Andrew: Hey there Freedom Fighters. My name is Andrew Warner. I’m the founder of Mixergy.com, home of the ambitious upstart and home of the in-depth interview with entrepreneurs who talk about how they built their businesses and all the steps along the way that only real entrepreneurs care about and the rest of the world would probably be bored by, but for us, it’s vital.
Now, three years ago I invited the co-founder of Trello to talk about how he launched his startup. Well, today I’ve got Trello’s other co-founder on to talk about how the company grew to six million users. My goal for this interview is to understand how the way a product is built can help you acquire more users.
Michael Pryor is my guest. He is the CEO of Trello, the free, flexible and visual way to organize anything with anyone. When I organize even events over at my house, Sunday brunch, I use Trello to make sure that I keep track of everything. You’ll find out more about the tool in a moment.
But first, I should say this interview is sponsored by LeadPages.net. If you’re like me, you need a way to get people who are just hitting your website to convert into users. One good way to do that is to create a great landing page that explains what your product does and explains to people why they should give you your email address and start a relationship with you.
Well, I’m not great at designing those pages, so I got to a company that is great at it. They’re called LeadPages, LeadPages.net. Go check them out by going to either LeadPages.net or if you go to Mixergy.com/LeadPages, I’ll take you to a bunch of their best designs so that you can see why I’m so excited about using them.
Michael, welcome.
Michael: Hi, Andrew.
Andrew: Hey. You know, Joel was on here. I’m so surprised that you are on here because when I Google Trello, when I Google other businesses that Joel’s involved with, his name is so attached. I keep wondering, do you ever feel overshadowed by him?
Michael: Well, I’ll tell you something. This is my first live video interview with anyone ever.
Andrew: You know, thank you so much for doing it.
Michael: I imagine as CEO of Trello, I’ll probably have to be doing a lot more of them. But Joel has spent the last 15 years building up his own brand. His blog is read by millions and millions of people and parlayed that into a successful software company, Stack Exchange, that all the people initially started using Stack on referral were from reading Joel’s blog and also Jeff Atwood’s blog.
We both have different strengths. I remember the first time I ever went to see Joel speak it was probably about 15 years ago. It was a room of like 20 people. The difference between that time that he spoke and when he goes in front of thousands of people now and does a keynote, it’s a big difference. So, it’s something you’ve just got to learn. I’m taking a bigger role at Trello now while Joel focuses at Stack Exchange. So, it’s something I just have to learn to do.
Andrew: You guys met up when you were working at Juno, right?
Michael: Yes.
Andrew: That’s the company that was going to focus on email and give you free email that you needed a whole other piece of software to connect to.
Michael: That’s right. That’s back in the day when everyone was dialing up on AOL and you had to pay a monthly fee. Their idea was, “We’re going to do the dial up part for free and it will be advertising sponsored.” First it was just email and then they added in internet access to compete with AOL and kind of merge with NetZero, who was the other company that was doing that at the time.
Andrew: I read in the book “Founders at Work” that Joel said that you and he worked together. You had no idea what business you wanted to start. But you just thought, “You know what? We can build a business where developers are treated well and the idea for what those developers build can come later.” Is that right?
Michael: Yeah. If you look back at the history of Fog Creek, there are probably about 13 different projects that we did and a lot of them were failures. Stack Overflow Careers, which is the way that Stack Exchange makes money, was born out of Joel’s job listing board that he had attached to his website. One day we had the idea to make an Indian job board for Indian programmers to find jobs. I think we made 24 rupees on the whole thing. But it was a huge failure.
We made a content management system that was a desktop application that you ran on your computer and it uploaded all the files to the server because back in the day in order to have a CMS, you had to have shell access to the web servers so that you can install these Perl scripts. People didn’t have that at the time. If you wanted to run Typepad, you had to install these Perl scripts. The web hosts at the time didn’t allow you to get shell access.
So, we thought, “Oh, we’ll build a desktop app that you run on your computer and upload.” At the same time, Typepad was seeing the same problem and was like, “Okay, we’ll just host it ourselves.” Now that model won. If you want a CMS, you basically sign up and create an account immediately. You don’t have to install any scripts.
So, we had a lot of different things that we tried over the years. Some of them were failures. Some of them were successes. Stack Overflow turned into a huge success. FogBugz and Kiln, our developer tools business, the biggest part of Fog Creek right now, is a huge success. We made a screen sharing app called Copilot before LogMeIn existed. It was a very small success for us, but that idea could have been turned into a giant business. And then Trello is a huge success. But there’s a lot more work that we have to do with Trello.
Andrew: Fog Creek Software is the main business that you guys started together. From there, all the other ideas sprang out, some of them failures, some of them successes. When I look at where Trello, the business that we’re here to talk about, came up, I found this quote from Joel where he said, “Bing. A light went off in my head. The great horizontal killer applications are actually just fancy data structures.” Is that really how the idea that now has six million registered users came about?
Michael: Think about a Word doc. It’s just text on a page. A PowerPoint presentation or slides are just those pages presented in a visual format on a screen. A spreadsheet is like a table of numbers. At its core, those things, that object, the idea of that didn’t exist 40 years ago. Somebody had to create a spreadsheet and sort of tell people why it was useful and what it was good for or even a database. People just assume those things existed.
I think what we’re trying to do at Trello is create another horizontal document type to give people structure to parts of their life and their work that they didn’t have before. So, that’s kind of a tall order. It’s a challenge because the word for that doesn’t really exist.
Some people call Trello a project management app. I think those are people that are using it at work for a software project. Some people call Trello a to-do list. People are using it themselves to keep order in their own life. But really, people use it in so many different ways. It’s so lightweight that we’re trying to get people to use it to just give structure to their life. A tool like that doesn’t exist currently. There’s no word for it. So, people aren’t going to search for it.
So, if you want to plan your Thanksgiving dinner or figure out what gifts the extended family is going to buy for the kids this Christmas and you want to collaborate on that, there isn’t a thing that you search for on Google and it just pops up. So, I think that’s a huge challenge for us, to sort of get people to understand that Trello actually fits all those use cases.
Andrew: I get that. Frankly, even for myself, I had Joel on here to talk about Trello. I signed up for an account. I used it but I wasn’t fully sure what to use it for, how it would fit into my life. It would just be another place to store some ideas. Then I had a guy come into my office and he said, “Here’s how I work with my virtual assistant. She has this board where she and I collaborate and I put my to-do lists on this list. When she’s done with them and she’s ready for me to see them, it goes on to another column.” I said, “Oh, yeah, I can use that.”
Now I use it for both collaboration with coworkers but also I create these Trello templates that I hand out to anyone who wants to do an interview and wants to keep the interview booking process organized. I say, “Look, you have a column for suggested interviewees. Then you have a column for finding their email addresses. Another column for inviting them and another column and so on. And each guest that you potentially invite gets a card on the first column.” and then you keep moving them over column by column until you’ve done an interview.
I get it. I feel like it is a tough thing to explain and to understand. So, I want to understand how you guys communicated it to people. But let’s first go back and see where the origin started. You didn’t set out to create a project management tool or a horizontal structure for organizing data. You guys said, “You know what? We want to go after civilians,” is what I read Joel Spolsky say. Why civilians as opposed to developers?
Michael: Well, there are a bunch of ideas that went into Trello. One of them was that we didn’t want to build an app that was just for developers. We had built FogBugz, Kiln. We had built Stack Overflow. We had a huge success from Stack Overflow was actually getting it broader and going out to Stack Exchange, even though Stack Overflow by itself is a huge success.
We wanted to do something where we create a mass market product that there is just a huge demand for it and tons of people get value out of it. That’s also why it had to be free because we didn’t want there to be any friction to it spreading. We knew that we could find the people that are getting a lot of value out of the product and get them to pay a small amount.
I know it was just one of the things that was a challenge. We saw a lot of different pieces coming together, like at the time, you would go to any software company and you would look where the developers were and they would all have a whiteboard and it would have Post-It notes on it and they would be in columns. They would manage their software development process by moving those Post-It notes across the board. It’s a style of working called Kanban, which is very familiar to developers, but not really known outside of that.
That was going on. We were seeing that happening everywhere. That was really useful for these agile teams that were moving really fast and switching away from the old way of tracking things like where you basically had this huge database of items that you had to do and you were just adding more tasks and more tasks and the old style project management app where you would be like, “What things do we have to do for this milestone?” And it would give you a list of 100 cases.
That worked well. I think the way that companies used to work, where you would ship software on a fixed schedule, you would put a bunch of features into a release and then you would ship that release on a certain date. Now you have more web apps where it’s like, “I’m going to make a feature and ship it next week and then I’m going onto the next thing.” It doesn’t make sense for me to think about things that might happen in three to six months because the app can be totally different at that point in time.
So, we sort of felt the need for that ourselves. We also felt the need for collaborating with our company and sort of figuring out what was actually going on to sort of get a higher level view of what people are working on.
Andrew: You mean you internally needed this tool too?
Michael: Yeah.
Andrew: Okay.
Michael: Like, you would be at work and you’d be like, “I don’t know what everyone’s doing.”
Andrew: I get that feeling.
Michael: Not on specific micromanaging, but just is the company moving towards its goals? Is the company getting closer to shipping?
Andrew: Okay. So, FogBugz is bug tracking software built for developers. Stack Exchange, question and answer site that you guys built starting with developers and then moving on to the rest of the world, including cooks and people who are into grammar and math, etcetera. Now you said, “We’re going to start with a mass market product. We can reach a big audience.”
Joel, at the time, told me he wanted to reach 100 million people. You told me when I started this conversation with you before I even hit record, “Yes, we’re up to 6 million. Our goal is 100 million. So, we’re only 6% of the way there.” I see how you wanted to think bigger. I’m also getting a sense of where the idea for where the structure of the software came from, those Post-It notes that people used to keep on whiteboards, etcetera.
Before you built the first version, did you interview people internally at the company and say, “Look, we’re thinking of building this. How would it fit your needs and solve a problem that you have?”
Michael: So, that genesis of the idea was that at Fog Creek we have a culture of innovation. We try lots of different things. I talked about that before. At this moment of time, what it was four years ago when Trello first launched, before that when we were first building it, we had a couple of ideas. We took a few developers and worked on those different ideas.
One of the ideas was Joel had this idea for a to-do list that you could only put five things in. It was called Five Things. There were two things that you’re working on now. You could switch between them. If you’re waiting for this code to compile, you could work on the other thing. Two things that you’re going to work on next and then one thing you were just not going to work on at all. And that’s it. Those were the only things you could put on the list. If you wanted to put something else on the list, you had to take something off.
That was just sort of like this idea that was trying to get at, I think, that feeling of, look at the way people use email. Their inbox grows and grows and grows because they put too much stuff in it and then they get frustrated. We have a hard time as humans, I think, letting go of things that we might want to do in the future. This was like forcing you to do that. It was like, “You’re never going to get to those things. So, if you keep them around, it’s just making you feel worse because it’s this backlog of stuff and you’re slowed down by it.”
So, this idea of Five Things was just to free you from that and just to accept the fact that you could only really concentrate on a couple of things at once. That worked together with the whiteboards that we saw and the Post-It notes and we sort of used some of those ideas and Trello grew into what it is today. Essentially we started using it internally. We saw the devs were actually using it after a couple hundred lines of code, like 900 lines of code were written. They started using it to help them build Trello.
Andrew: How many developers did you have at the time?
Michael: Oh, like three.
Andrew: Three or four developers then were using this as a way of keeping up with what they were all working on together.
Michael: Right.
Andrew: Okay.
Michael: And then Joel was mentoring a Techstars company, which was a couple of people. We had them start using it. We basically took people that we knew and showed it to them and asked them if they would use something like that and just kept iterating on that.
Andrew: Do you remember one of the pieces of feedback from your users that helped shape the product, the early users?
Michael: Well, one thing, I know the reverse of that, which is that because of the people that were using it initially, a lot of the feedback that we got was from developers trying to turn it into a developer tool. If you look out there and you look for Kanban or Agile management tools, there are like a million. They’re very structured about the workflow that you have, the terminology that it uses.
They’re tools for developers, like the websites are trying to attract developers to use it, the marketing materials are for developers, the features are for developers. When we started Trello, we knew that was something we weren’t going to do. So, when people would give us that kind of feedback, it was nice to understand that this was something that we weren’t ever going to add, like adding priorities or swim lanes, having the cards be in a certain row and things like that.
Andrew: I see. Because you said, “We are going to go for a mass audience,” even though you had user demand that they really needed this extra feature, you said no.
Michael: Right.
Andrew: I see.
Michael: Now, one thing we realized later was that we had lulled ourselves into thinking that all of the initial users were more technical. So, it meant when we would launch a new feature and get feedback that sometimes we were able to discard what they were saying because we were thinking, “Oh, they’re kind of jaded because they’re developers,” which meant that it was much harder for us to figure out whether the feature was actually good or not. I’ll give a more specific example.
Andrew: Yeah, please.
Michael: We added stickers that you could put on the front of Trello cards. So, there would be like a sticker pack. There was no semantic meaning for that. It was just a sticker. It was just visual. We thought that developers would hate that. We didn’t do it because we thought developers would hate it. But we thought that the feedback from developers was they wouldn’t like it because it was just adding cruft. There’s no way to search for the types of stickers. They would want labels and tags and things like that. This was just more of a fun thing to make Trello more fun because that was also one of the tenets that we wanted. We wanted the product to be fun.
So, when we first started getting the feedback from the beta testers, we initially thought, “Oh, well, all the people that sign up for the beta were sort of self-selecting.” They were people that really were into what we were doing. They followed us. We sent out an email. They replied. So, those people are generally more technical in nature. So, their feedback isn’t going to be legitimate for the consumer market that we’re really going after. That may or may not be true.
What I realized at that point in time was that the way that we select people to give us feedback has to be completely random. It has to have a mix of all kinds of people, not people that self-select to be beta testers. Because if you just go out and ask for beta testers on Twitter, you’re going to get a certain type of people. Does that make sense?
Andrew: Yeah. Number one, a user who is a Twitter user back then and number two, someone who’s following you is probably more into development and startup environment.
Michael: Right.
Andrew: So, what did you do to broaden the pool of beta testers beyond the tech community that you guys had close relationships with?
Michael: One of the things that we saw happening, we were focused on just building a really good horizontal product and then letting people talk about it and spread it through word of mouth. We had hoped that would happen organically because we were just a small company. We couldn’t spend a lot of money on marketing. In fact, we just hired a marketing person this summer and just started doing those sorts of activities.
Andrew: This is Stella?
Michael: Yes. What we saw was a bunch of people started writing blog posts about how they were using Trello for this or for that, like UserVoice wrote a blog post. Ryan Carson from Treehouse wrote a blog post. There was a guy from PayPal that wrote a blog post about how he uses it for recruiting. Heroku was using it for onboarding. They would create a board with all the things that a new employee had to do or learn about. And then every time a new employee would start, they would clone the board and put the person on that new board. And then if the process changed at all, they would just modify that original onboarding template.
Andrew: Forgive me, but all of those people who you mentioned are very technical, very startup-oriented people. So, by even having-
Michael: But they were using it in ways that weren’t software development ways.
Andrew: I see. So, even though they were using it to onboard new employees, it’s not a technical use. It’s more of a business-oriented use. That’s what broadened… I see.
Michael: So then you would see if it got into a company and it went into a tech team, it would jump over to sales and marketing or it would jump over to HR. We were seeing that by the blog posts people were writing. We knew that that was happening. Now, to what extent that was happening it wasn’t really clear and we’re not going to look at people’s data and we don’t ask people what department they’re in.
Andrew: What about this, what did you build into the product that allowed it jump? It wasn’t just happenstance.
Michael: That’s like asking me like how do you make a great product. There are so many things that go into it.
Andrew: Give me some of it. What’s one big important thing?
Michael: It had to fun. It had to be free. If it wasn’t fun, people wouldn’t use it outside of work. If it wasn’t free, there’s a lot of friction. You’d also have a lot of competition. It’s collaborative by nature. So, Trello is much more useful when you’re working with other people. You can certainly use it for yourself as a to-do list. But it was designed to be used by a whole bunch of people.
I think it’s so lightweight and horizontal that if you saw somebody else using it at your company, you’re in sales and the developer, you’re trying to manage some custom work for this one product and you see them using it, you can immediately understand that metaphor because it’s so visual and it already exists in the real world. People do this. They put Post-It notes up on their monitor and it works with the way people think spatially. You put things on your desk in certain places and they mean certain things or even on your desktop.
So, it just took advantage of a lot of those different things and just all came together to be a great product. And then people, you just cross your fingers that they’re going to recommend it and then other people are going to use it and then they did.
Andrew: I see. So, if I, as a business owner, used it internally but wanted to recommend it to a friend or have someone else in the company use it, before they got to use it, I’d have to buy a seat for them or another license. That would be too much friction and I wouldn’t pass it on. That’s why you decided that you would make it free. So, free allowed it to grow.
Michael: Right. Because we felt that the business that we wanted to build was a huge audience of people, like Evernote, for example. It’s the same model or Dropbox, where you’re giving a lot of value to people but the people that get a ton of value, those people are paying you.
Andrew: You know what though? Evernote didn’t realize the collaborative part of their product until later on in their evolution.
Michael: Very early. Yeah. I think Evernote is more of a personal product. They’re struggling now, well, not struggling, I shouldn’t say that, but they’re trying to figure out now, “How do we push the business aspect? How do we get people to put their business stuff in here?” It’s much easier to make money from businesses than it is from consumers.
Andrew: Right.
Michael: So, they started from consumer, work by yourself, put all your notes in there, not data that you would want other people to see. Now they have to convince people to use it collaboratively, whereas Trello started collaboratively and lots of people use it by themselves because it’s already on their phone.
Or you use it at work and you invite your husband or your wife to use it at home because you want to get more structure to all the projects that you have to do around your house or you’re doing a kitchen renovation and you invite your contractor on a board because you really want to know whether that renovation is moving forward or not. You just want to give yourself more clarity into that project.
Andrew: Michael, as far as I can tell, I’m an outsider and that’s why I have you on, so that you can give me a check on this. I feel like it’s more than just the inherent nature of this kind of product that allowed it to spread. I’m looking here at one of my Trello boards, the one that I use with our developers. On the right side, I see an activity list. So, I can see what he and other people on the team have worked on.
But at the very top, even as I scroll, I always see a button that says, “Add members.” By making it so prominent, by making it so easy to add new people, you encourage me to bring more people in this conversation and allow the software to be more viral.
Michael: And we’ve done a lot of different things related to that. We have a consumer product called Trello Gold where you get custom board backgrounds and more stickers and just fun stuff, like not things that you would need to get a lot of value out of Trello, but if you’re a super fan and you want to give us some money, there are people out there that were like, when we first launched and it was free, there were people that were like, “I’m not going to use it because I don’t trust that it will be around if I can’t pay for it.”
So, we kind of launched Trello Gold as a way for people to give us money and get some fun features. We also give it away for free if you invite more people to Trello. So, if you get a user to use Trello, then you’ll get a free month of Trello Gold.
Andrew: So, that’s another reason why the product spreads.
Michael: Yeah. I’m not trying to build a whole business on top of the revenue from Trello Gold. But it’s definitely useful to help encourage and give thanks to people that spread Trello and get other people to use it.
Andrew: How much revenue comes in from Trello Gold?
Michael: It’s negligible. It’s not a lot of money.
Andrew: I see. So, the only reason you put a price on it is so when I get it for free I feel like, “Oh, I’m not just getting these cool features.” You’re nodding. Yeah.
Michael: Yeah. It has a value. There are people that pay for it. It’s certainly the case that we’re asking you to do something for us and then giving you something in return. But the real revenue generation from Trello will come from businesses that are getting a ton of value out of it and are happy to pay for it. We just hired a sales person, actually, two months ago.
Going back to what I was telling you before, like one team will use it here and it will jump to another team that will use it here. We found that there were a lot of big companies where they had hundreds of people using Trello, but they were all spread out throughout the organization. It might be a team of ten here and a team of 20 here. They weren’t connected in any way.
But the organization, the bigger the company is, the more rules they have about purchasing. They’d rather just pay once for everyone and use the invoice and PO system. They have a 30-page contract that they want you to sign. So, there’s like a high barrier entry. We just started helping companies be legitimate. I think a lot of teams within these companies would just use it and not worry about that stuff. But once you sign the contract, your foot is in the door and it makes it a lot easier for people at a big company to use it.
Andrew: This is Trello Business Class that you guys have added.
Michael: Yeah. That product that I was discussing was Trello Enterprise, which is just sort of a layer on top of Trello Business Class. But Business Class is a bunch of features around an organization that’s using Trello. So, things like member management, somebody leaves the company and you want to reclaim the boards that they own, exporting your data because you’re worried that Trello might disappear or something like that. Things that are of a concern to a company that is running their business off of Trello.
Andrew: Who can make a board visible or non-visible.
Michael: Yeah, permissioning, things like that, but nothing that’s going to prevent a team, for example, at a company, from using Trello and getting a lot of value added. So, you’re at a big company but you don’t have any budget, you can use Trello because it’s free and you don’t need Business Class. You can just start using it. And then later once the business is getting a lot of value out of Trello, we can add on a couple features and services on the back and the business is happy to pay for it.
Andrew: How did you know what features to add in to that?
Michael: It’s not done. There’s probably a lot more stuff that we need to add.
Andrew: Did you have conversations with your customers?
Michael: Yes.
Andrew: You did. What’s your process for talking to customers and learning what they need?
Michael: We did a ton of interviews like this. We’d do video interviews and record them and transcribe some of the important details. Now the sales person that we hired is calling all these people and talking to them and getting data from them. We have one of the designers working with a lot of new users to Trello and sort of looking at the new user experience and how you get into Trello and is there anything we can do when people first sign up to get them much more engaged later on and not like lose them immediately. But it’s just people calling people, asking them if they want to talk.
Andrew: What’s the process? So, you call someone and say, “Do you want to have a conversation?” A customer says yes. You do video like this. If you’re trying to figure out what the onboarding is like, I’m guessing you want someone who’s never been on the site before, right?
Michael: Yeah.
Andrew: Okay, so you get them on.
Michael: There are a lot of different ways to do that. At Fog Creek sometimes we’ve actually paid people to come in like using Craigslist or something. The problem with that is you get, actually, it would be good for Trello, but in the past we were using it for products that were more geared toward technical people. You’d get like a very wide range of people if you just used Craigslist.
But if we contacted a company that was using Trello and had just started using Trello, it’s very easy for us to find people at the organization that haven’t had experience with it yet and talk to the person who does have experience in sort of introducing it to the organization and talk to them about how they’re inviting people to the boards and then later interview the people that got on the boards and see what was confusing about when they signed up and things like that.
Andrew: I see. Now, I’m acting like you’ve been on board from the beginning, which you were and you’ve been in there every single day. But you took some time away to work at Stack Exchange as the CFO, right?
Michael: That wasn’t my only job. I was sort of doing a lot of different things. Part of my role was doing the financial side of Stack Exchange. But my background is a developer. I have a computer science degree. So, that only got us so far. Thankfully, we hired a real finance person recently.
Andrew: How did you end up being the finance guy for a venture-funded, well-funded company?
Michael: When we first started Fog Creek, I think the first five years of Fog Creek, Joel did all the operational stuff, so, the payroll and the QuickBooks and all of those things and then we switched. We traded off. I kind of got stuck with it a lot longer. And then when we started Stack Exchange, there was no one else to do it, so I just took it on.
When there were like eight people, it wasn’t that much work. Gradually over time, as it got more and more people, it became more and more. Eventually we got to the point where we were like, “Hey, we need to hire a person with a real finance background.” It just happened to coincide with a point in time when we were like, “Okay, and also that will free me to spend more time focusing on Trello.”
Andrew: I think it was you on AlleyWatch.com who said, “Raising money was easy, relatively speaking,” even though you guys at the time had raised $10.3 million. “But separating Trello from its parent company was the tough part.” What was the challenge there?
Michael: It’s a people issue, the culture. We had been through this before with Stack Exchange. So, we knew what we were getting into. I think that’s partially why it took us… we bootstrapped Trello for three years before we took VC. The VC people that we had experience with at Stack Exchange, we knew people. They wanted to invest years before we actually decided to take money. But it’s hard to sort of separate a business and people’s identity with their company and, “I wanted to work for that company and now I’m working for this company. Those were my coworkers. Now I have different coworkers. Where’s our office going to be?” It’s a difficult process.
Andrew: Did it feel a little bit like playing dodge ball in school where you separate into two teams and you and someone says, “Why did I get picked on your team?” And someone else says, “Why did I get picked on that team?”
Michael: Yeah. It was a little bit easier because at Trello we kind of saw this coming. So, we were running the business where people were either on the developer tools business or they were on the Trello business except for a few operational people, like finance and recruiting. So, there wasn’t a lot of overlap in the group.
So, people knew that they were at Trello or Fog Creek. But we were happy to talk to anyone that wanted to be on one side or the other and figure out what was the right spot for them. It was actually, I think compared to Stack Exchange, the way that that went, it was much easier, a much better process.
Andrew: So, what’s the challenge? So far it seems like everything is fairly easy. Bing. An idea goes off. Boom. You end up sending an email to a few of your users and they come in. You get some feedback and the product evolves. Has there been any challenge or is it all just smooth sailing here?
Michael: I think when we decided to take VC, we signed on to a whole new challenge. If we just grow the way that we’ve been growing, that’s not succeeding at the challenge that we signed up for. We raised money so we could accelerate the growth and really ramp it up. Sitting back and saying, “Oh, we’re getting more members every week.” That’s not going to cut it for getting to $100 million.
Actually, I found a chart where Evernote published their user signup numbers and put Trello’s signup numbers on top of that to see if we were performing. They recently passed 100 million signups sometime this year. The far right side of that hockey stick gets really steep. It’s a big challenge to grow.
One of our top focuses right now is internationalization, just getting into a lot of different markets because we know those people are already using Trello but we can just open up a great product to a whole bunch of people that don’t speak English.
So, that’s a huge challenge for the company, not just from a technical nature but it adds all these things on top of your process now. You have to do it in a way that doesn’t slow your organization down because, “Oh, now we need to translate that to Portuguese or now we need more marketing materials for all these different languages. We need more support people to handle all the incoming queries.” It definitely gets more and more complicated as you get bigger and bigger and go after that consumer market.
Andrew: No. And I don’t see the chart that you mentioned, but I do see one in Business Insider about Evernotes growth to 5 million users. It looks like they hit 5 million users in 83 days. Is that right? No. That can’t be. No, no. I see there’s 83 days to go from 4 million to 5 million. That’s the difference.
Michael: Right. And then that starts to get much smaller. The time to add another million gets smaller and smaller. I think it was two months ago we were at 5 million or something like that. We did a big hoopla, a big marketing campaign.
Andrew: I saw that.
Michael: Yeah. So, that was five and now we’re at six and a half. So, it’s definitely growing faster. But it just gets faster. You have to grow faster and faster and faster.
Andrew: There’s another article that just happened to come up as I was searching for Evernote’s data that you mentioned earlier. It’s called “Five Moves that Fueled Evernote’s Impressive Growth,” and it’s by Mashable. If we had to say the same thing for Trello, what are five moves that allowed you guys to grow to 6 million people so fast, within three years?
Michael: Well, the most obvious one, which is not something you can just set out to do, is to build a product that people love. The thing about Trello that was different than all the other things that we’ve built because we’re building an app for everyone instead of just a particular audience was when we would see the tweets that people would say about how much they loved Trello, it was just totally different than anything we had ever built before. We knew that we were doing something right.
Andrew: What was that thing that you did right?
Michael: If I knew, I’d make that a product. It’s just a combination of so many different things. People would write tweets where they were like, “I wish I could marry Trello.” It’s like, I don’t know. They’re being goofy. But what makes people feel that kind of connection with a software product?
Andrew: Here’s what I’m getting. I’m over-simplifying it, but part of my job here is, and part of our job together, is to simplify the complicated for the audience and give them something meaningful that they can use. So, what I’m getting are conversations with the real users, including watching them use the product in person, a vision for what the product is that’s based in reality of seeing how people are using offline tools in the real world, dog-fooding, dog-fooding, dog-fooding, right? You guys have used your product from the beginning.
Michael: I think also a big one is design.
Andrew: Okay.
Michael: If you look at like why does Medium feel so much better as a content platform than like writing a Typepad blog or a Blogger blog. You know the hero bar, the big picture at the top, I just see that and I’m like, “Ah…” I feel good about reading that. The typeface just looks nice, the white space on the side. It’s not that different than other blogging platforms. There are definitely different features and those sorts of things, but what makes it so different, there are these little touches.
So, we have little touches all over the app. We have a little spokes-husky, a little dog, Taco, that pops up every once and a while, just rarely, to tell you new things that are in Trello. Just the way the tactile nature of dragging the cards around, the fact that it feels just like, on an iPad, actually, it’s a great experience. Just small little things, little things like that. The color of Trello probably has a lot to do with it, although we didn’t necessarily set out to scientifically determine the best.
Andrew: I do see a big difference here. My researcher put up the first HTML mockup of what was called Trellis before it became Trello, right? It’s grey and white.
Michael: Trellis, the idea was that this software gives structure to what you’re doing. So, that’s why the code name was Trellis, but we couldn’t get the domain.
Andrew: I see. And that’s why you changed it to Trello. But I do see now the difference. The color instead of being this rich, beautiful blue, it’s more grey and black and white. So, the color, you’re saying, helped bring people in.
Michael: Yeah.
Andrew: Okay. So, it’s the product. We described a little bit about how the product helped grow your user base. What are a couple of other things that allowed the user base to grow so massively in such a short period of time?
Michael: Well, the app is collaborative by nature. People are using it to solve problems when they are organizing with other people. So, by its nature, just because of the type of app it is, it helps itself spread because you invite more people to use it. That’s what it’s designed for.
Andrew: I see.
Michael: Giving away Trello Gold to people that add members. I think having… oh, here’s another thing. So, one of the challenges with building a horizontal product is what about those people that start to use it more and more and they have a higher need for more complex features? How do you do that without ruining that first user experience and making it so easy for the original people to sign up?
So, we struggle with that too. If you don’t add more detailed features, then people are going to switch to something else when they get to a certain point. One of the ways that we solve that was with power ups. They’re kind of hidden. But there’s a way to turn on certain features on a board. You can turn on like a calendar view. You can turn on voting. You can turn on a card aging. There will be more in the future. But the idea is to put more of that advanced feature stuff behind a curtain so it doesn’t have to confuse people.
One of the people that works for Trello, she was trying to organize holiday presents for her kids, like for her mom to buy presents for her kids and her brother to buy presents for her kids so that the people didn’t buy the same presents. So, she set up this Trello board. It had all these different lists. The first list was gift ideas. The second list was gifts that have been bought. There were a couple different lists.
So, she invited her mom to the Trello board who had never used Trello before. So, she signs in and the mother just starts making cards all over the place and just like typing questions to Liz, the person that works for us. Liz is kind of getting frustrated. She’s like, “She’s using it completely wrong. I had this whole process. I don’t know why she’s putting cards there.”
I said to her, “Yeah, but look, it works. You’re communicating with her. You’re going to answer her question and also tell her, ‘Mom, put the card for the gift in this column.'” It wasn’t that she had to learn that first in order to start using the product to talk to you, it was just flexible. It didn’t work the way you wanted it to, but that’s because there isn’t a set structure to it. So, it allows people that don’t have much experience with your app to just get right in and start using it.
Andrew: I see. Yeah. So, even if they’re not using it right, it still works.
Michael: Exactly. And then you get the opportunity to help them use it right. I think there’s a downside to that, though, too, which is sometimes people just get overwhelmed by that and they’re like, “I don’t know how to use it. Just tell me what the right way is.”
So, we’re working on ways to sort of give examples to people to show them ideas because I think when I talk about that onboarding Trello board where you have the new employee and you clone the board, when I give that example, people are like, “Oh, that’s a great idea. I should do that.” So, we have to figure out ways to help people see how Trello could be used in all their different areas.
Andrew: I always thought that would be helpful. When we start out, instead of being told we can create this generic board, it would be nice if we can say, “I need a board that allows me to collaborate with a developer or a board that allows me to onboard people. Can you give me a template that already does that?”
Michael: The template word, that word is kind of a weird word. I shy away from that a little bit because I think what you’re trying to achieve there is just to… in Trello world, a template, would that be some columns like labeled at the top? There’s not much to it. The important piece is somebody has seen a use case of Trello or certain uses of Trello and going, “Oh, yeah, that’s how I can do it.” And then specific to how they’re using it, they’ll be able to set the lists up the way they want, but there isn’t much in the board itself that would carry over from use case to use case.
Andrew: I see.
Michael: If we had an onboarding template, what would be in it?
Andrew: Just the column names, but then the column names tell me so much about what to do with it.
Michael: Yeah.
Andrew: But you’re saying maybe what I need is use cases, not columns.
Michael: I’ve done 90% of the work by just giving you that example and telling it to you. If you wanted to make your onboarding board now, you could do it in two seconds. That would be relevant to Mixergy, not to some other company like Heroku or something.
Andrew: I think the first time that I used it there were three columns; to-do, doing, done, right?
Michael: So, I think it still might be an A/B test, but we are experimenting to stop doing that because it sort of implied that Trello was a to-do list and it’s not. That’s one way that you can use it. So, it might have stopped the A/B test and now you just get columns. We stopped putting to-do, doing, done as the default board that you get.
Andrew: I see because then I do think of it just as a to-do list, where maybe instead my columns should be ways to improve the video interview, ways to improve the audio interview, ways to improve the transcript. Now we have three columns where we can have suggestions and vote and so on.
Michael: Yeah. Because the first thing that most people were doing was actually renaming those lists.
Andrew: Yeah. I get that. I do that all the time too. But this is a big challenge. For software that wants to be as ubiquitous as a spreadsheet and as flexible as a basic piece of paper, I feel like it’s a big challenge.
I remember talking to Justin Kahn about why they felt like they needed to change Justin.tv into TwitchTV. He basically said that it’s just too broad. We needed to narrow down to one topic. Twitch was just video games. So, people understood how to use us. And then he, of course, went on and created Exec, which was personal assistance for anything. And then I just he narrowed it down to home cleaning because it was too broad. How do you fight that?
Michael: I think part of this is just we’re figuring it out as we go along. We have this goal. That’s what we want to do. I think there was an article in Fast Company recently that was sort of that question, which was, “Trello is trying to be everything. Can they do that?”
It’s not really… I don’t think we’re trying to be everything. I’m certainly not trying to make Trello and applicant tracking system for a recruiter, for example. People can use it that way. Is it the best applicant tracking system? No. But what it is is a really lightweight tool that you can use to get clarity in your life. I think people are using it in many different ways.
So, I don’t see us struggling with adoption. I think if we were in a different place and we were like, “Trello can be used in so many different ways and people don’t know it and they won’t use it in these different ways,” then we may pull back and say, “Okay, let’s focus on a vertical. We’ll get traction there that way our marketing message will be consistent and people can pick it up and then we’ll worry about going to the next thing and then the next thing.” But we’re currently not having that problem.
People are understanding that they can use it for many, many different things. That’s what we want.
Andrew: You feel that that has helped you, the ability to use it for so many different things allows people to take it from the development side of their business to the sales side and then at home?
Michael: Absolutely. People just already get it. They’re using it at work. So, then it’s just easy to use it in other contexts.
Andrew: So, what are some home uses that we should be aware of? I always like use cases.
Michael: They called it a “honey-do” list, things to do around the house, like, “Those projects I’ve got to get done.” It’s usually the case where one person has to do it and the other person is like nagging and really wants you to do it. Let’s say the husband wants the wife to do something or the wife wants the husband to do something. The nice thing about Trello is then when you do it, they can see that it’s done. It’s a way to sort of say, “We’re making progress. I didn’t do that thing that you wanted because I was doing this other thing.”
The kitchen contractor, Joel redid his house and redid his kitchen. He was trying to figure out what they were doing because the date kept slipping and he was like, “What is going on?” like basically trying to get a picture into, “Is this project making progress?” essentially and loosely defined projects.
Andrew: Here’s what I use it for at home. As I said earlier, I have people come over for brunch, sometimes for dinner, sometimes for scotch night. I have a different set of things that I like to do to prep for each of those events and then one thing that’s for all of them, like, “Go check the bathroom, make sure there’s nothing embarrassing.” That I do for brunch, dinner, scotch, whatever.
So, I just have a different card for each one of my events and on it are all the little things that I like to have ready before people come over. So, if it’s brunch, I like to have some champagne and orange juice sitting right on the counter with some glasses so the first thing someone sees when they come in is a glass that they can have something to drink. I also like to have my coffee maker right there so I can offer them coffee if they prefer. Once that’s done, I check that off. If it’s dinner, it’s wine on the counter and music also playing on the stereo.
So, those little things, that’s one of my best use cases for Trello. I do that on my own. My wife does not believe we should be so anal about the way we prep for having people over.
Michael: So, some people are like, “Why do you do all that work?” I think then you don’t have to if you don’t get enjoyment out of getting that structure, then you don’t need something like that.
Another use case where it’s not necessarily a work flow or a checklist or to-do items, Joel has a board that’s all the art in his house because he collects different kinds of art. So, he has cards. On the front of the card is a picture of the artwork and on the back are all the details about it. In that case, it’s more for like just data. It’s not anything he has to do on that board. He’s not really collaborating with anyone. But Trello is already on his phone. It’s on the home screen. So, click, click, he can see all his art.
So, it’s interesting when you have a flexible tool like that to see how people use it in ways that you would never have imagined.
Andrew: I’m looking to see what other creative use I’ve had for it. You know, there’s one that I’ve subscribed to, and you can subscribe to other people’s boards, with suggestions and tools, two different ones, I guess. One is suggestions by topic. Each topic has its own column. Each suggestion has its own card and you can vote up and down. I think you guys have one of those for suggestions for how to improve Trello. The other is for tools. So, you get tools by category. Each category has its own column. Each product within that category has its own card.
You know what? Here’s another thing that I like about you guys, the API. So, there are some things that you don’t do. You can let me, via email, add a card to one column at a time.
Michael: Yeah.
Andrew: But because you have an API, I can go to Zapier and create an email system that allows me to send email directly to the column that I want or if I move something to a column, to send email to someone else and say, “Hey, Andrew just moved it to another column. Get on it.”
Michael: Yeah. So, actually on the back end, this will be more interesting to the developers, but one of the things that we learned from just our previous experience was that we treated our web app, actually, as a client just like the iOS app and the Android app. They all talk to the API. So, all of our apps are actually consumers of the API as well.
Andrew: I see.
Michael: So, anyone can use the API. Those apps, they’re not using a secret version of the API. The nice thing about that was that abstraction made things really easy for us to just sort of like compartmentalize things and iterate quicker because it wasn’t all connected and sort of mangled together.
Andrew: What do you mean? What’s an example of how, because your API and the product are separated, because your product is built on the API, that you were able to evolve faster?
Michael: Then the teams can be smaller. There can be a server team that’s sort of responsible for the API. So, when we do something like localization that’s going to have effects all over the place, we start there. We can say, “Okay, this is the common interface across all of our apps. What kind of things are we going to need to support or change in order to do that?”
It also means that performance is much easier to handle because everyone is using the API. So, the way to make sure that your performance is up to speed is just to measure how the API is performing and make sure that if there’s some way to optimize it, then do it and then everyone gets the benefit of that.
I think having your apps be clients of the API that you also publish for everyone else to develop on, it just gives a certain nice structure. I think the developers really like it. It just makes it cleaner.
Andrew: So, your first interview is seconds, maybe a minute or so away from being done. How does it feel?
Michael: It was easy. It was easy.
Andrew: Yeah. You know what? You were an easy guest to have on.
Michael: You’re an easy guy to talk to.
Andrew: You know, I didn’t ask you to do a pre-interview. Usually I would. But I peppered Stella with a lot of questions in preparation and I tried having a researcher put information together as an experiment to see, “Can we get the same level of information from researching and talking to Stella as we would from a pre-interview?” It was a really interesting example for me. It was a really interesting experiment, I should say, for me.
Michael: What was the result? Did it work?
Andrew: You know what? It worked in some levels but it didn’t in others. So, I felt that we could have prepped you a little bit better about what it was about the product that allowed it to go viral, that allowed it to grow so much so that you wouldn’t be put on the spot here trying to figure out how to explain to us what years and years of product development did that we could understand, then, to ourselves.
Michael: I know that’s what you’re trying to do, to distill that down for people and to figure it out. It’s so complicated. There are so many different things that go into that. It’s hard for even us to be self-aware of why Trello was successful and sort of Copilot didn’t turn in to LogMeIn. Why is Fog Creek the size it is and Atlassian, that makes the same products, the size it is? There are so many complicated things that go into answering that question that I don’t even know if we could solve it. But I hope that we got some wisdom out of this.
Andrew: I know we did because I’m a great interviewer and you are an incredible CEO. I know you’ve had the track record for years. I’m curious. Do you spend time internally or maybe outside of the office with a bottle of wine saying to yourselves, “Why did this work? What do we need to do more of because that’s what our strengths are?”
Michael: We definitely ask ourselves that every day because one of our biggest fears, this is the same in Stack Overflow too, like find a developer today that hasn’t heard for Stack Overflow. Huge success. But the big fear is what if everyone starts using something else. If we can’t put our finger on why people love Trello exactly, we might change something that makes people not like it. Like if it is the shade of blue that Trello is and then we change the color, we could get people be like, “I don’t know. I just don’t like this app.”
Andrew: That is a scary thing. I remember even Paul Graham and Jessica Livingston saying that whatever it is that Paul Graham had that allowed people to develop products, because they didn’t know what it was, it could have just disappeared. At least Paul Graham did. Jessica laughed it off. But over the years, you can see that they’ve refined their understanding of what it is that makes the startups they work with successful and what causes them to fail and that allows them to do more of what works. You guys are doing some of that.
Michael: Yeah. We’re definitely doing some of that and we’re getting better at it.
Andrew: Well, if you guys ever want me to come over there, bring a bottle of wine, actually, it will be scotch, and have a conversation like this where I drill you and pull out all this information from you, I’d be happy to come down.
Michael: I’d love that. That would be great.
Andrew: All right. Cool. I’ve admired your work for years. I’ve admired Joel’s, not just his articles, but frankly here’s a cool thing about Joel. Years ago, somebody wrote something nasty on my site and attached his name to it. I emailed him over and I said, “I’m sorry, what was it about what I said?” He could have totally just blown that off. He responded immediately back and explained it. I thought, “This is such a nice touch from some guy who must be incredibly busy.” Everyone in your organization has been like that.
Michael: Well, I thought one of the things you could have asked me was how do you find a good co-founder.
Andrew: How do you find a good co-founder?
Michael: I got so lucky. I found an amazing friend and a business partner and somebody that’s very just super smart and super honest. He’s just basically like a lot of the culture around Fog Creek and how we have profit sharing there and the way that we treat employees and the benefits and stuff like that, it all came from Joel’s sense of fairness, even our compensation system and those sorts of things.
So, he’s been an amazing partner. I got really lucky. I just got lucky because it was just serendipity that we were at the same company and we hit it off on a personal level. And then we just said, “Hey, the market is kind of tanking.” It was 2000. “Why don’t we go do our own thing?” And we did. The first four years, it was like us in an apartment.
Andrew: His grandmother’s apartment.
Michael: Yeah. Looking back now, it’s ridiculous that it turned into this. These companies collectively have over 300 people working there. But yeah, it’s been a fun ride. I’m really excited for the future.
Andrew: Well, congratulations on it. I can’t end this without saying thank you to Stella for introducing us. Stella Garber is amazing.
Michael: I had my eye on her for a long time. I had worked with her or I had seen her and talked to her and seen what she had done at a previous company. I knew when I started Trello I had to get her.
Andrew: For PR, you mean, you saw what she did there?
Michael: Yeah. I remember poking at that over the years. There was never like a right position for her. But when it came up… she had a startup at the time that she was working on. So, it took a little bit of convincing. But I’m really excited to be working with her.
Andrew: I’ve always been amazed by the way she promotes the companies she works with. It’s a human touch that also understands my position. I wanted to have you on, but how do I have you on? How do I tell her I’m not sure what to do with this? Well, she totally understood and she said, “Look, here’s the angle that we could take and here’s some research that will help make this into a great interview,” which I now have on my screen here.
For anyone who hasn’t seen her, her name used to be before she got married Stella Fayman. She did one of the first Mixergy courses where she taught how she gets publicity for the companies she works with. You’ve got to go and watch it. She’s fantastic. If you ever get to go see her at a conference, go see her.
Michael: Yeah.
Andrew: Congratulations on hiring her. Congratulations on hitting six million. And more important than all of that, congratulations on stepping up and doing more public interviews. I’d like to see more of you and I’m proud that this was the first place.
Michael: Okay. Great. Thanks, Andrew.
Andrew: Thank you. Thank you all for being a part of it. Bye everyone.
