say, if you're flirting with 1% or so, you're in good shape, but that's going to be the end outcome for you. And then what happens is, again, the danger of churn is not only one of the most important metrics for these businesses these days, but it's also a lagging indicator and somewhat of a silent killer. So you've got to develop a comfort around a leading indicator that you know about within the first 30, 60, 90 days of a customer life cycle that you have confidence will correlate with whatever churn you're solving for, okay? So I'll give you a specific example. In HubSpot, the marketing product years ago, the correlation for us was the product was pretty broad. There's probably 17 different features that you could actually use from the analytics system to the blog and the social media tracking tool, et cetera. We found that if someone used five of those features within the first 90 days of the customer life cycle, that that correlated with really great retention, and if they didn't, it correlated with poor retention. So, that was a great example of a very tangible metric. It's hard to tell a customer success team get churned below 1%. It's hard to just feel that. It's more tactical to say, " Get all of your customers using five apps within the first 90 days." And so we'd done that analysis and now that was a really nice short term goal that we could rally around. And because we had grown and scaled, we had statistical confidence that if we did that, the lagging indicator would be there for us. So it's a little bit of this. Maybe it's something like usage in the first 60 days or something like that. And then over time, make sure you validate that your theory was correct.
KK: Awesome. Well, join me in thanking Mark for coming by.
Mark Roberge: Thank you inaudible.