Ian Tien and Corey Hulen started Mattermost, the team collaboration and messaging service, in 2011. One year later they were admitted to the Y Combinator Summer ’12 (YC S12) batch. At that time, the company was called SpinPunch and had almost nothing in common with messaging solutions, it was a game platform. The idea of creating a messaging workspace for security-conscious teams emerged from a poor experience delivered by a messaging platform the company relied on back then. Today, thousands of companies are downloading Mattermost every month, and the company is proudly serving enterprises like Samsung, Nasdaq, SAP, Cigna, Tesco, and others. More interestingly, the company grew this much without having an actual office — Mattermost team has been remote and distributed since day 1.
YouTeam co-CEO Yurij Riphyak spoke with Ian Tien, CEO and co-founder of Mattermost, about the motivation for being a 100% remote company from the very beginning, their decision to open-source, all the challenges of managing a remote team and even about the workplaces of future.
Yurij: Hi Ian. To start with, let’s talk about you and the origins of Mattermost. What did you do before creating this current company, and what about that experience led you to the idea of starting a business messaging company?
Ian: My background is Microsoft. Right after college, I was in program management and product management working on Microsoft Office. I ended up running Outlook.com and OneDrive businesses — well, at least their predecessors. After my time at Microsoft, I went to graduate school — it was a little bit of a break. While I was in grad school, I was always coming up with different business ideas and trying out a lot of different things. Eventually, I ended up founding Mattermost, which started off as an HTML5 game platform. The concept was to let you build for web and mobile at the same time, and we created technology that actually let us launch fast action games using HTML5 — it felt like it could be pretty interesting. So that was the original company, but then our technology sort of didn’t work out. We couldn’t grow fast enough on Android because of the hardware limitations.
After that, we spent two years as a video game studio because we had a bunch of prototype games, and games that were making money. At that time, we relied on a messaging platform that existed before Slack. Everything we did, analytics, data, collaboration, was happening over this messaging platform. It was built by a startup, which later got acquired, and then things started going wrong. The program started crashing, losing data, etc. We were really upset. So we tried to export our data out of this messaging platform, and they didn’t let us! They didn’t give us access to our own data. And when we stopped paying our subscription, they paywalled our own information. That emotion of really trusting a service, and then not being able to get your data back, and being really trapped in that system… we had something like an allergic reaction to it.
So we thought: inside of our video games we have around 10 million hours of people using our messaging service: why don’t we try to create our own messaging platform? First, we started using it internally and ended up rewriting the whole product three times and making it open source. And then it kind of took off, people really liked the way we built it. It was easy to use, easy to install, and it just turned into a business overnight. It seemed to be very promising, so we changed our business again — from video games to messaging. This is the story of how it started — we were solving a problem we had ourselves. We knew how important it is to have control over your own data in a messaging platform that you can trust. All we wanted was wanted control and transparency, and now we provide the same to our customers.
Yurij: Let’s speak about your decision to go open source. This was a conscious part of your strategy, right? To have the full transparency over the codebase, or were there other reasons why you decided to go open source with it?
Ian: So I actually was showing our product to Sid [Sijbrandij], CEO of GitLab. I explained what we built and why we were building it — and he suggested making it open source. I came from a pretty strong anti-open-source environment at Microsoft and was skeptical. Sid was quite compelling. He said, “It’s a prototype.” So, if you make it open source, and people like it, then you can always close source it later. But if you make it open source and nobody cares, you should just give it up. If you think you’re building something that’s valuable and you give away all the source code, and still no one cares, there’s just no point in building it. That logic made sense, so we decided to go open source. GitLab showed us how open sourcing works. It just really helped in general. We embraced that open source culture of sharing mistakes and successes. And, yeah, the way our business ended up being very, very similar to GitLab.
Yurij: I see, so there was a strong GitLab influence in your decision to go open source. Do you think being a part of this open-source community has influenced your company culture and your way of working at Mattermost?
Ian: Yeah, absolutely! Open source values are so deep in our culture. I’ll give you an example. When Corey, the CTO at Mattermost, and I worked for Microsoft together, we founded Magnus. When Corey and I left Microsoft, we left with nothing. Nothing, except all the years we spent there, all the technology that we created together. I’ve got about 30 patents, you know, in Microsoft Office. And these are in the closed repositories and some folder archives in the giant Microsoft database, but none of the engineers that worked there, including ourselves, have access to them. And that’s kind of not okay. A whole bunch of engineers from that original Microsoft team are now working with us at Mattermost. Our engineers use their personal GitHub accounts to code, not Mattermost accounts. We open-source most of the stuff we do by default. And what that means is, if you’re working here, you can kind of go back in time, and you can look at all your past work, and share it with the community.
There’s really a ton of transparency: it’s really important to us that people who work at Mattermost have it out in the open as part of their career and part of their history. There’s one engineer who was posting on social media the other day, and he was talking to his daughter about what he does at work, and she was looking through his code and asking questions about why his code looked different from hers. Then he was explaining that this software is helping people — it’s helping space programs, it’s helping the government, it’s helping financial institutions to have this massive impact in the world. So she got to read through it and see what her dad does at work while she’s learning to code herself. So, here’s your answer on how open source makes such a difference. I think it’s actually massive.
Yurij: Wow, that’s amazing! I know that you were operating remotely in the early days, even before going open source, right? Do you remember your reasoning behind it? Did you weigh the pros and cons of remote versus co-located? What made you decide to go remote?
Ian: Yeah, it just made 100% sense from the start. I started a company when I was in graduate school. Our first mailing address was my dorm room. While I was building this company, I got free office space in Palo Alto as a Silicon Valley startup. I showed up there with the person who was writing the game engine code. But the problem was, I was on the phone all the time, recruiting people, and the other person was writing code the whole time. Being in the same office made no sense because we were just interrupting each other all the time. So I went back to my dorm room, and my friend went back to his place, 15 minutes away. And we just kept going that way for years, once in a while getting together for dinner or something. There was no need to meet. We definitely didn’t hire in Silicon Valley because we hired artists. We had them in Los Angeles, in Toronto, a lot of freelancers. When we needed to have a conversation, back then we used Skype. We would just hit the button and jump on the conversation. It could take five minutes, it could take three hours, but we were tremendously more productive than if we all were in the same room. And then from there we became addicted to our ability to source really great talent from anywhere in the world, being free from waiting in the lobby for a meeting room, and then getting kicked out of the meeting room, and eventually, you just have this massive overhead.
“So remote for us is really about productivity and talent.”
Yurij: That makes sense. I’m sure you heard this remote-skeptical thesis that remote hiring is always a plan B because the winners can easily hire the best talent in their geography. Can you reflect on that?
Ian: We can hire people from the Bay Area or Toronto out of other companies because we provide a much stronger value proposition over local companies: no traffic and commuting. If you only hire people in your area, it means they will have to commute to work. Commuting can take hours sometimes. And here’s the logic, what if at the end of the interview loop I say, “By the way, we’re going to ask you to spend 500 hours a year just sitting in your car, are you okay with that?”. Like, imagine you had to say that out loud. And when you think about it this way, you immediately want to start working at a remote company. An hour and a half, two hours long commute is not uncommon, so we think it’s insane to think of hiring people to work in your location. The math doesn’t work out, and even the value prop because people can be just as effective or more effective working remotely. All in all, the math doesn’t work for being an office.
Yurij: Yeah, but there are still challenges, right? You mentioned Sid from GitLab. I remember in my interview with him we were going through Seven Deadly Sins of remote, the most widespread challenges of remote work. For example, limited spontaneous communication or distractions associated with working from home or a lack of synchronicity between parts of the team and so on. So, what are your biggest challenges? And how do you address them?
Ian: I think that a lot of the things people mention, like timezone issues and distractions, aren’t actually that big of a deal. Working remotely does not necessarily mean working from home. It just means not working together in the same place with your team. You can find a WeWork or coffee shops where you can stay if this helps to avoid distractions.
One thing that works really well in the remote setting is when problems are being solved by a team of only two people. When a problem is talked through and resolved by two people, instead of a whole committee, we call it “pair design”. Working in pairs can be really powerful. When two people agree on a solution, they can take it to a broader committee.
What’s not easy in working remotely is that sometimes there are certain problems which require five to seven people to camp out in a conference room for a whole week, knocking these problems out. A good example is annual planning. When working in the same office, you can sort of force people to go into a room and make decisions. But it gets much harder when working remotely because everybody has their own important meetings. So it might be hard to get those four to seven people into an online conference room to work on a single topic for an extended period of time.
So, in my opinion, you can get over most of the challenges when working remotely. But there is a thing that remains particularly challenging.
Yurij: Let’s talk about the onboarding process at a remote company. How do you handle this?
Ian: We actually get a lot of positive feedback for our onboarding process because it is all documented. When in the office, nothing is written down, you just talk with certain people and they basically pass on their knowledge about how things work. When you’re a remote company, you’re forced to document everything, so onboarding includes a lot of reading through information. Still, even remotely, you can talk to a lot of other people, and you feel welcomed, and you spend some time getting to know each other. But because everything is written and prepared beforehand, you get higher quality interactions, having removed all the administrative stuff out of the way. And when you’ve read all the documentation, you understand the context of the upcoming meetings better, and you can start asking much deeper questions even in your first interactions.
Also, the tricky part about working remotely, especially over so many countries, is hardware and equipment. The person who procures laptops for employees from 18 different countries has to have multiple credit cards because it’s required by law. It turns out that when you buy laptops and ship them all over the world, it looks like credit card fraud. It’s just harder to onboard people really quickly and to make sure all the hardware is there in all the different countries. These complicated, weird patterns get flagged for fraud all the time. But onboarding, documentation, — it works really well in remote settings. But there are some snags around logistics.
“There’s also a misconception that culture is stronger in a building versus remote. And I would disagree with that. I would say culture is stronger when you write it down and document it because everyone on board is going to get a very consistent and similar experience.”
And since it’s consistent, it’s much more scalable. So I like to say that remote onboarding when done right is much more powerful and efficient.
Yurij: Okay, that makes sense. Thanks for sharing that. So, everyone is different, right? We are all wired in different ways. Do you think remote work is a good decision for everyone, or are there specific individual traits that make one person more or less receptive and successful in a remote setting?
Ian: Unfortunately, yes, there are people who are less suitable for working remotely. Even before the pandemic, there were people who really liked working remotely and those who really didn’t like being remote. Now, everyone has to work remotely, and you see that there are those who miss the office and those who think working remotely is amazing. I think communication skills are what make people successful at remote work. So the answer is yes, there are people who will function much better in an office, for example. It’s much harder to work remotely if you don’t have strong written communication skills.
“I would say that the number one characteristic that determines success with remote work is how strong of a writer you are, and also how well you consume information through writing.”
Yurij: I strongly agree with that. I think writing is the number one thing that makes a difference. And it’s actually good because communication becomes much more organized, discussions are easier to revisit, it also forces you to keep it short, and so on. So, yeah, there are lots of quirks with remote work. Let’s talk a little bit about the tools. Have you ever thought about building a tool yourself that would aid you in managing your remote company?
Ian: Oh, good question. I would say a whiteboarding tool is something that really hasn’t been solved yet. But it’s more of a hardware problem than a software. There is no hardware yet, and I think people miss whiteboards.
Yurij: Makes sense. So, that would be a whiteboarding tool. But what is the problem with it? What’s missing from the existing tools?
Ian: I’ll give you an example. You and I are having a conversation and I want to explain a concept to you. In-person, I would want to take a pen and just sort of draw and explain to you what I’m thinking. But working remotely, I have a PowerPoint slide and I just can’t do it. I can’t explain something to you as fluidly as I could on a whiteboard. Miro and Mural are good solutions but they’re still on the way to simulate physical collaboration better. There’s a gentleman named Hiroshi Ishii at the MIT Media Lab. He has been studying for a decade about how to bridge the gap between the digital and physical world. I think that is one of the pieces we’re really missing. Everything else with a voice and video and screen sharing, and messaging, and workflows will come and we’re doing a pretty good job at it already. But what I think the world right now is really missing ist that physical to digital connection. And I think whiteboards are a very important piece that we don’t have yet.
Yurij: Got it. This brings us to my last question. And it’s broader, it’s about the future, and about the impact of remote work. So let’s say we have borrowed a time machine and traveled into the future. Let’s say 20 years from now. So the year is 2040 now. The remote work paradigm has of course taken over the world. Every company has a significant part of its staff working remotely. So, what will be strikingly different from now? Something that is not here today, but that in 2040 every business has.
Ian: Great question. I think sports will be very different. If you look back 30 years ago, there were lots of different sports done in person but computers were not so amazing. And if you look at today, eSports is really big right now, it’s filling stadiums. You can see it is trending even more than physical sports. Look at TikTok, look at Fortnight, mobile games. That’s how people are spending their time, and that’s what they’re going to be nostalgic for. And I see how our free time is becoming much more digital and much more alive, and the way we spend our personal time is really going to transform.
I think if you look demographically at how the younger generation is spending their time, it is completely different from that same age group 30 years ago. I think that’s going to have a pretty big impact in the future. Live sports are just one example. People are going to be online naturally, like digital natives. And that’s going to really shape the work environment, the way we do business conferences and networking and business development, it’s going to be a lot more like Ready Player One.
Yurij: Do you mean things are going to happen in virtual reality?
Ian: I don’t know much about virtual reality and headsets but I think that online is going to be rooted in us. Let’s take the G Suite and Zoom. They have massive adoption and shared traction because it’s what we also use in our personal lives. And, eventually, your work stuff looks the same as your personal stuff, right? And now think how Fortnight or Discord would look like in the workplace.
Yurij: Very interesting. Well, that would be all from me. Thank you so much. It’s been an amazing interview, lots of insights, I’ve heard lots of interesting things I didn’t expect to hear.
This interview is just a tiny part of our huge project – a series of founder stories shared by YouTeam. Enjoy your reading and keep coming back for more opinions from the world’s most successful tech founders!