Dharmesh Shah, CTO of HubSpot at Business of Software 2011. Building Big Ass Software Businesses.

1 comment

Dharmesh Shah is the founder and CTO of HubSpot, a venture-backed software company offering a hosted software service for inbound marketing. Prior to HubSpot, Dharmesh was the founder and CEO of Pyramid Digital Solutions. Pyramid was a three time recipient of the Inc. 500 award and was acquired by SunGard data Systems in 2005. Dharmesh is also the author of OnStartUps.com, a top-ranking startup blog with over 20,000 subscribers and 100,000 members in it’s online community. He has a B.S. in Computer Science from the UAB and an M.S. in the Management of Technology from MIT.

Dharmesh is also the co-author of the book Inbound Marketing: Get Found Using Google, Social Media and Blogs.

Dharmesh talks quickly and packs a huge amount of insight and data into all of his talks. This one is no exception. Here are a few of the highlights, quotes and things you will learn if you watch this video:

BoS talks to your inbox.

Don't miss another talk! Leave your email address (unsub anytime and we'll never sell your details).

Be the first to hear when new BoS talks go live
Hear about chances to talk to speakers online to answer questions you have about their talks
Reminders so you never miss a conference again
Hear about scholarships and special offers to attend BoS Europe and BoS USA

For more like this, visit the BLN’s talk library and find seven years of insights on building a great tech business.

Golf tips from Dharmesh

Don’t play golf. (Unless you enjoy it).

When Dharmesh set up his first business, an enterprise sales guy told him he would have to play golf. That sucked as he hates golf, but he discovered, you don’t have to play golf to be successful as a software entrepreneur. Take people for meals instead.

“I’ve been doing meals, all the time, every week. Every city that I’ve traveled into, I will randomly send emails out to folks, either to people that I know or people that I don’t know, and I have a specific pattern to it now. I figured out I like round tables, I like acoustics that are good (so you can have a conversation), so I’m trying to reproduce the golf environment without the frustration and I can trace back millions of dollars, millions of dollars of value, to conversations that would never have happened had it not been for that meal policy.”

You are a unique snowflake

“The reality is, you’re a unique, precious little snowflake just like friggin’ everyone else, get over yourself. Because what happens, essentially, I’ve talked to hundreds of entrepreneurs now, hundreds of people that are in the software business. We make remarkably similar mistakes. So, if you’re unique, you’re not using that creativity on the kinds of mistakes you make, I promise you. You’re using it somewhere else. You’re making the same friggin’ mistakes that everybody else is making.”

SaaS and Subscription Business Models

In the subscription business, one thing you learn — my sympathies for those that are not in the subscription business yet; you’re missing out; it’s awesome you should move over — is that the cancellation rate is likely the most important thing that will impact your

Customer Churn defines long-term sustainability

Understand the differences between

  • Revenue churn rates
  • Discretionary churn rates
  • Involuntary churn
  • Negative churn – how the heck can you have negative churn. Simple!

Customer Happiness Index

Used to be to predict customer churn. Hubspot now uses it to predict which users will become marketing superstars. This is a far better measure of the likely value of a customer to you.

“So not only do we tell our customers, “Here’s how you’re doing from a marketing perspectives”, we have leaderboards now. Right? It’s like, “Here’s how you’re doing compared to other people that started at HubSpot the same month.”

Freemium vs Cheapium Models

Why cheapium models get a better class of users.

Raising prices in SaaS businesses.

“One thing they did not account for at Netflix for example, and this is the one lesson that we have, especially in the SaaS world now, is you have to assume that of all your customers are connected and united. Because they are. So, if you go in the back room and calculate pricing chances. What it is doesn’t account for is the fact that when the people get pissed off… everyone has a microphone now and they can unite… and it’s super, super painful. Netflix learned this lesson about as hard as you can learn it.”

At what point you need humans to sell software?

Why does enterprise software sucks from a startup’s perspective?

Strategy

“Whenever I hear the word “strategic” it’s like, “Oh, we did this for… We did this acquisition infrastructure for strategic reasons.” It’s basically, “We really just couldn’t justify the math…””

MBAs

“The running joke on Silicon Valley and startups everywhere is, if you want to try and value a startup, you know, it’s about a million dollars of value for every engineer and subtract $250,000 for every MBA on the team. As it turns out, of the first ten people at HubSpot, eight — including me, including all the management team, are all MBAs.”

Recipes for failure

“The closest recipe for failure is taking a business guy that trivializes the value of product creation and engineering and design and those kind of things.”

Oops

Dharmesh tells the story about when he suggested, over dinner, Drew Houston shouldn’t pursue the idea for DropBox.

The ‘Secret’ of success

The secret of success is to work your collective asses off. Dharmesh uses the experience of an early HubSpot employee and Olympic Hockey player to show the difference between sport and business, and a big lesson that business can learn from the sports world.

You can read a full transcript of the talk below.

Dharmesh Shah

All right. So you can tell I’m not quite a professional speaker yet because I still don’t bother with these little things like winding down my shirt, you’re supposed to do… Hello everyone!

This is… Always fun to be here at the Business of Software conference. My topic today, I made this up late last night, is building badass software businesses. So I’ll talk you through it. I’m Dharmesh. For everybody on the Internet too—I guess we’re Livestreaming this—guys, hello, thanks for joining.

And usually when I do these sessions I open with an apology. An apology about for public speaking skills, because I’m a coder, I write code every night, I don’t speak for a living…

This time I’m going to aim higher. I’m not going to apologize for my lack of presentation skills, and I’ve been practicing. So let’s see how that works out. So no apologies this time.

And, the folks that are watching on Livestream just so we can do this even next year, like, make some noise, make some noise on the internet so that we can go back to Neil and Mark – the guys from the Business of Software and say, “Look at all the attention that we got as a result of making this available on the internet.” So, make some noise, folks.

One thing that I’ve learned about presentation skills…

[Laughter, clapping]

…is that there’s some value to, like, breaking the ice, because my first five minutes are always the worst because I’m nervous. So, I’m actually going to have two jokes. I thought this one might go over pretty well because it’s kind of a localized joke. This one actually is better, and it’s also localized.

So earlier today… [Laughter] …I would say… Mark made this casual comment. [Laughter, applause, whistling] And so this is my joke. Fresh baked this afternoon. I’m going to let you ponder on it for just two seconds because I was proud of myself. It’s really deep in the Inception movie kind of way if you just think of it for just a little while. It’s like multiple layers…OK. So this gives people a chance to walk in, which they have, so I’m going to run through this. I had hoped to create this cohesive narrative, this story, which is the other thing you are learning about which I’ve read about how to do great presentations. I feel to do that, in terms kind of bringing a cohesive story to this thing, so what this is essentially…

How many people have heard me before at Business of Software either online or here? Most of you? OK. That’s awesome because I’m not going to be repetitive. I haven’t written a bestselling book, so I don’t, like, put the same brilliant messages out there. This is all fresh. It’s all new, both good and bad. So, I’m going to talk about things I’ve learned since the last one or about things that I didn’t get a chance to talk about last time, so most of this stuff should be relatively new to you. I mean, it’s new to me!

So one thing that’s happened the last year, obviously, is Steve Jobs passing away. Lots of people have had personal anecdotes so I’m going to give you a very quick story. So my first Apple product was the Apple IIe. Back when I bought a used one, I’d mid-eighties-ish, maybe it’s ’84-’85, somewhere in that range. And for the subsequent 20+ years I didn’t spend a single penny with Apple. Now, one: (but I was and am a huge Steve Jobs fan) and the reason that I’m going to talk about this, the reason I’m a huge Steve Jobs fan, is that he’s had a major influence on a lot of us—including me, including HubSpot, which I’m going to talk about—But that’s my baby, another thing that happened is that I’m a father now.

[Applause]

Name’s Sohan and I’ve said this before now. So I had kitten photos in my presentations prior to becoming a father. Babies work much better than kittens. It’s like… On the spectrum, in terms of audience empathy, it’s hard to beat babies, as it turns out. So…

What was interesting about Steve was, in addition to creating great products and the reason I didn’t buy Apple products was not this philosophical resistance to Steve and his approach at Apple, was around I wasn’t a target customer.

Like he was making things for mere mortals, and I didn’t think I liked him from a technical perspective — I’m a mere mortal, obviously — but he actually helped us believe in this power of innovation, of doing something dramatically new, and going after the non-consumers as Clay said earlier today, right, the idea is go out for those green fields, ideas that no one else has tried before and that it can work and it does work. And he obviously was a genius and Apple’s an awesome company, but I think that’s a wealthily powerful message, and I’m not going to talk about what HubSpot does that much at all because I’m not a sales guy, I say this every time, I’m not here to sell you HubSpot at all.

But when we started HubSpot the idea was around this integration story inspired by Clay Christensen who spent lots of time with and in small groups-was around non-consumers. Essentially, like, how do we, as a software business in the marketing space — which we happen to be — pull these non-consumers into the market?

So, kind of closing Steve Job’s story here, we have… When we started adding conference rooms, which is probably just nine months after the company started, because it was originally just the two of us and there were four of us and there were six of us, and it was just a great, big room, we just got bigger rooms. There were no conference rooms at all at HubSpot for a long time. When we started putting conference rooms in, we named them after people that had inspired the company and inspired the founders, and so our first room was Jobs. And it’s our boardroom now at HubSpot. Anyway, that’s where… I actually cried that night, so I just… I’m not going to get emotional here because that’s the stupidest thing to do in a professional business presentation for software guys.

[Laughter]

OK. So, I’m not going to talk about HubSpot and how we’re a marketing software company. This I’m going to put up here because I put it up every year, and I put it up here for two reasons: one is, to brag. So, this is the HubSpot customer growth chart and the revenue growth, and that’s the axis got cut off there? It’s not cut off there. So, we’re over 5000 customers. If you plotted the revenue growth… By the way, we call customers people that pay us money. It’s not a proxy for something else, it’s people that pay us money.

So, the revenue growth it’s actually steeper because the average revenue per customer has historically gone up over time at HubSpot. So, the business is doing well. And part of why I share that other than to just brag is to, you know, build some credibility with you folks. And we’ve learned a lot of things along the way, which I’m going to share with you both from HubSpot, and then we’ll talk a little bit about folks I’ve learned from. It sort of works. The company’s doing really well. We’re about 290 people here in the other Cambridge, a few miles away. And it’s going remarkably well. It’s… Whatever it is, it’s sort of working. We had lots of luck along the way, but I think the smart folks at HubSpot got a few things right.

So, we were voted the second-fastest-growing software company here in the United States by INC Magazine for privately-held companies, for those that are into that kind of thing. And the other thing I do other than HubSpot — this is just me going through the warmup sequence; it gets better, you know that if you’ve watched me — is, I write this blog called “On Startups,” and On Startups has been immensely popular. It’s not clear why yet. I think the right place, right audience. But one of the things that’s interesting, a little hack I want to share with you, I thought about what kinds of topics do I want to cover. I have some super-geeky topics, in terms of like RAW, like looking at metrics and things like that, we have some of those. I’m going to share one personal hack with you, because this one started early in my software career. And very few people have heard this story. It relates to that last bullet point.

So, when I first started in the software business, my first startup was ’94-ish. My co-founder and I, I was 23 and a half, my co-founder was 17, and we were in the enterprise software business and selling, like, big pieces of software to banks and mutual funds and life insurance companies for half-million dollars, somewhere in that range. We talked to each other and we didn’t have a whole lot of mentors because we were still young and we didn’t know what mentors were; we didn’t have access to very many people. But we did know one salesperson that worked at the company that we had, left prior to starting the company. And he said, “OK, if you’re going to sell enterprise software, you needed to play golf. Because that’s where… If it costs more than $100,000, that deal gets done on the golf course.”

And we’re like, “Well, that kind of sucks”, because we’ve never played golf. From what we know of it, it’s a very frustrating game; it takes a long time to get good at it, and we’re super, hyper competitive, and it’s expensive. So, we went to this analysis in our heads and we’re like, “OK, well, if we don’t play golf, that means there’s some greater than zero percent of the population that also doesn’t play golf.” We’re geeky guys that way. And so, we decided that day, essentially, to say, “You know what? We’re not going to pick up golf and frustrate ourselves. We’re going to do meals.”

And our decision was, at any point any one of us there were only two of us, but this carried through 10+ years, anyone could buy anyone else a meal. Potential customer, customer, another employee, partner, random person I ran into on the street, didn’t matter. That was a one line item on the expense sheet that we never questioned. And so, we carried that through for a long time and I carried it when I went to grad school and at IT; I would do the same thing. Like, I was a grad student, I would walk down the halls, and it’s like, “Oh, are you hungry?” As it turns out, students are hungry, more often than not.

And I would make up any excuse, which I did in my professional… I had different excuses, “If you’re hungry, I’ve got this expense account.” You can make that up. It was a small lie. I think… Sorry, Jason, for the lying part.

[Laughter]

But it’s worked. And so what I’ve been doing for now — I guess, wow, it’s for 15+ years. 27, 17 years? — it’s been doing meals, all the time, every week. Every city that I’ve traveled into, I will randomly send emails out to folks, either to people that I know or people that I don’t know, and I have a specific pattern to it now. I figured out I like round tables, I like acoustics that are good (so you can have a conversation), so I’m trying to reproduce the golf environment without the frustration and…

[Laughter]

…the competitiveness. And it’s worked out blindingly well. So, I can trace back millions of dollars, millions of dollars of value, to conversations that would never have happened had it not been for that meal policy. And the reason I tell you this story is that a lot of what I’m willing to share with you — I’m not going to tell you details, because, obviously, most of this stuff is shared in confidence — but a lot of what I’ve learned comes from those kinds of meals. Essentially, the On Startup meals.

So, thanks for everyone that’s participated in those things. And this is the only way I would name drop and say, “Here’s all the cool people I’ve had meals with.” They’re really cool people, trust me. It’s… I might drop a few names along the way, the book has done reasonably well, all right. So, it’s enough about me. Let’s talk about you folks. And all of you are sitting there…

[Laughter]

…and you look at… And this precious, unique snowflake, “I have my own set of problems and our business models are different, and we’re doing this unique thing on pricing or whatever.” And the reality is, you’re a unique, precious little snowflake just like friggin’ everyone else, get over yourself. Because what happens, essentially, I’ve talked to hundreds of entrepreneurs now, hundreds of people that are in the software business. We make remarkably similar mistakes. So, if you’re unique, you’re not using that creativity on the kinds of mistakes you make, I promise you. You’re using it somewhere else. You’re making the same friggin’ mistakes that everybody else is making.

So, we’re going to run through some stuff. And the reality is, I haven’t figured it out all either. Most of the time I have no idea what I’m doing. All right. So, this one, last opening slide here, is that normally I like to talk about startups and entrepreneurship and how to make your startup better and things like that, we’ll talk about founder issues and financing and things like that. And by the way, HubSpot raised another $32 million since the last time to bring the total to $65 million of venture capital in terms of updates, so… I know something about raising venture capital now. Who would have thought it?

But… So, my goal here is that… So, as I go through these things I’m going to be somewhat specific. So, I’m going to be talking more about B-to-B than B-to-C, essentially. I’m not talking about mobile apps. Not that I don’t like them or care about them, but that’s my context. But my hope is that the stuff I’m going to be talking about is going to be relevant to you whether it’s not this company or the next one, because you guys will go on to do different things.

It’s just… Anecdotally… Very few entrepreneurs stop at the first one. It’s… All right. Jumping right in. We’re going to talk about customers. And I talked last year about the importance of creating customers — not just creating customers, obviously, — but creating delighted customers. And I’m going to reiterate one point that I think was important from the last one. One of the most important things I said is that the goal is not to make customers happy or delighted. The goal is to make happy customers. Right? So, making customers happy sounds like an after-the-effect kind of thing. It’s like, “Oh, I have these customers and then I have my support team and other people that are there to kind of make them happy, we’re going to pick features or whatever.” No. To manufacture happy customers, which is the purpose of the business, you start from the very, very, beginning, from how you do your marketing, how you do your sales. Because if you end up signing a sub-optimal customer, you have already put a ceiling on how happy that customer is going to be simply by virtue of choosing them. All right. So, I’m going to dig into this one a little more because it’s subtle, but it’s important. We’ll come back for that one. The other thing is, most… How many people are in the subscription-software business of some form? Most of you. How many people are in the enterprise software… Big ticket…OK. And how many people just won’t raise their hand for anything? Doesn’t really matter. OK.

[Laughter]

So, in the subscription business, one thing you learn, — my sympathies for those that are not in the subscription business yet; you’re missing out; it’s awesome you should move over — is that the cancellation rate is likely the most important thing that will impact your long-term sustainability, survivability, successibility, which is not really a word, is thinking through cancellations. Well call them churn rates, because that’s the industry jargon. That’s what… Especially VCs will ask you about your churn rate.

And we’re going to dig into this because unless you’ve been around it, unless you’ve done a bunch of runs of capital or have read a lot or whatever, churn rate actually means different things to different people. I’m going to walk you through the types of churn. Because just from a raw here’s how we should talk about it perspective… the kinds of churn. These are not mutually exclusive. The first is customer churn. And customer churn is the number of customers that canceled on a percentage basis. Let’s say we’re doing a month-to-month business month after month.

So, if you started the month with a hundred customers and five of them left, you have a 5% churn rate, right? That’s roughly simple. Revenue churn is essentially… Let’s say you had $200,000 of revenue, starting the month, and customers representing $20,000 left. You have a 10% churn rate. Right? So the difference here is that the number of customers… So, your revenue churn could be much higher or lower than your customer churn based on what kinds of customers are canceling. Right? So, you might have a 2% customer churn rate — 1% customer churn rate — but your revenue churn is super high because the ones that you happen to be losing are the ones that are paying the most money. Did I make sense? OK. By the way, pause me along the way if you have questions. Otherwise, I’m going to run through this which will be fine, we can chat in the halls later. I’m going to pass out immediately after this, but then I’ll come back and chat with you folks.

So, then we’re going to talk about discretionary churn. And discretionary churn… So, if you’re in the subscription software business, many of you, either now or in time, will likely have options for the customer. “Oh, you can buy any old contract or pay upfront and get a 15, 16, 17, 18, 19% discount.” Or something like that. And discretionary churn really measures churn based on the number of people that had the option to cancel. This is the number in the denominator. So, let’s say we started with those 100 customers, right, let’s say… try to do the math here… Let’s say 50 of them were on annual contract.

And they didn’t have the option to cancel. Right? It was not at their discretion. They couldn’t cancel even if they wanted to. They were on an annual contract, they’ve already paid you money. They’re not allowed to cancel. And then of those, let’s say fifty people – ten people canceled… five people canceled, it’ll make the math easy, let’s say 10% discretionary churn rate, because of the customers that had the option to cancel, same customers or whatever, you shouldn’t get credit for the people that didn’t cancel, because they didn’t have the option to cancel anyway.

Right? Makes sense. So you need to… And different kinds of churn makes sense when you’re trying to optimize different parts of the business. Right? An involuntary churn is when customers didn’t mean to cancel, in theory, and canceled by virtue of their credit card didn’t go through, they changed credit cards… something happened; you weren’t able to process the transaction. You tried to bill them for their monthly subscription fee. That monthly billing transaction failed. The net impact to your business, if it continues to fail, is exactly the same. They churn… They didn’t mean to churn. They didn’t make an active “I want to cancel.” And we can do… I’m not exactly sure what the industry stats are on how many people canceled ended up being through involuntary churn or the other way around. Anyway, these are important.

When you’re talking to, let’s say, VCs or the industry or you’re going to go public someday, the churn, if you read public statements as I do sometimes because I’m that kind of guy, the churn number they’re talking about in the docs that are publicly available are usually revenue churn. They’re talking about how much revenue did the company lose based on the revenue they started with, of their subscription revenue, essentially. And that’s an interesting way to approach that. I’m not going to say it’s right, wrong, or indifferent, but it’s important to understand. So, if you have revenue churn — and this is the weird thing — if you sell upgrades to those same customers, so let’s say you had a customer, some percentage of customers went from paying you $50 a month to $100 a month, that reduces your revenue churn rate because upgrades count as a positive against the cancellation rate from a revenue perspective.

So, as you might expect, and it’s actually so mathematically possible and happens that some publicly traded companies have negative churn rates. Right? So you would think that’s not possible. How is it possible to have negative churn? Well, the reason is, their upgrades are exceeding the amount of revenue they’re losing through cancellations. These have lots of up, we’re going to talk about why that’s important, all right.

So, the other thing I’ve talked about a lot over probably the last two or three years, we have a number at HubSpot called the “Customer Happiness Index”, and this Customer Happiness Index is basically a single number that quantifies, in our opinion — not in our opinion; based on the data — and adjusted the algorithm over time to what degree a customer is happy, essentially. Right?

When we started building the Customer Happiness Index, which we’ve had since our first year, and we’ve been building for a while, the original goal of the algorithm was simply as a predictor of churn. And I talked about this last year. And it was awesome for that. So the idea is we can go through all of our 5000 customers, we can run the customer happiness index report and we can tell you, “Here are the customers that are most likely to cancel.” It’s a probabilistic model, which has been great for us. And then we had lots of internal debate. It’s been useful and we still do that, but that Customer Happiness Index as it exists now at HubSpot, the big change that we made is instead of solving for predicting churn, the algorithm now attempts to predict the degree to which a customer will be successful at extracting value from HubSpot and being marketing superstars, which is what we’re trying to build. Now, this one’s a subtle one, right?

So, economically, measuring churn, awfully useful, right, the downside — the reason we changed it to looking at… our customers actually… are they doing the right marketing things? Because if they fail at marketing, regardless of what we do, they will be out of… it won’t work, right? So it could be that they are extremely happy and doing the exact, wrong things. They can be logging 14 times a day and not doing the right things. Right?

So, just calculating their likelihood to churn, intentionally churn, is not that revealing. So, what we do now is we try to measure the kinds of things that we think correlate with marketing success which is the value we’re trying to build that says, “OK, well, are you writing a blog? If so, how much traction is it getting? Is it getting comments or not?” Because these are the things that we care about. And so, those things play a much larger role in the algorithm. These are the updates since last year in terms of what’s changed. So, we… I would encourage you to do at least achieve 1.0 customer happiness index and just try to measure and predict churn. This next one is still, the jury’s still out, I happen to believe in it, thankfully, but anyway, so that’s what’s changed, super useful thing for us, a continuous to supply large. One of the questions online that got posted when I made the request for questions was: “Is CHI still important?” Thanks, for the question. Is CHI still important? Do we still use it? The answer is yes, and we’ve tweeked it now to be this…

The other thing, I’m going to skip by this slide just so you don’t try to read it because it doesn’t have anything worth reading, there’s no revealing data or anything like that. So, when we made the switch to CHI one of the reasons that was coming back is that, you know, our VP of customer happiness, he essentially came back and said, “Well, Dharmesh, we’re measuring CHI, we have all these reports and management team looks at it, customer support team looks at it, sales team looks at it, everyone looks at this number of sales, people have commented on it. That’s all awesome, but it doesn’t really measure whether the customers are doing well or not on marketing.”

And the other thing is it’s a very insular thing. We didn’t share with our customers and say, “By the way, your CHI is only 87. That means you have a 13% probability of canceling next month. You don’t know this yet, but you have a 13% probability of canceling next month.” We didn’t… It was not a transparent thing that we exposed to our customers.

The new one is… So the new one now… So not only do we tell our customers, “Here’s how you’re doing from a marketing perspectives”, we have leaderboards now. Right? It’s like, “Here’s how you’re doing compared to other people that started at HubSpot the same month.” And the model — the metaphor that we’re using which is working really well for us is an academic one because a lot of us came from grad school recently.

But the idea here is to treat each group of customers that starts in a given month as a cohort. So, imagine it’s a class that got started in grad school, undergrad or whatever it is. And those people are going to move through the kind of on board, learn HubSpot process and then they’re going to graduate, let’s say, in 90 days or so. And so now, we measure. It’s like, “Here’s how you’re doing compared to other people that are essentially the same stage as you.” And we’re trying to build this community around cohorts as they move through and try to get educated and go. And that’s working really well. It’s…

This was said earlier today, so I added this slide. Thanks, Alex. The other thing we’ve learned is that it’s very, very dangerous to get married to just about anything about your business too early on, either target customer or product, pricing, feature set, whatever it is. We’ve changed a bunch, so my thought to you here is that: don’t get emotionally attached other than your grand vision or things like that, which is awesome to have one, but the tactics around the business and how you run it don’t get too easily married to specific parts of the business, because you will learn things. They will be surprising, and if you have too much emotional attachment, you tend to be less apt to change. That’s something we’ve learned.

I had this long — probably 20-ish odd minutes — segment of this presentation on pricing, and then… but… The CEO of Twilio, Jeff, I think, is speaking on pricing, so I’ve moved those slides to the end. If we have time left over I’m be happy to chat about them. But he’s going to spend a lot of time talking about pricing, so I don’t want to be duplicative. I want to share a couple of things with you that I don’t think will be duplicative. One is the brilliant insight, you know, going through two years of MIT, we had a pricing class. Is this from one of our marketing professors, who’s a brilliant, brilliant, brilliant guy. He’s from New Zealand, super smart — those two things are not necessarily correlated, just saying, it’s an observation, he’s from New Zealand…

[Laughter]

But he’s super smart. He doesn’t just sound smart because he’s got a New Zealand accent, he’s actually really smart. And it’s that, this is really hard, so of all the time the management team at HubSpot has spent debating issues, which we do a lot, with rigor and lots of hand-waving, pricing is the most hotly-debated topic. It’s the one we’ve spent the most number of hours on. Pricing and impact because usually it’s kind of two things that go together. And it is super hard, super hard to get right.

And as it turns out, the price is always… So, you always look at someone else and it’s like, “Wow, they have, you know, they provide only a 12% discount for cash,” or something like that. There’s a bunch of different things you’ll look at. You’ll drive yourself crazy if you try to analyze other peoples’ websites and say, “Well, here’s how we need to do it.” You need to test it yourself, because yeah, there’s some data out there, but you need to test that yourself, in terms of what works for pricing.

So, this is where the brilliant pricing slides would be. I still have them. Let’s see… We’re actually doing pretty good on time. I’m going to take a pause so I don’t pass out… Everything OK so far? Nodding your heads. OK. All right?

So, I want to spend a little bit of time on one of the topics we’ve debated a lot at HubSpot. A lot of you may be considering or already doing is the freemium option. How many people have a freemium? Which is, you have a free version of the product and then you how some people upgrade, which is awesome, and it works.

The most common resistance to freemium from SaaS companies usually kind of falls into a general set of buckets, one of which is: it actually costs us money to support the free users and that makes it really hard. So our upgrade rate has to be x percent in order for us to afford the servers and the support time and all those things, which is…

The other one is, well, it devalues the product because we have a free version of it then people are like, “Well, you know, is the paid version that much better than the free version? Do I really care that much?” Those are part of, there is a bunch of them. From one of the thoughts that I had, I’m exploring this, we haven’t tried it yet, is a notion of thinking of as cheapium. I’m making this up and cheapium… The idea here is to dress some of those kind of big concerns. One is… so imagine if… let’s say you are… Dropbox is a bad example because they have 240 million dollars in revenue, 4% uptake rate, they are doing super-well… But let assume you are having problems, you know, getting upgrades because you couldn’t afford the free users…

You could make your cheapium algorithm, the price that you charge equal to your cost of being sold. For instance, let’s say you are charging 100 dollars a month but it actually costs you, so let’s see, do the numbers, which you should, let’s say it costs you 4 dollars and 40 cents to support the customer, right? That’s the bandwidth we’re using, Amazon EC2 hosting costs, we have two people in support and they serve X numbers of customers and you do the financial model and you figure out your actual cost of getting sold.

Not what it cost you to build the product, right, the actual cost from the finance perspective, become marginal cost. Oh, if I add another customer, how much is that additional customer costing? Not existing ones. Just the new ones in order to get that one customer and make your price equal to that cost of getting sold. So you could charge 4 dollars or 5 dollars a month or whatever it is. The idea here is the upside is that it then gives you the ability to have literally an infinite number of free users and kind to figure out a way till the upgrade rate kind of picks up on … The other incidental benefit is, you get a better class of users.

I’m not trying to pass moral judgment in terms of… choosing customers, the kind of people that pay you 4 dollars a month, whatever your price, your cheapium price, are likely going to be better long-term customers that will pay you money for a longer time on average, right? So, you get a better group of customers into the funnel to start with. That’s something I think too many of us resist the freemium model for, you know, all kind of good reasons. I think this addresses some of them, so something to think about.

The other thing we’ve learned, fortunately we didn’t learn this the hard way, we stand on the shoulders of giants and people that learned the really hard way, and so there’s lots examples you could do searches and Zendesk put up there because I happened to know the CEO, spend time with him personally, went up to his office, and asked him what the hell happened. Because he’s a super-smart guy, very high emotional IQ, he is not by any stretch of the imagination naive or stupid. Not by any stretch. And so what Zendesk ended up doing essentially is raising prices, right? They did the internal math like… we are going to … and it really, really, really blew up on them to the degree that they ultimately ended to retract that price increase, which kind of adds insult to injury. Cus what ended up happening, they’ve got all that bad karma and bad press and bad everything.

And after all is said and done they weren’t able to rip the benefits of the price increasing because they have to retract it anyway. So, it was like super-painful. I want you to viscerally feel that pain because we are going to… yeah… it’s super-painful. So, next time you are thinking about pricing changes we’ll talk about what can be done to fix those. And then a much bigger more public example is the Netflix price change and business change and things like that. So, this is the public stock price for Netflix and I… If you look at the chart of the NASDAQ or any kind of… Public markets have been volatile, but they haven’t fallen off a cliff like this, right? So, Netflix destroyed 4+ billion dollars of value no matter how you slice it as a result of this decision, destroyed shareholder value, right? Billions of dollars.

And the reason it turned out that way… Once again, very smart management. One thing they did not account for, and this is the one lesson that we have, especially in the SaaS world now, is you have to assume that of all your customers are connected and united. Because they are. So, if you go in the back room and calculate pricing chances, it’s: “Oh, well, we are going to change the price by this, we did a survey, you know, 5000, 500 customers and we expect 8.2% of our subscribers to cancel as a result of the subscription change or whatever, and then we work that into an economic model and say oh well, if we changed the price by x, more then pays off for the loss of subscribers, this is a good business decision, increases shareholder value”.

What it is doesn’t account for is the fact that when the people get pissed off… everyone has a microphone now and they can unite… and it’s super, super painful. Netflix learned this lesson about as hard as you can learn it in terms of just of having… What was roughly a good stock, and you solved that price… It’s unfortunate.

So, if and when you get into the point where you are changing prices, which I encourage, I think it’s healthy thing to do because it’s unlikely that you’ve got the price right the first time regardless how much research you put into it, is to think through the implications of what would happen if everyone that knew anything about your brand was on the Internet and responded, essentially, with their…

So don’t… It’s not a survey of a thousand, it’s a survey of everyone essentially because that’s effectively what happened. And the Internet amplifies both happiness and anger. And so they got… So, don’t do that. And one very easy hack — and it’s expensive except compared to the alternative — is to grandfather or grandparent your existing customers. So, we have a rule at HubSpot, it’s been in place for five years, we’ve changed prices, increased prices, consistently relatively for 5 years maybe twice or so a year, we continue to do that step up and there’s lots of goodness that comes out of that, but we don’t screw existing customers.

So, whatever it is we do… and it’s like, “Oh, well, whatever you are paying for HubSpot right now even that was… your price does not change.” And that creates cost for us, not that we lost the upgrades. The reason it’s painful is because now we have this other calculation we have to do because there are other [skills] and product units that we have to track and…

But it’s completely worth it. So we did another big increase, you know, recently whatever once again when back to our existing customers and said this is what we are doing on pricing, this is why our new customers that sign up will be on the new pricing for obvious reasons and all of you essentially get to benefit from your early belief in our company and thank you for being loyal customers essentially. And that helps a lot. In two ways: one is it sends the signal our customers have watched us do this.

It’s a weird thing, I have to share one little tactical [note] with you because it’s interesting. And we are maniacal about metrics and we have I think 80 people on the sales team at HubSpot now and we tracked sales metrics like you would not believe like… lots of data, lots and lots of data. We’re maniacal about it. Every time we announce a price… and this is obvious now to me but it wasn’t obvious at the time, every time you announce a price increase, the sales go out the roof. Of course they do, right? This is your last time to get in locked into this pricing before we increase it next month.

So don’t make it a surprise. Put it out there, reward your existing customers and then that pays for a lot of the other stuff. So, when you change prices, grandparenting existing customers make it public how they make consistent approach to how you deal with pricing. I like increasing pricing because going down on pricing is actually much harder than going up. And the reason is when you go down on pricing you have to deal with cannibalization arguments inside the company. It’s like, “Oh, well, if we reduce it, then how many of our customers are going to want the lower price?” And the answer is all of them. And if you can’t do that, then…

So, what is happening you are almost never lower the prices because of this cannibalization, like your finance people will come back or you will have conversations with yourself if you are the finance person. And it’s like, “Oh, we can’t do that because the revenue will fall out the basement.” OK.

One of the other questions that came up recently which is a very good one is: at what point you need humans to sell software? So, ideally your business will look exactly like that, right? People come to your website at 3 a.m. on a  Sunday, whip out their credit card and start paying you money and life is good, right?

That’s the way. That’s like the dream all of us have. It’s like to have a server in our basement that’s getting 50 bazillion hits a day, whatever, some people convert the customers pay as credit cards, they are happy with the product and life goes on, that’s awesome.

As it turns out that’s hard in terms of selling and so the answer to this question… I think the only one with bullet points, but it’s warranted here and no pictures…. So these are the three reasons why you might have the need for humans to sell. One is the product itself is complex. None of these are counter-intuitive. They make sense I should probably talk through them… If the product is complex you likely need humans to help people understand just to get value out of the product. So, sales is not just closing deals, sales is education.

The market might be new, so people are not searching on Google, they are not looking, they don’t not wake up Monday morning and say oh, I want to buy enterprise security software for X or whatever, that’s quite a bad example maybe people do look for that, but whatever your market happens to be people if aren’t searching the market is just new it’s an evangelical sale, you are trying to educate the market, you are going to likely need humans to educate them in some form or fashion. And if the price is high, right, this is…

There’s no real good rule of thumb, but I think it starts to go much above a hundred dollars, it starts to get harder and harder to convince humans just to whip out their credit card and start paying you a hundred dollars a month. So, there’s a price… The guy from SurveyMonkey has a comment around: there’s like a point of indifference, which is around 30 to 40 dollars a month, is that there’s a point in there somewhere where humans seem to at least, in theory, seem to not care that much. So they’ll say, “Ah, it’s 30, 40 dollars a month. As long as I can cancel next month, I’ll go ahead and buy,” and they tend not to cancel, so there’s this magical point, you know, sub-50 dollar or somewhere. But as you start approaching 100 dollars, you see less of that happen. There’s usually some humans involved in the process.

And the way to get around this – not get around it, but the way to kind of address… If you do want to sell human-less essentially, and still have higher revenue per customer is to do something like a freemium or cheapium or a low-end product. So, they’re starting out paying you very little money, which is when they’re making the purchasing decision, you deliver value and there’s a natural upgrade path. Dropbox is a great example, right? So, it’s a freemium product, I think it’s 2 GB for free, and then after that point you kind of follow a certain percentage of the customers, just start getting value and they start paying more money.

Drew, I know really well. I actually have really good Drew story. I’ll see if I can… so the other thing you need to do, in terms of thinking through economics is: really get good at picking your customers. And this is going to sound platitudinal, but you have access to many more customers than you probably realize. And it’s super important. Oh, so it’s like, oh, so we talk a lot about like: here’s the right mix of features that we want in the product that customers are asking for.” And that’s important.

I’m not suggesting you shouldn’t spend time on that. It’s as important to decide just like you decide what features go into the product to decide what features go into the customers you’re selling to. What do they look like? How do they spend? Who’s doing the approval process? What causes them to use it or cancel it or whatever? You should have a detailed profile, just like you have for your product; this is what’s going into it and you should be making decisions just like you do about your product, and you should have a customer profile, just like you have your product roadmap, you should have a customer roadmap. Here’s how the customers looked when we sold a year ago. Here’s how we want them to look. Here’s things that we’ve learned when we sell to law firms, you know, they pay us a bunch of money but they end up canceling, because lawyers are the second-hardest segment to sell to, next to software developers. It’s hard.

So, anyway, you should understand that because it has massive, massive impact on the business, long-term. So, I’m going to take a little pause… OK. Yes, a question?

Sp2: Does this apply to Enterprise Software sales?

sp1: Yep. Yep. So the question is, if you’re in the enterprise business, where your revenues and growth are dependent on customer revenues, can you afford to be choosy, or do you just take what you can get in terms of customers?

And the answer varies. A lot of it is based on whether you’re boot-strapped or funded, right, so one of the downsides to having a boot-strapped self-funded company is that you can’t afford not to have revenue, because revenue drives growth. So, you tend to… It’s natural to conceive more, in terms of the customers that you sell to, and you shouldn’t conceive too much, but that’s… It’s a hard line. But this is one of the reasons.

So, when I wrote my thesis in graduate school, one of the things, one of the big “Aha!”s Once again I have an uncanny knack for the obvious, one big “aha!” was enterprise software sucks from a startup’s perspective. And this is one of the reasons it sucks, because you have so little control over your destiny, because you’re essentially beholden to convincing large scale customers with a very binary outcome. Right?

So, it’s like: “They either buy it for $200,000 or they don’t. It may take six months. We’re not sure if we’re, like, 90% of the way into the sales process,” which is what your salesperson will tell you, month after month, quarter after quarter, including in month ten. It’s like: “Oh, weren’t we 90% like sixty days ago?” Yeah… Anyway.

So, one of the advantages, I think, — and this is why my belief, if you’re doing a startup, if you have a choice, try and find a way, especially if you’re going to be self-funded, to start smaller. Essentially, where you have the ability to be more nimble, sales cycles are shorter, so you… It’s going to sound trite, but you do have the ability to pick your market , right? That’s one of the luxuries startups do have. So, my advice is to find a way to pick a smaller market instead of going after elephants, because that’s a very, very hard business. It’s very hard to be agile, it’s very hard to change, it’s very frustrating, and it’s just a non-fun existence.

Yeah?

sp3: XXXX customer that came XXXX

sp1: Yeah, totally. So, yeah. So, it’s… So this is the interesting thing… so, right, in the system dynamics model. So if you look at marketing in terms of, you know, pricing and positioning and product features and support and how… You know, do you have a 0800 number, what’s the support hours, how many people do you staff, what’s your SLA in terms of email responses. All of these things are related so when you’re sitting down and thinking about the startup, thinking about the software business. What you want to do is, you want to kind of start, to the degree possible, measuring it and saying, “OK, here’s what it costs us to support x amount of customers. When we raise prices by x, when we do y, like, this is the impact it has.” Because what happens if you do this correctly and most really big SaaS companies — several of them jump to mind — do this really, really well.

So, what you’re trying to do is build this cocktail, right, where you have all these knobs and dials  in the business. It says, “Oh, I want to turn up, I want to increase the number of support hours from, you know, eight hours a day to ten or twelve hours a day. And I’m going to run this experiment. I’m going to have that offering for, let’s say, a month or two months and see what comes back.”

So, what you want to do is you want to be able to kind of tweak the business. It’s one of the luxuries that SaaS businesses have, is, you can change lots of things about the product. You can test lots of things. You can measure what people are or not using.

Like, consider taking a feature that’s being debated within your company. It’s like: “We’re not exactly sure whether this particular feature is, like, useful or not”, and you can do usage studies, like: “Oh, yeah, well, all of our customers, 23% use that feature. If we took it out, 23% would be unhappy with the features gone.” Not true. You’re jumping to a conclusion. Take the feature out for your next 100 customers and see if… And watch that cohort, see if anything happens. So, when the feature was there, it was like, “Yeah, it’s there. Well, I might have just clicked on it just to see what was there, but I may not have cared that much.” And see what happens. Right?

So, you are allowed to have testings, you are allowed to take things out, put things in and run tests. And the best software companies do that exceptionally well. And that’s one of the advantages that we have of being in the SaaS business vs. enterprise software: we have to ship stuff and then pray that they’re using the product someday. And you’ll never know to what degree they’re using features, regardless of what they ask you. It’s like, “Oh, can you add these seven features to this particular area of the…” It’s like, “Have you used that?” Like, they won’t be able to tell you, this just sounded like a good idea to them, you know, so…

All right, so I’m going to talk about some other software stuff and… Whenever I hear the word “strategic” it’s like, “Oh, we did this for… We did this acquisition infrastructure for strategic reasons.” It’s basically, “We really just couldn’t justify the math…”

[Laughter]

“…But we think it’s a good idea.” Which is completely legit. I’m not disparaging strategy, but it’s… Yeah, you don’t have the data but you have really good intuition, hopefully. So, we’re going to talk through some of these things we’ve learned — I’m going to try to talk slower, I really talk too fast…

One of the things we’ve learned at HubSpot is, you know, we’ve got 290 employees now, 5000 customers, we’re reasonably smart guys and gals. It’s really hard to change things, and there’s like all the.. and the reason I use the words ‘donuts’ is because I’m a secular kind of guy… And it’s like that’s where the value is. It’s like you have to really say, day by day… Like, we like to idealize this kind of, it’s like, “Oh, I had this brilliant idea in the middle of the night, and then I went off and coded or whatever. And this happened and the world changed. It was like…” And that happens. Like, you will… Your day to day existence of just carefully crafting the business will be periodically interspersed with these moments of brilliant insight that actually move the business. But it doesn’t happen that often.

And so, you really need to get good with the details and get good at metrics. I’m not going to talk a lot about metrics because David, the CEO of Performable has an entire session coming up on Wednesday about building a data-driven business. As it turns out, he was our first acquisition, so David now works for HubSpot, so now he has, kind of, both sides of it, and he owns a product there.

A weird little twist of fate, like, there are three speakers now at HubSpot. I mean, three speakers that are from HubSpot at this year’s… And they weren’t at HubSpot… two acquisitions. Both of them are speaking here, for some reason. Completely…

sp4: ?

sp1: Yeah, exactly, yeah, yeah. And the other thing we’ve learned: this is a weird, weird thing, so we’re very data-geeky — very, very data-geeky — and I found this strange thing that happens. It’s, like, so we’ll be sitting around, and the first time it hit me it was like, “Oh, this can’t possibly be true.” What happens is that someone will present, it’s like, “Oh, we ran this particular test, and here’s the insight. But it can’t be right because it makes too much sense.” I’m like, “No, on average, most of the things you come up with if it mirrors a real world, if you have reasonably good intuition, it should be exactly what you thought.”

And every now and then you’ll come up with a counter-intuitive insight, like, “Oh, well, we thought x, but actually it was y. But that’s not the norm. So don’t be…” So what I find people doing is like: “Oh, well, yeah, we’ve ran these tests and whatever and we want some big, brilliant counter-intuitive insight that, like, people will pause and write books about and it’s like: ‘Wow! Those guys figured out that…’”

Yeah, that happens, but don’t be searching for that counter-intuition. Just search for being right. Right? That’s all the matters. So… Once you start looking at data, you will find this. It’s a weird psychological human thing. I do the same thing, I say, “Wow, I couldn’t write a blog article about that. That’s kind of boring. Let’s keep looking at the data until we find something counter-intuitive.”

So, let’s talk about people a little bit. We’ve grown a fair amount at HubSpot, and we have… And last time I talked a lot about culture. One of the things I want to talk about is people in culture. So, this is unique to HubSpot. I think it’s unique, actually.

So, the running joke on Silicon Valley and startups everywhere is, if you want to try and value a startup, you know, it’s about a million dollars of value for every engineer and subtract $250,000 for every MBA on the team.

[Laughter]

And as it turns out, of the first ten people at HubSpot, eight — including me, including all the management team, are all MBAs. And… so, I believe this in terms of MBAs are not intrinsically toxic. You know, and maybe you’re, like, saving one more soul from investment banking, so there’s… it’s a noble calling getting MBAs out of that path…

[Laughter]

…and putting them into startups. But the pattern that we’ve found, though, is that what’s worked for  HubSpot, when we did our early recruiting, early hiring just about everyone the we recruited and brought on didn’t come from that particular discipline. So our VP of sales has never been a VP of sales, or a director of sales, or a sales anything. Before, our VP of customer of has never been a VP of customers, a VP of customer support or anything like that.

So we essentially solved for two things. And Joel kind of characterizes which I think is relatively close, is smart and get stuff done, and so we think of this smart as just raw wattage. Because there’s two things that I’ve learned about the software business. One is, it’s a complicated business. There’s a lot of moving parts, it’s this weird mix of creativity and science. And average IQ doesn’t cut it at a certain scale, just doesn’t cut it. It’s a complex problem. Just like you wouldn’t expect folks with average IQ to come go off to med school. It’s a hard problem. It’s an intellectually challenging thing to build a software business. So you have to be reasonably smart. Now I’m not going to… I’m not that smart. Actually, I’m reasonably smart.

[Laughter]

Reasonably smart. So let’s talk about people a little bit more. The other thing we’ve learned is as a coder myself I always wonder, “What do business people actually…” Especially in the early days, like, “What do they people do, why do I need one?” And there’s a reason for this belief, because it’s mostly rooted in truth. So it’s much more likely that a coder hacker type, the classic hacker type, will learn the things they need to learn about business than the other way around.

So I’m an amateur angel investor, I’ve made 29 angel investments now, I just had to count because I couldn’t remember them all. And one of the things I’ve learned is that you find like a sole founder that says: “Oh, I’ve got this really great idea. And I’m going to outsource the development and it’s going to be really cheap to produce this app, or I’m going to go off and find a VP of engineering — not a co-founder, mind you — a VP of engineering, I go running, screaming in the other direction. Right?

That’s the closest recipe for failure is taking a business guy that trivializes the value of product creation and engineering and design and those kind of things. Which they sometimes they’re up to do. But it’s equally —not equally but close to equally as bad as for us to not value the stuff that goes into being a business geek, let’s say. It’s like, “Oh, I have this passion for customer metrics and things like that”. Not that you couldn’t do it; you’re smart, you’d be able to do it, but they wake up every morning working on this stuff. They’re smart, they have this natural intuition, a tolerance for a different kind of pain than you have essentially and it’s super useful. And it’s necessary to build big, sustainable kick-ass software businesses, as it turns out. So, you need both.

And, so I think we tend to discount business geeks too much in the technical geek community. So one thing that we’ve learned.

By the way, so this is a new to me and I usually talk about startups and the early stages and how they get started so most of this talk has been around. If you are out to build a multibillion dollar business and put a dent in the universe, how would you do it, so it’s intentional and that’s why I’m doing it and I’m going to…

And I’m not suggesting, obviously, that there’s anything wrong with independent software businesses that want to be that. Lifestyle businesses, I don’t think that’s a derogatory term at all, that’s completely OK. Although I will say I think I want Peldi after Balsamiq, which is awesome Peldi, great. I want him to go take down Adobe some day.

[Laughter, clapping]

Like he needs to. Noble calling he needs to do it, because they’ve had a good run, great software company in their day, but Photoshop and their integration with Illustrator, that stuff sucks!

[Laughter]

I’m sorry, it sucks. They have a monopoly. They need to be taken down. But anyway I’m talking about how to build big-assed… Or kick ass — big ass sounds kind of down but it’s… Software businesses. So one of the people that I respect the most… I have time, I’m going to tell you this story because where else would you hear this story…

So how many people know Dropbox, and Drew from Dropbox—he’s on the cover of Forbes by the way, this coming issue. OK. So Dropbox… so Drew from MIT, from Cambridge, this Cambridge, not the other Cambridge, and I’ve known him before he ever did his first startup. I’m going to tell you this story because it’s a good story. So I ran into Drew because he’s in the MIT community working for a company named BIT9, which is a bit low-level security company doing low level Windows internal stuff. And I had lunch with him a couple of blocks from my office and our IT blocks were within one mile radius, and I met him and he said, “Oh yeah, I’m working on this new startup idea”, and I go,”Oh, yeah? What is it?” I haven’t met him and he said “Well, it’s an SAT prep software to help people prepare for the SATs. It’s a standard test here for… in the United States.” And I said, “Drew, that’s a really bad idea. Please don’t do that.” And he listened, and he didn’t do that.

And then we kept in touch and we would have lunches and we would have dinners, we would like talk about startups, super-cool guy, and then he’s like,  “OK, I’m not doing the SAT prep thing. I’ve got this new idea.” And I go “Cool, what is it?’ He said ‘I’m going to do this kind of file syncing software so people can sync up multiple devices’, and I was like “Drew, please don’t do that.”

[Laughter]

He did it anyway, which is awesome. So, it’s 4 billion dollars in market cap of valuation. This is now in Forbes so I’m can actually talk about it which is nice. 240 million dollars in projected revenue, set up about the same time as HubSpot about six months, but it’s about ten times as successful a business on whatever metric of revenue that you look at. So it’s a supremely successful company. He’s an awesome, awesome entrepreneur. And I do not say that lightly. He’s great. So, a lot of the stuff and it’s been great watching him grow up,  you know 4-5 hours dinners and get drunk off our asses and chat about software companies. But he’s a cool guy. So one of the things that came out in the conversation is around this kind of working your asses off. I reckon there is no other more gentle way to put it. And the question is like how important is that? In lots of ways you’ll be asking yourself. What about work-life balance, like we the new age thing, he said, “Yeah. I want a family. I want this, or whatever, it’s not all about the company,” and that’s a reasonable question to be asking yourself. And I have a very controversial that’s about to get even more controversial because I said it now publicly and I’m going to write about it, but I have never, ever, ever met a super successful software company where the early team did not work their asses off. Not one! Not Facebook, not… Nobody. Nobody,— that didn’t work their asses off. Even the ones that got lucky, still worked their asses off. So the question is, well, to what degree is that necessary and how do you balance, I was going to talk about that.

This actually comes from Netflix, which is one of the best presentations on culture, and it’s about employee culture, not pricing culture obviously…

[Laughter]

…but in terms of team culture, and it’s a brilliant deck, you should look it up you should look for Netflix culture. You should go through it, and you should go through it multiple times as we have…

One of the points that they make in that deck is that they think of their team not as a family, but as a professional team. Like maybe an Olympic Team or an athletic team, whatever it is. And so they define this like: OK, there is this common goal, yes, we have respect for each other, yes we like being around each other, as teams need to be able to do, but this is a specific goal, it’s not family, and the reason it’s not family — they don’t say this, this is my interpretation — is that you get to pick your team, you don’t get to pick your family.

Right? I mean family happens to you. Good or bad, that’s luck of the draw.

The team you get to kind of choose. And so, you know, building companies, I think it’s… That particular metaphor works a lot better. So, I’ve been thinking a lot about this because I have a company meeting, all hands meeting on Thursday with all of HubSpot. And the scary thing is, as I’m looking across the room now, there are going to be more people in that company meeting than there are in this room, which is kind of scary. Or close to it, ah maybe not that much, but close. It’s like 290 people which we’re all mostly here. And so, I’ve been thinking about, like, how do I, you know, how do I talk about culture, how do I talk about this?

Like, we’re growing. It’s like, “Oh, does that mean it’s not necessary now to work as hard as the early team?” Or whatever. How does HubSpot think about work-life balance? We have unlimited vacation, lots of cultural attributes. And it’s squishy so, I’ve been kind of digging into this. Is it really, really, necessary? I’m a father now, to what degree does that, you know, impact my thinking and it’s like we’re at a different scale now, different thing. So, I dug in a little bit. So, this is a picture of the 1998 woman’s hockey team from United States that won the gold medal. And one of the people in that picture, — this is her, Colleen Coyne — is an employee at HubSpot, as it turns out. So we have a gold medalist… And she’s got a super fascinating story, which I’ll write about — I’m not going to talk about it here — but I asked her some of these questions. Like, once this kind of team thing went into my head, it’s like, “Oh, well, I know nothing about athletics at all, which should be obvious.

[Laughter]

And so, I asked her and I had this really long and I just actually might post an email of her response to my email, you know just verbatim; that might be an interesting thing. Because I was trying to ask her: “OK, well, how did you guys deal with this so you know, if you’re an Olympic team, my thinking is well, if you’re in the Olympics, and you’re there to win, — that’s why you’re there, right — and you train for this and it’s like, OK, well you expect everyone else in the team to have trained really hard, but the other thing is that everybody else, every other team there are by definition in the top. .01 percentile. That’s how they got there. So simply being somewhat talented — maybe exceptionally talented — that’s just a cover charge to get in. It’s not enough to win. Right? So I’m digging, so I asked her, without leading the witness too much, “Well, how important was it for you guys to work really hard?

And she had a couple observations that I think were fascinating. One was — and she’s been at HubSpot since year one, she was like one of the first ten employees, so she’s watched us grow and watched the culture evolve — and her statement was that one of the problems we have in the business world is in athletics: you have eight off-seasons when you’re not playing. You kind of thinking and … it’s the off-season for whatever sport it happens to be, usually.

And you have injuries. Like physical injures.  “Oh, I played too hard, I did this or whatever, I twisted an ankle. I can’t play.” And that is the world’s way of telling you ‘you shouldn’t play’. A measurably, visible surfaceable thing. And we don’t have that in business. We may have someone that’s completely burned out, that’s not productive, whatever, it’s damaging to them and the company and everything, but is not outwardly visible. You may see some outward symptoms, but there’s no way to definitely say, “You have a twisted ankle. Like, you need to take next month off”, or something like that. And so, this is one of the things that we struggle with. How would we know? How do we convey that people take enough downtime? So anyway, there’s a bunch of things out of that particular interview… One of the things out of that interview…

So I asked her the question, “When did you decide that you wanted to be on the Olympic team?” And my theory behind asking the question was, “OK, she’s really good, was born super-talented, and it’s like someday a recruiter came along and says, “I’ve seen, I’ve watched you play,” like you see in the movies.

And she’s like she had decided, and she publicly stated this, that she wanted to be on the Olympic hockey team when she was 16. They have it on video. At the time that she made that decision, there was no Olympic hockey team in the United States. An Olympic hockey team was not formed until ten years later from that point in time. And my point here is that companies do get successful, you do get to scale or whatever, but most of the successes that I’ve seen there is deliberation there. It’s hard to build a multibillion dollar big software company accidentally.

Like, I woke up one day and: “Wow! There’s a billion-dollar software company!” It doesn’t work out that way. And my point here is that as you’re working through this stuff, you need to make some deliberate decisions and there’s no wrong way. I mean, I’m just taking one extreme position. I have a bias because HubSpot happens to be trying to build this multimillion dollar juggernaut. But if you’re looking to do that, you have to make deliberate decisions based on that. That’s like, that’s the goal, that’s what we’re trying to accomplish. Because lots of things will fall out of that and if you behave like you’re not going on the Olympic… that’s not what you’re doing, you won’t get there. It’s, like, it’s very, very rare for someone that doesn’t design something to be this massive… based on how much capital you’ve raised, how fast you grow, and all those things. Anyway that’s my message there.

The other lesson, once again this is from Colleen is… So, when we started HubSpot we hired lots of generalists. Generalists are awesome in the early days of the company, and it’s like “Oh, you can do a bunch of different things.” As you grow, as we’ve grown, we’ve found that specialization starts to matter a lot more. And where it starts to matter is because you’re in a competitive market, you’re competing with something that does this for a living, they’re gifted, they’re world-class at x, and it’s difficult to compete with that with generalists. It’s like…

So, her statement was, she plays defense or whatever. She plays that position and she’s like: “I’ve had coaches ask me to do different things.” Like, “You have the physique and build.” And she’s like, “No. This is what I do, this is what I do best, and that’s what I, so this is what you need to start getting more specialists in.

So, I’m going to close on my baby again. So, I have 2 baby photos. First time ever in a single presentation. So, I want to close. It’s awesome being here. You guys are friends, so I see you guys year after year, I love this stuff, it’s an honor and privilege to speak. Thanks for your time.

[Applause]

BoS talks to your inbox.

Don't miss another talk! Leave your email address (unsub anytime and we'll never sell your details).

Be the first to hear when new BoS talks go live
Hear about chances to talk to speakers online to answer questions you have about their talks
Reminders so you never miss a conference again
Hear about scholarships and special offers to attend BoS Europe and BoS USA

One Response to “Dharmesh Shah, CTO of HubSpot at Business of Software 2011. Building Big Ass Software Businesses.”

  1. [...] Dharmesh, HubSpot, Onstartups. Building big … – Business of Software This entry was posted in Uncategorized by Dennis. Bookmark the [...]

Leave a Reply