/
RSS Feed
In episode 734, Rob Walling interviews Ian Landsman, founder of HelpSpot, about his 20-year bootstrapper journey. They discuss Ian’s transition from on-prem software to SaaS, the challenges and benefits of each, and the early days of building the business. They wrap up by discussing the potential impact of AI on the customer service industry.
Topics we cover:
- 1:11 – Ian, the OG bootstrapper
- 2:22 – Benefits of on-prem software in 2024
- 5:46 – Slow, steady, profitable growth through the years
- 9:20 – Embracing a risky start
- 14:11 – Getting early awareness
- 18:52 – Transitioning to SaaS
- 26:37 – Laravel raises $57M
- 28:59 – AI impact on customer service
Links from the Show:
- The SaaS Playbook
- TinySeed
- Ian Landsman (@ianlandsman) | X
- HelpSpot (@helpspot) | X
- HelpSpot
- Podscan
- Accel invests $57M into Laravel Products & Open-Source Framework
- Mostly Technical
If you have questions about starting or scaling a software business that you’d like for us to cover, please submit your question for an upcoming episode. We’d love to hear from you!
Subscribe & Review: iTunes | Spotify
Rob Walling:
Welcome to this episode of Startups of the Rest of Us. I’m Rob Walling and this week I talk with Ian Lansman, the founder of HelpSpot. Ian has been bootstrapping for almost 20 years and he started with on-Prem software and then after about 10 years launched a SaaS version. Ian and I have known each other almost that same 20 years from back in the days of Jolen software’s business of software forums. And over the years, Ian has started a few of his own podcasts as well as spoken at MicroConf. I believe I were to guess it was probably around 2014 or 15. So it’s a great conversation because Ian and I have history. I think we were able to pull out some really interesting bits about why he’s bootstrapped for 20 years, why he hasn’t sold and moved on to his next act, how he thinks AI is going to impact the customer service and support space. And if you stick around until the end, you’ll get to hear the company name that is most often confused with HelpSpot and it wasn’t the one that I thought. So with that, let’s dive into our conversation. Lansman, welcome to the show. Thanks for having me on. It’s great to be here. Cannot believe you have not been on this show. No, you’re like OG Bootstrapper.
Ian Landsman:
Somebody on Twitter reached out and said, I should come on here. And I was like, man, I think I’ve been on there. And then I searched and I was like, man, I haven’t been on there. Wow, that’s crazy.
Rob Walling:
Yeah, I am glad you reached out. I think someone, your podcast co-host maybe mentioned it on a podcast and I use arvid calls pod scan, and so my name popped up and I went and I’m like, wait, what? Someone wants to go out and start. So I was like, yeah, dude. And I thought the same thing. I was like, I think Ian’s been on it, but to give people context, like 20 year, I mean really early man back in the day, remember how it was, there’s going to be a lot of jokes about how back in the day, there was no customer development, there was no AWS, there was no SaaS right? Where
Ian Landsman:
Before all that, man, it’s hard to believe we’ve become the old timers. How did that happen?
Rob Walling:
Oh yeah, big time. Just so people know, you’re the founder of HelpSpot. It’s at helpspot.com, it your H one is your customer service at scale Amplify your support with HelpSpot, the streamlined solution for scaling customer service effectively. So it’s email ticketing, reporting and metrics knowledge base. And since you started so long ago, you were on-Prem for years and then you launched a SaaS. So you have both on-Prem and SaaS, and I think most folks listening know what on-Prem is, but it’s where they actually download your code, your source code, PHP, Laravel, and they will install it on their own server. And then I guess what’s the benefit? Why would today, I would just do the SaaS personally, but the people who still use on-Prem today, why do they do that rather than pay you for the
Ian Landsman:
SaaS? Yeah, so definitely the majority or SaaS using the SaaS version these days, but we still have new customers in the on-prem and basically there’s some advantages. Big companies or certain specialized companies like in finance, healthcare, they have different rules, potentially laws. They’re also the type of companies that still have IT departments. And so they want to be in control of their own data is usually the main reason in terms of they control the database and they’re backing it up and those things, but also sometimes they’re just running completely off the internet. It’s like, here’s our help desk, it’s for maybe IT in that case and it’s literally not connected to the internet. And so it’s just fully encapsulated in their network behind their firewall. And so they want to do that. And it is an area I think people have abandoned completely that maybe shouldn’t be totally abandoned because there are some, you could charge more. We actually charge the same price currently, but I think there’s a lot of opportunities there to charge a lot more for the people who want it on premise and things like that. So yeah, it’s not that bad if you have a modern SaaS app, it’s not impossible to make an on-premise version. Some things like if you depend on 30 services, that’s going to be hard, but with a little bit of forethought, it’s really not that bad.
Rob Walling:
We have a few TinySeed companies and I don’t know how many exactly, but when they apply they’ll often say, so we’re SaaS, but we also do on-prem. Is that going to be a deal breaker? And I was like, no, we funded a handful. And again, I don’t know if it’s five or if it’s 10 out of one 70, but there are especially like you said in finance or in certain regulated industries where you really do need or want control of that data and just went on some random spot.
Ian Landsman:
Yeah, it also gives you some advantages of you’re a tiny bootstrap company, you’re not going to be able to be SOC two compliant, for example, and there’s going to be these type of companies that are like, well, you’re SOC two. Let me talk to your security people, all this stuff. And it’s just like you, it’s you and a couple of people, and so this gives you some outs with those big companies. It’s like, Hey, you know what? You can install in your own SOC two data center and you’re already SOC two and you can, it’s all good and so we’ll charge you for that and everything. And those tend to be also some of the best customers. I mean those are the customers that stay with you for a decade. You’re in there, you’re on a server, just not even the type of companies that are in the mindset of like, oh, somebody launched a new fancier version of this tool we use. Let’s go out and find the fancy. Let’s take a look at this and let’s see if we could switch over. It’s like they’re not, the culture in these companies is not like that, which is great for you as the bootstrapper who isn’t always moving as fast as a VC back company can or some big company that enters your territory. It’s like you got a lot of customers that are just there and they’re happy with your product and they’re not looking around for other solutions.
Rob Walling:
And I want to be clear, I want to get back to on-prem probably halfway through the interview, what we’re not saying is once.com, right? It’s not you pay once and that’s a different thing. I guess it’s related, but to circle back to where I wanted to start is where is the business at today?
Ian Landsman:
So we’re profitable doing well, a little under 2 million A RR five employees, I think five, yeah, five full-time employees and a couple of part-time people. So it’s kind of been running, it’s just going, it goes up a few percent every year and it’s fine and we’re profitable and it’s great, and I’ve been running it in sort of that not people say lifestyle, I wouldn’t know if I’d go all the way to lap, but that applies a certain ease of life that I don’t know if I’m all the way achieved. But yeah, not super stressed about that stuff. It’s been nice, especially the last, I think we’re going to talk about the early years. There’s probably more action in some ways, but it’s kind of the last 10 years, my kids have been getting bigger. My oldest just went to college, so it’s sort of been this, I’ve been cool with it being profitable and running well and that’s great. And yeah, I think I’m starting to turn a little corner of some new things I want to do with it as I get more time, kids get older, all that stuff. But yeah, it’s doing well.
Rob Walling:
This is going to sound like a negative pejorative question, but it’s like how have you not gotten bored? I would get bored. Where 19. It’s next year, 20 years,
And we all have different personalities and I’m the person who’s never worked the same job even, let’s see, I think MicroConf, well, MicroConf doesn’t totally count, but even for me it was a hobby for many years. But MicroConf and TinySeed now I think are the thing I’ve worked on the longest six years because even Drip from the founding to selling was three and a half years, and then I stayed another two years, so even Drip was five and a half years. So you could tell I just had that personality of like, Ooh, I need to do the next thing. But you’ve stuck with something for 20 years.
Ian Landsman:
I don’t know. I guess I wouldn’t have thought it would be exactly like that in the beginning necessarily. I think what I’ve done is since it’s been profitable, it’s given me flexibility to do other things along the way. So I ran conferences both in world and online. So similar to what you’ve done with MicroConf as a way to explore things. We built a job board for the Laravel and we run the official job board. I built a product called Thermostat that didn’t really work, and I recently sold off. I built another product that I sold off that didn’t work. So every three or four years I get into something else and that’ll distract me for a little while, probably to HubSpot’s detriment, but also probably practically just something that needed to happen to keep me when I come back to HubSpot after my little excursions, it’s like, oh, it’s still here.
It’s still doing great. Energized to take on some new stuff and do some new things. And so that’s kind of how I’ve done it. And I think it’s something I’m sure you hear a lot and Bootstrapper sort of problems let’s say, is that you get to a certain level of success and it’s just a little bit of a weird zone where it’s not big enough to sell for the amount of money that’s just like, oh, I’d never have to work again, money, I’m totally set. Don’t worry about it. It’s not quite big enough for that really. Maybe if I sold in 2021 right at the top there. So it’s never been really that. I’ve never had any really appealing offers that’s like, oh yeah, I should sell it. Obviously you could just sell it to sell it and move on to the next thing and still have a nice amount of money, but it’s like, ah, who knows if the second product’s good? I’ve never built a really good second product. I mean, some of these things have been okay, but it’s hard to build a second product, so it’s like, ah, I got the first product that’s done so well. I’m just going to stick with that. That makes
Rob Walling:
Sense. Let’s go back to the beginning. So you launched it in 2005 and you told me that you quit your day job and coded for six months, which is exactly what we tell people not to do today, right? Don’t quit your day job and live off savings or you said your wife was basically supporting you during that time, but what a gamble To me, it’s terrifying to do that, right? Were you scared at all and why do you think it actually worked? What was that time?
Ian Landsman:
Yeah, it was, man, it’s so wild to think about now. The world was just incredibly different. In 2004, it started, I wasn’t even a programmer in college or anything. I learned the program on my own and then I was like, okay, I want to do a product. We went through a million product ideas, whatever, finally came on this HubSpot idea because I used a really awful help desk at work that was mainframe based and didn’t accept email and all this stuff. That’s where the world was back then. Just to set the stage, A lot of big companies just used had no help desk or a very poor help desk solution. So it was like, okay, I want to do this. Both me and my wife agreed it’s a good idea. So we had just bought this condo, which was a little bit pricey. I was a little nerve wracking.
We sold my car, so we had that money, put it in the bank, so we just went down to the one car. My wife kept working, and then, yeah, it’s like I would never give anybody this advice now, but it’s just such a different world because it’s like there was no choice, there was no real frameworks, there’s no Ruby on Rails, no Laravel, none of these things. And so it’s like, okay, I know PHP, I’m just going to have to write it. I don’t even know JavaScript. So I’m literally sitting there with the JavaScript Bible, which is a four inch thick book, learning JavaScript while I’m building the app. And yeah, it’s like I always say I could build V one of HelpSpot today probably in three weeks or something with if I just use all the modern tools and everything we have and just do the very basic version that it was in version one. It’s like, yeah, you could do it, keep your job, do it on the side, all that stuff. But it just would literally never happen. I mean, I worked six days a week, 12 hours a day for six months, and even then it wasn’t great. It was just functional. And so that’s just what you had to do back then, or it wasn’t going to happen. Those were the options
Rob Walling:
When back when people would raise half a million dollars with just an idea that really doesn’t happen anymore. But you had to do it because as you said, you had to write every line of code and there were really no libraries, and it was, yeah, it was a lot more to be done.
Ian Landsman:
Even the raising money is something I thought about, but you got to remember in 2004, the dotcom bust just a couple of years before that, the whole VC world was kind of a mess. They weren’t looking to fund individual guys with an idea too much. I’m sure there was some investments, but it wasn’t very common. So we were all, you were there, we were all in these bootstrap circles, Joel on software forum and all kind of just working together to figure out how do we ship software without really any big funding.
Rob Walling:
That was the thing is the narrative was still raise a bunch of money, but there was Joel and there was, well, it was you
Ian Landsman:
Eric Sink and
Rob Walling:
Eric Sink and the
Ian Landsman:
Wedding Patty 11 was there, Patty
Rob Walling:
11, and, and that we all knew each other by name, and it was like, oh, yeah, and we’re all trying to figure it out on our own, but also together of like, is this even possible to do? There was no narrative, there were no books on this topic. There were no podcasts. You just, it was forums, right? I mean, that’s what I remember.
Ian Landsman:
Yeah, just forums. That’s it. There was not much else. Those of us in the community had blogs or whatever, and we talked about it, we talked about in the forums, but yeah, no Twitter, right? No YouTube, nothing. There was not a lot of nice ways to get information about how people do it and different strategies and all that stuff. Joel was writing some books. Eric Sink wrote some awesome books. That was kind of it.
Rob Walling:
And I remember around that time, I think I had gotten an invoice in oh five or oh six, and I didn’t know, I was like, well, how do you market do SEO AdWords, whatever? So I would go look at, there were no books about startups doing this. It was all the info marketer internet marketing stuff. And so I went and learned copywriting from them and applied it to startups, and I was like, oh, they split tests info. A lot of ’em were very scammy info internet marketers course. So I would take that and then translate it to me, which was the not scammy version. So they were doing split testing, they were doing AdWords, they were doing SEO O, and I learned from those folks and translated it in, and no one was really talking about that in the startup space. It was like startups were raise a bunch of money, go viral and have a big launch party, buy billboards. It was this really odd thing of, it was a weird
Ian Landsman:
Time,
Rob Walling:
It really was. But folks like you starting HubSpot, you needed nuts and bolts. How do I get in front of people? How do I sell so that I can have a paycheck next week? Let’s flash forward six months. So you spend six months building, you get to launch day. How do you get the word out?
Ian Landsman:
Man, it is crazy. You’re going to appreciate this being an email newsletter kind of guy. So I built a list, an email list of, and this was really basically just people from the Joel Software forum and a few people follow my blog, and it was literally like 84 people right now. People launch are like, oh man, I only have a thousand people on the list. What are we going to do? 84 people? And the first month we had $4,000 in sales, and it’s like some people brought it into their companies and brought it to their boss and whatever. And I had started on the SEOA little bit before launch two with the landing page and stuff and help desk software in terms of web-based help desk software was new. So we were first or second for the term help desk software for several years, just kind of like Blue Ocean or whatever, and that was our main marketing channel. But yeah, the initial launch was this 80 person email list. Then the SEO kind of picked up over those first few months, and then that was, it just started going, and we still have a lot of those customers today, but that was the kickoff.
Rob Walling:
So someone listening to this who has tried to launch and failed thinks you got $4,000 of sales in your first month, what a month. I do want to remind them it’s not 4,000 of MRR.
Ian Landsman:
No,
Rob Walling:
It is 4,000 effectively one time. Did you have a maintenance fee each year or how did that work?
Ian Landsman:
Yeah, so that’s a great point to nobody out there listings even going to think of it that way. Yes, it is not 4,000 MRR and you’re like, woo, baby. We’re launched and going, yes, it was $4,000. There was a owned license. So you buy the license and you own it, and then yearly there’s a support maintenance fee where you pay that and you get updates and you get support. And so if you don’t pay that, you don’t get updates and you don’t get support. So the good thing though is that it was effectively a recurring revenue, and this is all annual, so it was 4,000. It was like a 30 something, 30%, let’s say was the support fee. So it would be whatever it is, like 1500 bucks or whatever a year after that. So each year, that first group of $4,000 would pay us $1,500 to maintain support and updates. It’s a B2B app and it’s a heavy use B2B app where it’s not just like we use it once in a while. It’s like, no, you’re paying people full salaries to sit in this app 40 hours a week and use it. And so generally most companies are going to pay that support, and they did. And so that worked out good long term. But yeah, absolutely. It wasn’t like, oh man, this is monthly. Four grand a month was definitely not that.
Rob Walling:
Do you remember what your pricing was? I’m just trying to get an idea of how many copies sold to get four grand?
Ian Landsman:
Oh, geez. I think it was one 50 a license,
Rob Walling:
So it would’ve been like 30, well, not quite 30, like 26 or 27, somewhere in there. Licenses.
Ian Landsman:
Yeah, something like that. That’s pretty good. Sounds right.
Rob Walling:
That’s not bad at all.
Ian Landsman:
Yeah, I was happy. I was like, whoa. And then we mined that mailing list and so the second month was like a thousand dollars or something crazy, and it was a little nerve wracking there, and then it went up after that and just kept going. It’s an ASCO,
Rob Walling:
It sounds like.
Ian Landsman:
Yeah, exactly.
Rob Walling:
Because the thing that you don’t remember, if you haven’t run a onetime sale business, the sales numbers are a lot higher. T net invoice was my first product and it was 300. Well, it was a hundred when I bought it. I acquired it and then I raised it to 300 and I would sell seven to 10 copies, sometimes up to 13 copies or whatever. So you’re talking like two, three, 4,000 a month. And we did have 20% maintenance, but it really was one time. And so if we lost our SEO rankings or during the financial crisis, we would lose 80% of our revenue overnight. So we’d go from three grand down to 500 because nothing was recurring. So you basically front load it, right? It’s like they’re paying for a year is, which is actually good when you’re bootstrapping, so you get a lot more cash upfront,
Ian Landsman:
Keeps you in business.
Rob Walling:
Yeah, it’s pretty nice.
Ian Landsman:
Yeah. Well, I think that’s too, with the invoicing software, it’s like it’s a little more like, Hey, we use it, it works. Well, maybe with some of us sometime we’ll pay him his renewal, but we don’t really care that much. It’s operational. It’s like if you can get to that next layer of like, no, we have $2 million in payroll using this software, and now we’re taking this big risk. If we’re down for a day, it’s a big problem and we’re losing money and more directly. And so I think when you’re thinking about the kind of business to get into, there are advantages to that kind of thing where it’s a product that people use very heavily, very competitive nowadays too.
Rob Walling:
So you had this on-prem software from oh five until you told me 2015 is when you launched a SaaS version. And the question that popped up in my mind around that is 2015 feels late to launch. SaaS SA first became kind of in the zeitgeist in the, I would say oh seven to oh nine, that MailChimp became a thing. And there was other, I mean certainly there was SaaS before that, but it wasn’t called that. Remember ASPs and whatever the other acronyms, Basecamp, Basecamp, Salesforce, constant Contact, they were around. Even AWeber was really early, but I remember by 20 11, 20 12 SaaS, that was a thing. It was like going, so I would think you having this on-prem would have moved to SaaS, not moved, but just deployed a subscription version earlier. So what was the delay? And I’m not acting like you launched it late, but it’s just in my head. Yeah, that number. So what took that time?
Ian Landsman:
I think it was really a couple different things. Definitely when I launched, I knew about SaaS and made the decision to not make it SaaS because I didn’t know anything about running servers. There was no money to hire somebody else to run servers. And so I just feel like that wouldn’t have ultimately worked out very well. So it was on premise and then it was kind going along and it was busy and it was fine. And the big, obviously when I took more notice in terms of direct competition was you had Zendesk, which I think was maybe 2008 ish, something in there, and it’s like, oh, okay, this is really becoming a thing now, so I should probably start thinking about more. But we did even before Zendesk earlier on, have a partnership with a hosting company and they ran the hosted version of HubSpot for people.
And so it was a separate relationship with them. It’s like you came to us, you bought the licenses, you went to them, you paid the monthly hosting fee, but they would actually install help spat and run it for you, and if there was a problem with the server, they’d fix it. So we had this sort of in-between, so that did let us delay. Then another sort of aspect to it is that HubSpot was not in any way conceived as a SaaS, so the data structures are not correct for SaaS and things like that. And so in that earlier phase of the SaaS rise, I think it would’ve been pretty tricky to do it. And what kind of happened later on is we had, then you have AWS, you have really cheap servers and things like that, and so it made it simpler for us to make the move because how HubSpot SaaS version runs is actually everybody gets a tiny AWS like nano server and everybody’s on their own server.
We do have a centralized big RDS database with multi failover and all that stuff, but we basically deploy it as your own little HubSpot server, and so we don’t have to make the database work for multi-tenant and all these things, and the cost is outrageous. By that point, it was like, okay, two bucks a month you can have a little mini server on AWS. And so that’s the way we rolled it out, and this way we can keep the code base the same for on-premise and cloud and as we go, I think it’s going to start to, we do already make a lot more accommodations for the cloud version since it’s more of the new customers, but that was kind of the transformation there basically.
Rob Walling:
How painful was that
Ian Landsman:
To go on
Rob Walling:
Premises? Oh, it’s painful. Was it brutal?
Ian Landsman:
Brutal. Very painful. I mean, because there’s a lot involved had to switch. We moved to subscription pricing at that time, but I didn’t want to force everybody’s subscription pricing. And so it’s like if you want to be on the cloud version, you have to move to subscription. So even on-premise customers, now if you’re a new on-premise customer, you’re on subscription. Everybody’s on subscription and it is annual subscription, which is another thing I think founders don’t think enough about is just only having, I’ve only ever had annual, and I think it’s served me pretty well. We are going to experiment with the monthly option here, but again, if you’re in the kind of business where it’s often, it’s literally a committee we’re dealing with and they’re evaluating multiple choices and they’re not really going to switch off on a whim after three months, they’re making a commitment on their end, so not that big a deal.
So that was kind of a big one. The subscriptions, the tech. Yeah, it was, oh, the other thing is moving people. So we have all these existing customers and they want to move, and it’s a big app. We have customers at 70 gigabyte databases, and so moving them onto the cloud is quite a production. We also allow a lot of customization. It’s kind of one HubSpot’s differentiators from other solutions. There’s a lot of knobs and dials and things you can do. You can even write your own PHP in certain spots if you want to do something very custom, which is something that other help desk software won’t let you customize on that deeper level, obviously, because no way for them to really safely do that. So we still to today, so we did that 10 years ago. Today we still will have at least usually one a month, but sometimes more of customers coming from on-premise to our cloud solution. And yeah, it’s a multi-week operation usually to not pure time, not like 80 hours, but going back and forth and figuring it all out and they check it and blah, blah. So yeah, it’s a big process.
Rob Walling:
I was going to ask if you still offer on-Prem or if at a certain point, obviously if you sell, have a bunch of customers using it, paying you, you would let them keep going, but the idea of just going to hubspot.com and the only thing you can sign up for is the SaaS. You haven’t done that. You still offer both. What’s the thinking there?
Ian Landsman:
And we have had competitors. I think Kyoko was also an on-premise like us, and then they moved to SaaS and then they said, that’s it. No more on-premise. Spiceworks also said no more. So people definitely abandon the on-premise. But yeah, for us, I just feel like we definitely obviously have a good amount of customers who still use it. Like I said earlier, those are some of our biggest, best customers. Been with us 14 years and pay us $20,000 a year, big accounts for us. And so definitely eliminating that would be tricky if I’m forcing them to come up to my cloud or go elsewhere, some percentage are obviously going to go elsewhere just naturally, so I don’t want to do that. I could, of course, like you said, just going forward, but again, it’s like a lot of really great customers are in the on-premise.
So even though it’s only probably 10% of our new business, it’s a valuable 10%. And also I think we haven’t done a great job marketing it recently, and that’s something I want to actually do better and try to get that actually up a little bit. There are interesting trade-offs with on-premise in that it’s more support for sure. There’s just server issues and sometimes the support person takes half a day on somebody else’s problem ultimately that we end up being responsible for helping them, things like that. But again, they’re oftentimes much more higher revenue than the day in day out customer. And they also, there’s other little interesting things like they often pay by check or transfer, which costs either nothing or $10 versus cloud customers that often pay by credit card and costs 5% ultimately and stuff, which when on $10,000, 5% is a lot of money.
So there are these little other things too. But yeah, that’s kind of the reasons I still think that there’s something there. And also I think in general, we’re seeing Elasticsearch and databases and a lot of the structural things still are even more so offering on-Prem. It might be literally on-Prem, it might be in your AWS cloud, but it’s your cloud, not our cloud. And so I think there’s just still opportunity there that I wouldn’t want to be totally out of the game, especially because no other help desk software, nobody knew when Help Desk software is offering it. None of the existing players are offering it who didn’t already offer it in the past. And so I feel like it is a little bit of a competitive advantage there where it’s like we’re going up against other poorer solutions than just the intercoms of the world with infinite money. It’s like, no, this can be a little niche that we have some advantages in.
Rob Walling:
No, I think that makes sense. I may seem off topic, but I have a question for you about, well, you were written on some PHP that you hacked together 20 years ago, but then I know at a certain point you integrated and you’re pretty heavily involved in the community. I think you mentioned some of that earlier. Laravel just raised 57 million, I don’t know, two weeks ago maybe. What’s your thinking on that and what that means for the future of Laravel?
Ian Landsman:
Yeah, so I think it’s really exciting. I mean, I run the official Laravel job board. Taylor Twell, the creator of Laravel, worked at UserScape with me for three years. My company’s name is,
Rob Walling:
I didn’t know that. Oh, what a trip.
Ian Landsman:
Yeah.
Rob Walling:
Did he work on HelpSpot?
Ian Landsman:
Yeah, yeah. And some other new stuff. In one of my distraction periods, he worked on some new stuff. He worked on HubSpot. But yeah, I actually found him, he was working at a shipping company and I found Laravel and I was like, wow, this is awesome. I want to have a framework to work on new versions of HubSpot with as well as other ideas. And I didn’t like any of the other PHP frameworks even up to then, which is 2010 or something like that. So I reached out to him. I was looking to hire a developer. Anyway, I ended up hiring him. He worked the first three or four months at user scape, just building out Laravel actually in some of the more enterprisey things. It didn’t have already caching and some other things. And then, yeah, we did the conferences together and stuff. Then ultimately he built Forge, which is the main revenue producer at Laravel.
It’s like a hosting platform or manager. Then that started making a lot of money, and I was like, okay, you should go run that. And so he did. But we’ve stayed close friends and I’m definitely friends with a lot of people in the Laravel community. So yeah, I think the thing is he turned that into a bootstrapped and very profitable business, and again, kind hit that point where he had some bigger ambitions basically to launch this thing called Laravel Cloud, which is going to be Heroku, but for PHP and Laravel, basically, it’s kind of like the shorthand, but that kind of thing obviously requires just even if you have a very profitable business, it’s ultimately you’re going to need a lot of very expensive engineers and lots of servers and security people and all this stuff that starts to get into, if I’m paying, I’m making up numbers here, have no idea what they’re paying. But if you’re paying a security guy $800,000 a year, let’s say, or whatever, that starts to be a lot of money. So to chase those ambitions, that’s what he’s done. So I think so far so good. Like you said, it’s only been a couple months since they raised, so
Rob Walling:
Now I want to ask you something that I asked you offline, and it was about what’s your biggest fear today? Looking ahead, you’ve run this business for 20 years, and you mentioned it was like how can AI potentially impact this? You want to talk us through your thinking?
Ian Landsman:
Yeah, so I think this is a topical one. So I think customer service is more directly in the crosshairs of even current AI capabilities to some degree. Obviously you can make the case for like, well, every job is going to be affected or whatever, who knows? But I think this is one where it’s like it’s a cost center. A lot of companies look at As, and they’re like, okay, if AI can do a reasonable job, maybe we’re willing to make changes there. So yeah, I just think it’s a big unknown. I think so far it hasn’t been really disruptive at all, but that doesn’t mean it won’t. There’s definitely been some companies that have made big moves, like Klarna got a lot of press for moving almost all their support to ai, but again, that is a very particular use case where they don’t have a lot of different types of questions.
Basically it’s like, I want a refund, how do I pay my bill? So I think a very good use case for an ai, they had poor service ratings to begin with, so it’s like, are the AI people going to do any worse? Probably not. So you have some differences there, but yeah, who knows? So I mean on the one hand it’s like those fears of just might consolidate down to a few big AI players who do a fantastic job with it and they just gobble everything up, or is there going to be, is it not going to be a thing at all or more of a middle ground where since we have access to a lot of similar technology, at least ostensibly, right, is we can pay for open AI or whatever else, we can add those tools for the most part, I think there’s going to be probably some things that are going to be beyond our capability, but so I think that’s definitely just a big unknown and we haven’t had that in a long time. And the help desk space, it’s been pretty stable. So this is definitely a new thing that’s out there and it’s like, we’ll see how it goes. We’re definitely adding AI features and maybe that’s just kind of be kind of where it ends up, but we’ll see.
Rob Walling:
That makes sense. Just out of curiosity, what’s the first one or two AI features you’re adding?
Ian Landsman:
Yeah, so very early on we did your standard writing helpers. We went farther than most of the other help desk tools I’ve seen have gone where you can define your own prompts and tools for the agents, but still it’s ultimately more on the writing and human creation side. But from there we are working on doing things like auto triaging where it can route inbound tickets to the right categories or right agents and things like that. So that’s going to be one of the first more sort of offensive minded, more active AI elements that we’re going to be adding. And then from there we have some different ideas about how we might do auto responding in a way that’s maybe a little bit safer. So obviously everybody’s trying to go for the holy grail of the AI just responds. You give it anything and it can just respond, but maybe there’s some in-between ground there where it’s more of a human response, but we use the AI to maybe figure out what’s the right response or some different things like that. So some stuff in r and d, some stuff very close to shipping. So that’s kind of where we’re at with it. But it is a big sort of sea change much when I started the company, which is like everybody’s using client server apps or just pure email, and that was the big shift and now it seems like there might be another big shift. I don’t know if it’ll be, it could be bigger than that. It could be less than that. I dunno, who knows, but we’ll see.
Rob Walling:
So last question for you today before I let you go, I asked offline, I said you named HelpSpot. How did you feel when all these other help desks launched with help? Right, so there was Help Scout and help somewhere else, and what did you say to me?
Ian Landsman:
Yeah, so those don’t bother me at all. There’s lots of whatever variations, but I thought the spot was kind of my unique thing. I liked the spot and Dharmesh Shaw, who’s the co-founder of HubSpot, was in Joel on software with us and kind of in our circles. And then he starts HubSpot and stole my spot like dude, and now it’s HubSpot. And obviously anytime I say HubSpot and HubSpot and people get confused even more than with Help Scout, although that one does get confused as well sometimes. But yeah, I’d say in terms of naming irks HubSpot,
Rob Walling:
HubSpot
Ian Landsman:
Kind of takes the pride from me.
Rob Walling:
Sometimes that happens, sometimes it happens. Seemed to have
Ian Landsman:
Done it. I like Dharmesh though. DH is great. Alright. He’s a great
Rob Walling:
Guy. He’s so gracious and he still, when I wrote this SA playbook, I emailed and said, would you give me an endorsement? And he’s like, of course. And he gave this great endorsement. I put it on the cover of the book. It was so nice. It was like, I’ve been learning from Rob for years. You should too or something. I’m like, oh, you’re the
Ian Landsman:
Awesome man. That’s crazy. He’s so good. Yeah, a few times I’ve talked to him over the years too. It’s just like, yeah, he’s still like the guy from in the for, you know what I mean? It’s like he’s a billionaire now, but he’s still kind of just humble. The same guy.
Rob Walling:
Humble and just,
Ian Landsman:
Yeah. That’s great.
Rob Walling:
Ian Lansman, thanks for joining me today. Folks want to keep up with what you’re working on, obviously helpspot.com to see the app, and you are Ian Lansman on X Twitter. Anything else you’d like folks to check out? I know you have a few podcasts.
Ian Landsman:
Yeah, the other thing, I guess the other main thing, if you want to keep up with what I’m doing is mostly technical is my podcast I do with Aaron Francis, and it’s more somewhat technical. There’s a little bit technical, but mostly it’s business and other topics, so it’s more entertaining, I’d say. Yeah, so check that out. We have a lot of people seem to like that, so that’s been fun the last year or so, working on that. Again, one of my little, I can’t really justify it on an ROI basis purely, but it’s a fun thing that I get to work on and is exciting to do.
Rob Walling:
Awesome.
Ian Landsman:
Yeah, so check me out
Rob Walling:
There. Thanks again for coming on the show. Thanks for having me. Thanks again to Ian for coming on the show, and thank you for listening this week and every week. This is Rob Walling signing off from episode 734.
Welcome to this episode of Startups of the Rest of Us. I’m Rob Walling and this week I talk with Ian Lansman, the founder of HelpSpot. Ian has been bootstrapping for almost 20 years and he started with on-Prem software and then after about 10 years launched a SaaS version. Ian and I have known each other almost that same 20 years from back in the days of Jolen software’s business of software forums. And over the years, Ian has started a few of his own podcasts as well as spoken at MicroConf. I believe I were to guess it was probably around 2014 or 15. So it’s a great conversation because Ian and I have history. I think we were able to pull out some really interesting bits about why he’s bootstrapped for 20 years, why he hasn’t sold and moved on to his next act, how he thinks AI is going to impact the customer service and support space. And if you stick around until the end, you’ll get to hear the company name that is most often confused with HelpSpot and it wasn’t the one that I thought. So with that, let’s dive into our conversation. Lansman, welcome to the show. Thanks for having me on. It’s great to be here. Cannot believe you have not been on this show. No, you’re like OG Bootstrapper.
Ian Landsman:
Somebody on Twitter reached out and said, I should come on here. And I was like, man, I think I’ve been on there. And then I searched and I was like, man, I haven’t been on there. Wow, that’s crazy.
Rob Walling:
Yeah, I am glad you reached out. I think someone, your podcast co-host maybe mentioned it on a podcast and I use arvid calls pod scan, and so my name popped up and I went and I’m like, wait, what? Someone wants to go out and start. So I was like, yeah, dude. And I thought the same thing. I was like, I think Ian’s been on it, but to give people context, like 20 year, I mean really early man back in the day, remember how it was, there’s going to be a lot of jokes about how back in the day, there was no customer development, there was no AWS, there was no SaaS right? Where
Ian Landsman:
Before all that, man, it’s hard to believe we’ve become the old timers. How did that happen?
Rob Walling:
Oh yeah, big time. Just so people know, you’re the founder of HelpSpot. It’s at helpspot.com, it your H one is your customer service at scale Amplify your support with HelpSpot, the streamlined solution for scaling customer service effectively. So it’s email ticketing, reporting and metrics knowledge base. And since you started so long ago, you were on-Prem for years and then you launched a SaaS. So you have both on-Prem and SaaS, and I think most folks listening know what on-Prem is, but it’s where they actually download your code, your source code, PHP, Laravel, and they will install it on their own server. And then I guess what’s the benefit? Why would today, I would just do the SaaS personally, but the people who still use on-Prem today, why do they do that rather than pay you for the
Ian Landsman:
SaaS? Yeah, so definitely the majority or SaaS using the SaaS version these days, but we still have new customers in the on-prem and basically there’s some advantages. Big companies or certain specialized companies like in finance, healthcare, they have different rules, potentially laws. They’re also the type of companies that still have IT departments. And so they want to be in control of their own data is usually the main reason in terms of they control the database and they’re backing it up and those things, but also sometimes they’re just running completely off the internet. It’s like, here’s our help desk, it’s for maybe IT in that case and it’s literally not connected to the internet. And so it’s just fully encapsulated in their network behind their firewall. And so they want to do that. And it is an area I think people have abandoned completely that maybe shouldn’t be totally abandoned because there are some, you could charge more. We actually charge the same price currently, but I think there’s a lot of opportunities there to charge a lot more for the people who want it on premise and things like that. So yeah, it’s not that bad if you have a modern SaaS app, it’s not impossible to make an on-premise version. Some things like if you depend on 30 services, that’s going to be hard, but with a little bit of forethought, it’s really not that bad.
Rob Walling:
We have a few TinySeed companies and I don’t know how many exactly, but when they apply they’ll often say, so we’re SaaS, but we also do on-prem. Is that going to be a deal breaker? And I was like, no, we funded a handful. And again, I don’t know if it’s five or if it’s 10 out of one 70, but there are especially like you said in finance or in certain regulated industries where you really do need or want control of that data and just went on some random spot.
Ian Landsman:
Yeah, it also gives you some advantages of you’re a tiny bootstrap company, you’re not going to be able to be SOC two compliant, for example, and there’s going to be these type of companies that are like, well, you’re SOC two. Let me talk to your security people, all this stuff. And it’s just like you, it’s you and a couple of people, and so this gives you some outs with those big companies. It’s like, Hey, you know what? You can install in your own SOC two data center and you’re already SOC two and you can, it’s all good and so we’ll charge you for that and everything. And those tend to be also some of the best customers. I mean those are the customers that stay with you for a decade. You’re in there, you’re on a server, just not even the type of companies that are in the mindset of like, oh, somebody launched a new fancier version of this tool we use. Let’s go out and find the fancy. Let’s take a look at this and let’s see if we could switch over. It’s like they’re not, the culture in these companies is not like that, which is great for you as the bootstrapper who isn’t always moving as fast as a VC back company can or some big company that enters your territory. It’s like you got a lot of customers that are just there and they’re happy with your product and they’re not looking around for other solutions.
Rob Walling:
And I want to be clear, I want to get back to on-prem probably halfway through the interview, what we’re not saying is once.com, right? It’s not you pay once and that’s a different thing. I guess it’s related, but to circle back to where I wanted to start is where is the business at today?
Ian Landsman:
So we’re profitable doing well, a little under 2 million A RR five employees, I think five, yeah, five full-time employees and a couple of part-time people. So it’s kind of been running, it’s just going, it goes up a few percent every year and it’s fine and we’re profitable and it’s great, and I’ve been running it in sort of that not people say lifestyle, I wouldn’t know if I’d go all the way to lap, but that applies a certain ease of life that I don’t know if I’m all the way achieved. But yeah, not super stressed about that stuff. It’s been nice, especially the last, I think we’re going to talk about the early years. There’s probably more action in some ways, but it’s kind of the last 10 years, my kids have been getting bigger. My oldest just went to college, so it’s sort of been this, I’ve been cool with it being profitable and running well and that’s great. And yeah, I think I’m starting to turn a little corner of some new things I want to do with it as I get more time, kids get older, all that stuff. But yeah, it’s doing well.
Rob Walling:
This is going to sound like a negative pejorative question, but it’s like how have you not gotten bored? I would get bored. Where 19. It’s next year, 20 years,
And we all have different personalities and I’m the person who’s never worked the same job even, let’s see, I think MicroConf, well, MicroConf doesn’t totally count, but even for me it was a hobby for many years. But MicroConf and TinySeed now I think are the thing I’ve worked on the longest six years because even Drip from the founding to selling was three and a half years, and then I stayed another two years, so even Drip was five and a half years. So you could tell I just had that personality of like, Ooh, I need to do the next thing. But you’ve stuck with something for 20 years.
Ian Landsman:
I don’t know. I guess I wouldn’t have thought it would be exactly like that in the beginning necessarily. I think what I’ve done is since it’s been profitable, it’s given me flexibility to do other things along the way. So I ran conferences both in world and online. So similar to what you’ve done with MicroConf as a way to explore things. We built a job board for the Laravel and we run the official job board. I built a product called Thermostat that didn’t really work, and I recently sold off. I built another product that I sold off that didn’t work. So every three or four years I get into something else and that’ll distract me for a little while, probably to HubSpot’s detriment, but also probably practically just something that needed to happen to keep me when I come back to HubSpot after my little excursions, it’s like, oh, it’s still here.
It’s still doing great. Energized to take on some new stuff and do some new things. And so that’s kind of how I’ve done it. And I think it’s something I’m sure you hear a lot and Bootstrapper sort of problems let’s say, is that you get to a certain level of success and it’s just a little bit of a weird zone where it’s not big enough to sell for the amount of money that’s just like, oh, I’d never have to work again, money, I’m totally set. Don’t worry about it. It’s not quite big enough for that really. Maybe if I sold in 2021 right at the top there. So it’s never been really that. I’ve never had any really appealing offers that’s like, oh yeah, I should sell it. Obviously you could just sell it to sell it and move on to the next thing and still have a nice amount of money, but it’s like, ah, who knows if the second product’s good? I’ve never built a really good second product. I mean, some of these things have been okay, but it’s hard to build a second product, so it’s like, ah, I got the first product that’s done so well. I’m just going to stick with that. That makes
Rob Walling:
Sense. Let’s go back to the beginning. So you launched it in 2005 and you told me that you quit your day job and coded for six months, which is exactly what we tell people not to do today, right? Don’t quit your day job and live off savings or you said your wife was basically supporting you during that time, but what a gamble To me, it’s terrifying to do that, right? Were you scared at all and why do you think it actually worked? What was that time?
Ian Landsman:
Yeah, it was, man, it’s so wild to think about now. The world was just incredibly different. In 2004, it started, I wasn’t even a programmer in college or anything. I learned the program on my own and then I was like, okay, I want to do a product. We went through a million product ideas, whatever, finally came on this HubSpot idea because I used a really awful help desk at work that was mainframe based and didn’t accept email and all this stuff. That’s where the world was back then. Just to set the stage, A lot of big companies just used had no help desk or a very poor help desk solution. So it was like, okay, I want to do this. Both me and my wife agreed it’s a good idea. So we had just bought this condo, which was a little bit pricey. I was a little nerve wracking.
We sold my car, so we had that money, put it in the bank, so we just went down to the one car. My wife kept working, and then, yeah, it’s like I would never give anybody this advice now, but it’s just such a different world because it’s like there was no choice, there was no real frameworks, there’s no Ruby on Rails, no Laravel, none of these things. And so it’s like, okay, I know PHP, I’m just going to have to write it. I don’t even know JavaScript. So I’m literally sitting there with the JavaScript Bible, which is a four inch thick book, learning JavaScript while I’m building the app. And yeah, it’s like I always say I could build V one of HelpSpot today probably in three weeks or something with if I just use all the modern tools and everything we have and just do the very basic version that it was in version one. It’s like, yeah, you could do it, keep your job, do it on the side, all that stuff. But it just would literally never happen. I mean, I worked six days a week, 12 hours a day for six months, and even then it wasn’t great. It was just functional. And so that’s just what you had to do back then, or it wasn’t going to happen. Those were the options
Rob Walling:
When back when people would raise half a million dollars with just an idea that really doesn’t happen anymore. But you had to do it because as you said, you had to write every line of code and there were really no libraries, and it was, yeah, it was a lot more to be done.
Ian Landsman:
Even the raising money is something I thought about, but you got to remember in 2004, the dotcom bust just a couple of years before that, the whole VC world was kind of a mess. They weren’t looking to fund individual guys with an idea too much. I’m sure there was some investments, but it wasn’t very common. So we were all, you were there, we were all in these bootstrap circles, Joel on software forum and all kind of just working together to figure out how do we ship software without really any big funding.
Rob Walling:
That was the thing is the narrative was still raise a bunch of money, but there was Joel and there was, well, it was you
Ian Landsman:
Eric Sink and
Rob Walling:
Eric Sink and the
Ian Landsman:
Wedding Patty 11 was there, Patty
Rob Walling:
11, and, and that we all knew each other by name, and it was like, oh, yeah, and we’re all trying to figure it out on our own, but also together of like, is this even possible to do? There was no narrative, there were no books on this topic. There were no podcasts. You just, it was forums, right? I mean, that’s what I remember.
Ian Landsman:
Yeah, just forums. That’s it. There was not much else. Those of us in the community had blogs or whatever, and we talked about it, we talked about in the forums, but yeah, no Twitter, right? No YouTube, nothing. There was not a lot of nice ways to get information about how people do it and different strategies and all that stuff. Joel was writing some books. Eric Sink wrote some awesome books. That was kind of it.
Rob Walling:
And I remember around that time, I think I had gotten an invoice in oh five or oh six, and I didn’t know, I was like, well, how do you market do SEO AdWords, whatever? So I would go look at, there were no books about startups doing this. It was all the info marketer internet marketing stuff. And so I went and learned copywriting from them and applied it to startups, and I was like, oh, they split tests info. A lot of ’em were very scammy info internet marketers course. So I would take that and then translate it to me, which was the not scammy version. So they were doing split testing, they were doing AdWords, they were doing SEO O, and I learned from those folks and translated it in, and no one was really talking about that in the startup space. It was like startups were raise a bunch of money, go viral and have a big launch party, buy billboards. It was this really odd thing of, it was a weird
Ian Landsman:
Time,
Rob Walling:
It really was. But folks like you starting HubSpot, you needed nuts and bolts. How do I get in front of people? How do I sell so that I can have a paycheck next week? Let’s flash forward six months. So you spend six months building, you get to launch day. How do you get the word out?
Ian Landsman:
Man, it is crazy. You’re going to appreciate this being an email newsletter kind of guy. So I built a list, an email list of, and this was really basically just people from the Joel Software forum and a few people follow my blog, and it was literally like 84 people right now. People launch are like, oh man, I only have a thousand people on the list. What are we going to do? 84 people? And the first month we had $4,000 in sales, and it’s like some people brought it into their companies and brought it to their boss and whatever. And I had started on the SEOA little bit before launch two with the landing page and stuff and help desk software in terms of web-based help desk software was new. So we were first or second for the term help desk software for several years, just kind of like Blue Ocean or whatever, and that was our main marketing channel. But yeah, the initial launch was this 80 person email list. Then the SEO kind of picked up over those first few months, and then that was, it just started going, and we still have a lot of those customers today, but that was the kickoff.
Rob Walling:
So someone listening to this who has tried to launch and failed thinks you got $4,000 of sales in your first month, what a month. I do want to remind them it’s not 4,000 of MRR.
Ian Landsman:
No,
Rob Walling:
It is 4,000 effectively one time. Did you have a maintenance fee each year or how did that work?
Ian Landsman:
Yeah, so that’s a great point to nobody out there listings even going to think of it that way. Yes, it is not 4,000 MRR and you’re like, woo, baby. We’re launched and going, yes, it was $4,000. There was a owned license. So you buy the license and you own it, and then yearly there’s a support maintenance fee where you pay that and you get updates and you get support. And so if you don’t pay that, you don’t get updates and you don’t get support. So the good thing though is that it was effectively a recurring revenue, and this is all annual, so it was 4,000. It was like a 30 something, 30%, let’s say was the support fee. So it would be whatever it is, like 1500 bucks or whatever a year after that. So each year, that first group of $4,000 would pay us $1,500 to maintain support and updates. It’s a B2B app and it’s a heavy use B2B app where it’s not just like we use it once in a while. It’s like, no, you’re paying people full salaries to sit in this app 40 hours a week and use it. And so generally most companies are going to pay that support, and they did. And so that worked out good long term. But yeah, absolutely. It wasn’t like, oh man, this is monthly. Four grand a month was definitely not that.
Rob Walling:
Do you remember what your pricing was? I’m just trying to get an idea of how many copies sold to get four grand?
Ian Landsman:
Oh, geez. I think it was one 50 a license,
Rob Walling:
So it would’ve been like 30, well, not quite 30, like 26 or 27, somewhere in there. Licenses.
Ian Landsman:
Yeah, something like that. That’s pretty good. Sounds right.
Rob Walling:
That’s not bad at all.
Ian Landsman:
Yeah, I was happy. I was like, whoa. And then we mined that mailing list and so the second month was like a thousand dollars or something crazy, and it was a little nerve wracking there, and then it went up after that and just kept going. It’s an ASCO,
Rob Walling:
It sounds like.
Ian Landsman:
Yeah, exactly.
Rob Walling:
Because the thing that you don’t remember, if you haven’t run a onetime sale business, the sales numbers are a lot higher. T net invoice was my first product and it was 300. Well, it was a hundred when I bought it. I acquired it and then I raised it to 300 and I would sell seven to 10 copies, sometimes up to 13 copies or whatever. So you’re talking like two, three, 4,000 a month. And we did have 20% maintenance, but it really was one time. And so if we lost our SEO rankings or during the financial crisis, we would lose 80% of our revenue overnight. So we’d go from three grand down to 500 because nothing was recurring. So you basically front load it, right? It’s like they’re paying for a year is, which is actually good when you’re bootstrapping, so you get a lot more cash upfront,
Ian Landsman:
Keeps you in business.
Rob Walling:
Yeah, it’s pretty nice.
Ian Landsman:
Yeah. Well, I think that’s too, with the invoicing software, it’s like it’s a little more like, Hey, we use it, it works. Well, maybe with some of us sometime we’ll pay him his renewal, but we don’t really care that much. It’s operational. It’s like if you can get to that next layer of like, no, we have $2 million in payroll using this software, and now we’re taking this big risk. If we’re down for a day, it’s a big problem and we’re losing money and more directly. And so I think when you’re thinking about the kind of business to get into, there are advantages to that kind of thing where it’s a product that people use very heavily, very competitive nowadays too.
Rob Walling:
So you had this on-prem software from oh five until you told me 2015 is when you launched a SaaS version. And the question that popped up in my mind around that is 2015 feels late to launch. SaaS SA first became kind of in the zeitgeist in the, I would say oh seven to oh nine, that MailChimp became a thing. And there was other, I mean certainly there was SaaS before that, but it wasn’t called that. Remember ASPs and whatever the other acronyms, Basecamp, Basecamp, Salesforce, constant Contact, they were around. Even AWeber was really early, but I remember by 20 11, 20 12 SaaS, that was a thing. It was like going, so I would think you having this on-prem would have moved to SaaS, not moved, but just deployed a subscription version earlier. So what was the delay? And I’m not acting like you launched it late, but it’s just in my head. Yeah, that number. So what took that time?
Ian Landsman:
I think it was really a couple different things. Definitely when I launched, I knew about SaaS and made the decision to not make it SaaS because I didn’t know anything about running servers. There was no money to hire somebody else to run servers. And so I just feel like that wouldn’t have ultimately worked out very well. So it was on premise and then it was kind going along and it was busy and it was fine. And the big, obviously when I took more notice in terms of direct competition was you had Zendesk, which I think was maybe 2008 ish, something in there, and it’s like, oh, okay, this is really becoming a thing now, so I should probably start thinking about more. But we did even before Zendesk earlier on, have a partnership with a hosting company and they ran the hosted version of HubSpot for people.
And so it was a separate relationship with them. It’s like you came to us, you bought the licenses, you went to them, you paid the monthly hosting fee, but they would actually install help spat and run it for you, and if there was a problem with the server, they’d fix it. So we had this sort of in-between, so that did let us delay. Then another sort of aspect to it is that HubSpot was not in any way conceived as a SaaS, so the data structures are not correct for SaaS and things like that. And so in that earlier phase of the SaaS rise, I think it would’ve been pretty tricky to do it. And what kind of happened later on is we had, then you have AWS, you have really cheap servers and things like that, and so it made it simpler for us to make the move because how HubSpot SaaS version runs is actually everybody gets a tiny AWS like nano server and everybody’s on their own server.
We do have a centralized big RDS database with multi failover and all that stuff, but we basically deploy it as your own little HubSpot server, and so we don’t have to make the database work for multi-tenant and all these things, and the cost is outrageous. By that point, it was like, okay, two bucks a month you can have a little mini server on AWS. And so that’s the way we rolled it out, and this way we can keep the code base the same for on-premise and cloud and as we go, I think it’s going to start to, we do already make a lot more accommodations for the cloud version since it’s more of the new customers, but that was kind of the transformation there basically.
Rob Walling:
How painful was that
Ian Landsman:
To go on
Rob Walling:
Premises? Oh, it’s painful. Was it brutal?
Ian Landsman:
Brutal. Very painful. I mean, because there’s a lot involved had to switch. We moved to subscription pricing at that time, but I didn’t want to force everybody’s subscription pricing. And so it’s like if you want to be on the cloud version, you have to move to subscription. So even on-premise customers, now if you’re a new on-premise customer, you’re on subscription. Everybody’s on subscription and it is annual subscription, which is another thing I think founders don’t think enough about is just only having, I’ve only ever had annual, and I think it’s served me pretty well. We are going to experiment with the monthly option here, but again, if you’re in the kind of business where it’s often, it’s literally a committee we’re dealing with and they’re evaluating multiple choices and they’re not really going to switch off on a whim after three months, they’re making a commitment on their end, so not that big a deal.
So that was kind of a big one. The subscriptions, the tech. Yeah, it was, oh, the other thing is moving people. So we have all these existing customers and they want to move, and it’s a big app. We have customers at 70 gigabyte databases, and so moving them onto the cloud is quite a production. We also allow a lot of customization. It’s kind of one HubSpot’s differentiators from other solutions. There’s a lot of knobs and dials and things you can do. You can even write your own PHP in certain spots if you want to do something very custom, which is something that other help desk software won’t let you customize on that deeper level, obviously, because no way for them to really safely do that. So we still to today, so we did that 10 years ago. Today we still will have at least usually one a month, but sometimes more of customers coming from on-premise to our cloud solution. And yeah, it’s a multi-week operation usually to not pure time, not like 80 hours, but going back and forth and figuring it all out and they check it and blah, blah. So yeah, it’s a big process.
Rob Walling:
I was going to ask if you still offer on-Prem or if at a certain point, obviously if you sell, have a bunch of customers using it, paying you, you would let them keep going, but the idea of just going to hubspot.com and the only thing you can sign up for is the SaaS. You haven’t done that. You still offer both. What’s the thinking there?
Ian Landsman:
And we have had competitors. I think Kyoko was also an on-premise like us, and then they moved to SaaS and then they said, that’s it. No more on-premise. Spiceworks also said no more. So people definitely abandon the on-premise. But yeah, for us, I just feel like we definitely obviously have a good amount of customers who still use it. Like I said earlier, those are some of our biggest, best customers. Been with us 14 years and pay us $20,000 a year, big accounts for us. And so definitely eliminating that would be tricky if I’m forcing them to come up to my cloud or go elsewhere, some percentage are obviously going to go elsewhere just naturally, so I don’t want to do that. I could, of course, like you said, just going forward, but again, it’s like a lot of really great customers are in the on-premise.
So even though it’s only probably 10% of our new business, it’s a valuable 10%. And also I think we haven’t done a great job marketing it recently, and that’s something I want to actually do better and try to get that actually up a little bit. There are interesting trade-offs with on-premise in that it’s more support for sure. There’s just server issues and sometimes the support person takes half a day on somebody else’s problem ultimately that we end up being responsible for helping them, things like that. But again, they’re oftentimes much more higher revenue than the day in day out customer. And they also, there’s other little interesting things like they often pay by check or transfer, which costs either nothing or $10 versus cloud customers that often pay by credit card and costs 5% ultimately and stuff, which when on $10,000, 5% is a lot of money.
So there are these little other things too. But yeah, that’s kind of the reasons I still think that there’s something there. And also I think in general, we’re seeing Elasticsearch and databases and a lot of the structural things still are even more so offering on-Prem. It might be literally on-Prem, it might be in your AWS cloud, but it’s your cloud, not our cloud. And so I think there’s just still opportunity there that I wouldn’t want to be totally out of the game, especially because no other help desk software, nobody knew when Help Desk software is offering it. None of the existing players are offering it who didn’t already offer it in the past. And so I feel like it is a little bit of a competitive advantage there where it’s like we’re going up against other poorer solutions than just the intercoms of the world with infinite money. It’s like, no, this can be a little niche that we have some advantages in.
Rob Walling:
No, I think that makes sense. I may seem off topic, but I have a question for you about, well, you were written on some PHP that you hacked together 20 years ago, but then I know at a certain point you integrated and you’re pretty heavily involved in the community. I think you mentioned some of that earlier. Laravel just raised 57 million, I don’t know, two weeks ago maybe. What’s your thinking on that and what that means for the future of Laravel?
Ian Landsman:
Yeah, so I think it’s really exciting. I mean, I run the official Laravel job board. Taylor Twell, the creator of Laravel, worked at UserScape with me for three years. My company’s name is,
Rob Walling:
I didn’t know that. Oh, what a trip.
Ian Landsman:
Yeah.
Rob Walling:
Did he work on HelpSpot?
Ian Landsman:
Yeah, yeah. And some other new stuff. In one of my distraction periods, he worked on some new stuff. He worked on HubSpot. But yeah, I actually found him, he was working at a shipping company and I found Laravel and I was like, wow, this is awesome. I want to have a framework to work on new versions of HubSpot with as well as other ideas. And I didn’t like any of the other PHP frameworks even up to then, which is 2010 or something like that. So I reached out to him. I was looking to hire a developer. Anyway, I ended up hiring him. He worked the first three or four months at user scape, just building out Laravel actually in some of the more enterprisey things. It didn’t have already caching and some other things. And then, yeah, we did the conferences together and stuff. Then ultimately he built Forge, which is the main revenue producer at Laravel.
It’s like a hosting platform or manager. Then that started making a lot of money, and I was like, okay, you should go run that. And so he did. But we’ve stayed close friends and I’m definitely friends with a lot of people in the Laravel community. So yeah, I think the thing is he turned that into a bootstrapped and very profitable business, and again, kind hit that point where he had some bigger ambitions basically to launch this thing called Laravel Cloud, which is going to be Heroku, but for PHP and Laravel, basically, it’s kind of like the shorthand, but that kind of thing obviously requires just even if you have a very profitable business, it’s ultimately you’re going to need a lot of very expensive engineers and lots of servers and security people and all this stuff that starts to get into, if I’m paying, I’m making up numbers here, have no idea what they’re paying. But if you’re paying a security guy $800,000 a year, let’s say, or whatever, that starts to be a lot of money. So to chase those ambitions, that’s what he’s done. So I think so far so good. Like you said, it’s only been a couple months since they raised, so
Rob Walling:
Now I want to ask you something that I asked you offline, and it was about what’s your biggest fear today? Looking ahead, you’ve run this business for 20 years, and you mentioned it was like how can AI potentially impact this? You want to talk us through your thinking?
Ian Landsman:
Yeah, so I think this is a topical one. So I think customer service is more directly in the crosshairs of even current AI capabilities to some degree. Obviously you can make the case for like, well, every job is going to be affected or whatever, who knows? But I think this is one where it’s like it’s a cost center. A lot of companies look at As, and they’re like, okay, if AI can do a reasonable job, maybe we’re willing to make changes there. So yeah, I just think it’s a big unknown. I think so far it hasn’t been really disruptive at all, but that doesn’t mean it won’t. There’s definitely been some companies that have made big moves, like Klarna got a lot of press for moving almost all their support to ai, but again, that is a very particular use case where they don’t have a lot of different types of questions.
Basically it’s like, I want a refund, how do I pay my bill? So I think a very good use case for an ai, they had poor service ratings to begin with, so it’s like, are the AI people going to do any worse? Probably not. So you have some differences there, but yeah, who knows? So I mean on the one hand it’s like those fears of just might consolidate down to a few big AI players who do a fantastic job with it and they just gobble everything up, or is there going to be, is it not going to be a thing at all or more of a middle ground where since we have access to a lot of similar technology, at least ostensibly, right, is we can pay for open AI or whatever else, we can add those tools for the most part, I think there’s going to be probably some things that are going to be beyond our capability, but so I think that’s definitely just a big unknown and we haven’t had that in a long time. And the help desk space, it’s been pretty stable. So this is definitely a new thing that’s out there and it’s like, we’ll see how it goes. We’re definitely adding AI features and maybe that’s just kind of be kind of where it ends up, but we’ll see.
Rob Walling:
That makes sense. Just out of curiosity, what’s the first one or two AI features you’re adding?
Ian Landsman:
Yeah, so very early on we did your standard writing helpers. We went farther than most of the other help desk tools I’ve seen have gone where you can define your own prompts and tools for the agents, but still it’s ultimately more on the writing and human creation side. But from there we are working on doing things like auto triaging where it can route inbound tickets to the right categories or right agents and things like that. So that’s going to be one of the first more sort of offensive minded, more active AI elements that we’re going to be adding. And then from there we have some different ideas about how we might do auto responding in a way that’s maybe a little bit safer. So obviously everybody’s trying to go for the holy grail of the AI just responds. You give it anything and it can just respond, but maybe there’s some in-between ground there where it’s more of a human response, but we use the AI to maybe figure out what’s the right response or some different things like that. So some stuff in r and d, some stuff very close to shipping. So that’s kind of where we’re at with it. But it is a big sort of sea change much when I started the company, which is like everybody’s using client server apps or just pure email, and that was the big shift and now it seems like there might be another big shift. I don’t know if it’ll be, it could be bigger than that. It could be less than that. I dunno, who knows, but we’ll see.
Rob Walling:
So last question for you today before I let you go, I asked offline, I said you named HelpSpot. How did you feel when all these other help desks launched with help? Right, so there was Help Scout and help somewhere else, and what did you say to me?
Ian Landsman:
Yeah, so those don’t bother me at all. There’s lots of whatever variations, but I thought the spot was kind of my unique thing. I liked the spot and Dharmesh Shaw, who’s the co-founder of HubSpot, was in Joel on software with us and kind of in our circles. And then he starts HubSpot and stole my spot like dude, and now it’s HubSpot. And obviously anytime I say HubSpot and HubSpot and people get confused even more than with Help Scout, although that one does get confused as well sometimes. But yeah, I’d say in terms of naming irks HubSpot,
Rob Walling:
HubSpot
Ian Landsman:
Kind of takes the pride from me.
Rob Walling:
Sometimes that happens, sometimes it happens. Seemed to have
Ian Landsman:
Done it. I like Dharmesh though. DH is great. Alright. He’s a great
Rob Walling:
Guy. He’s so gracious and he still, when I wrote this SA playbook, I emailed and said, would you give me an endorsement? And he’s like, of course. And he gave this great endorsement. I put it on the cover of the book. It was so nice. It was like, I’ve been learning from Rob for years. You should too or something. I’m like, oh, you’re the
Ian Landsman:
Awesome man. That’s crazy. He’s so good. Yeah, a few times I’ve talked to him over the years too. It’s just like, yeah, he’s still like the guy from in the for, you know what I mean? It’s like he’s a billionaire now, but he’s still kind of just humble. The same guy.
Rob Walling:
Humble and just,
Ian Landsman:
Yeah. That’s great.
Rob Walling:
Ian Lansman, thanks for joining me today. Folks want to keep up with what you’re working on, obviously helpspot.com to see the app, and you are Ian Lansman on X Twitter. Anything else you’d like folks to check out? I know you have a few podcasts.
Ian Landsman:
Yeah, the other thing, I guess the other main thing, if you want to keep up with what I’m doing is mostly technical is my podcast I do with Aaron Francis, and it’s more somewhat technical. There’s a little bit technical, but mostly it’s business and other topics, so it’s more entertaining, I’d say. Yeah, so check that out. We have a lot of people seem to like that, so that’s been fun the last year or so, working on that. Again, one of my little, I can’t really justify it on an ROI basis purely, but it’s a fun thing that I get to work on and is exciting to do.
Rob Walling:
Awesome.
Ian Landsman:
Yeah, so check me out
Rob Walling:
There. Thanks again for coming on the show. Thanks for having me. Thanks again to Ian for coming on the show, and thank you for listening this week and every week. This is Rob Walling signing off from episode 734.
Leave a Reply