What are some of your earliest entrepreneurial stories?
I started my career as a writer a long time ago, almost before the Web. And before I even got a proper writer’s job, I did freelance writing, which is always pitching and selling. I would pitch ideas for stories because that meant I could get paid and buy beer.
Was anyone in your family entrepreneurial?
Nobody. I was really the first one who had a profound difficulty holding down a job. I was a terrible employee. As a pizza delivery guy and all of those other jobs that you have in high school, I was always anxious. Not nervous anxious but like, “Come on what’s next, what’s next?” I guess the combination of a short attention span and a curiosity about things made it hard to work for somebody else.
As a pizza delivery guy and all of those other jobs that you have in high school, I was always anxious. Not nervous anxious but like, “Come on what’s next, what’s next?”
What type of student were you?
I barely finished college. I got all A’s and F’s. Because I was totally into whatever I was into, and I couldn’t stomach anything that seemed pointless, which maybe leads to a myopic education. But it also taught me that if you’re really into something, go deep. I guess I call that entrepreneurial–getting obsessed with something over and over again and trying to find a way to make that work.
But it also taught me that if you’re really into something, go deep. I guess I call that entrepreneurial–getting obsessed with something over and over again and trying to find a way to make that work.
How did you get into design?
In college when I started really using Macs, I was fascinated with desktop layout programs like PageMaker and QuarkXPress. It was so cool how what you saw on the screen looked just like it would when it came out on paper. I tried hard to make my term papers beautiful. I used QuarkXPress to make a desktop-published book. It was a FAQ about the video machinery students had access to at my school. I remember that project sitting on a floppy and I needed the postscript version and what’s ATM? I figured all that stuff out, and loved it.
So when I finally got a job at a newspaper I wrote maybe a month’s worth of stories and then went over to editing. I liked the big screen and laying out the paper. I loved the inseparable connection between the words and the expression of those words on the page. When I give talks publicly, I work out what I’m going to say in Keynote instead of making an outline or a list of talking points. I like to work out what the story might be while I’m doing the slides. I see the two as inseparable.
That’s how I stumbled into design. I had no idea it was a thing and that people went to school to study it. But luckily I got started on the Web early enough where none of that really mattered. To get started, being able to do HTML and having a copy of Photoshop was enough.
How did you transition from newspaper layout to interface design and Web startups?
One pivotal moment was seeing Wired Magazine for the first time. I was 24 or 25 and at a party in Santa Barbara, California, and I looked down on the coffee table and there’s this fluorescent color magazine with Bruce Sterling on the cover. It was the first issue of Wired. I picked it up and literally read it from cover to cover. Then I put it down and left the party.
Wired paired the geeky Web stuff I was doing with design and page layout stuff. It was like an, “Ah, angels from heaven” moment. So I applied for a job by writing to them at [email protected] I said I’m a huge fan, I would love one of your open positions, here’s my resume, which was a pasted URL. I was the first one to do that at Wired. Surprisingly, there were no URLs in the first issues. I got the job, and then they found out I knew HTML.
How did you know HTML already?
Geekiness. I just loved messing around online. I had an account on the school UNIX system, MichNet, in ’89. I played with MUD and FTP. There were a lot of people swapping software, and I loved collecting fonts. Maybe I shouldn’t say that, but I guess it’s fine because that’s how you can get into things when you don’t have the financial means. You can call it pirating or you can call it learning to be a customer later.
So at Wired, I quickly got to know the creative director, Barbara Kuhr, who did the Web stuff. She said, “You know Web technology and QuarkXPress? You’re in! Put your desk in my office.” I became her design assistant, even though I wasn’t a designer. She said, “There’s probably something between being pretty technical like you are and understanding design.” And that became a sort of career grid for me. So for five years I apprenticed under her and then had my own team. We called ourselves the interface team, which was different from the design group. We built the front-end for HotWired, an online version of the magazine but with separate, new content and its own advertising.
Barbara mentored me in design, visual hierarchy, color theory, and whatever you can think of. There was actually a tremendous amount of color theory that was happening in Wired with all the inks and the dissonance in the color. So I started out self-taught but that’s how I learned properly.
Besides design theory and so on, what other lessons did you take away from Wired?
There was a very good thing that happened to me in that dot com insanity. Condé Nast bought Wired magazine but they weren’t interested in HotWired, which at the time was called Wired Digital. They thought, “The Web, bah, flash in the pan. We just want to get the hip magazine audience that tech advertisers will spend money on.”
So, Wired Digital, was sold to Lycos and I was there for about a year and saw just how profoundly bad it can be at a big company. I thought my career trajectory would be through publishing, and before I saw Wired, I had my sights on The New York Times. Then I worked at a big company and it sucked. In the first few years of a company, people are passionate. They throw themselves into their work and there isn’t all this bullshit that goes on in a big company. So I knew I had to get out of there but didn’t know what I would go do.
I had only tried publishing and a technology content publishing website. I’d never done e-commerce. This was around 2000, when the dot com bust happened. So there are interactive Web apps and Hotmail. I realized, UIs aren’t just in desktop software, we’re going to do it in Web pages.
And I ran into some friends who also had just either quit when their companies got acquired or got laid off when the company folded. We wanted to work together, and be smart. We all had rough experiences in our previous companies, so we decided to do consulting together. So Adaptive Path came about from our strong desire to work in all kinds of different areas.
What we sold to businesses was how Web design could be of financial benefit to them. These big companies had been funneling money into the Web for six years now, and it all busted out. We said, you’ve got to stay on the Web, obviously, but you’ve got to be a lot smarter about it. There’s ROI for good design.
How long did you stay in consulting?
I was at Adaptive Path for about five years before I realized I didn’t want to solve other people’s problems. I wanted to own the problem. In consulting, the diversity is fun at first. And the money is instantaneous—you start the project and then get paid. None of this, “Let’s build the product and we’ll find a business model eventually.” But on the flip side, you only get paid when you’re working. While your product can grow exponentially, your time cannot. And the diversity actually gets exhausting after awhile.
What did you learn from your experience at Adaptive Path?
As a consultant, one of the most valuable skills that I had was telling the story of what we were trying to accomplish—communication skills. I think every successful product has a narrative running through it. What are we doing here? Some people call that the vision. And so I put a lot of effort into making sure everybody knows what the vision is, hears the narrative a lot, agrees on it, and that it’s expressed in everything that we do.
Every successful product has a narrative running through it. What are we doing here? Some people call that the vision.
We also realized the importance of culture and of trust and spent a lot of time building those. If everyone on your team is smart and you trust them, then if they disagree with something, it matters. It’s not top down political, like it can be at a big company. If you’re working as a team, then dissent is vital to continuing to refine the vision and narrative over time.
When we started Typekit, the team was very clear about the vision. It wasn’t just an opportunity to bring fonts to the Web. There was another layer on top of that—advocating for Web standards. And not as a consultant, as in I’m a designer and I’m going to pitch standards for the clients. Our vision was you can actually build a service that helps people implement Web standards and charge for it. Typekit is a product and service that takes the values that I’ve been cultivating my whole career.
So rewind a bit. How did you start Typekit?
Part of why I stopped consulting was I wanted to work on a big project of our own. And so I convinced my partners at Adaptive Path that our consultancy should build a product, which became Measure Map [a Web analytics tool]. We were acquired by Google to stop Measure Map and to redesign from the ground up a very new product that they had called Analytics. The UI that people use today mostly came from our team. We also worked on some apps like Gmail, Calendar, and Docs with a bigger team.
I left Google a couple years later with two core guys from the team, Ryan Carver and my brother Greg, and then over time we hired some more people from that team as well. It felt like a rock band reuniting, like let’s do another album but on a new label. We had no idea what we were going to do. For six months we experimented.
So one of my projects at Adaptive Path was to redesign Blogger post acquisition to Google. And all that wraps up to the story of Typekit. It looks like we were really lucky, but man did we work hard to put ourselves in that position.
How many years?
Over a decade and a half. I did work on the W3C working group for Web fonts in CSS in 1996. So Christmas 2008, I was in Australia, where my wife is from, and had lunch with John Allssop, who is the Australian Web standards guy. He told me that the Opera browser had just implemented that Web font work that I had done in ’96 and that it was in the WebKit nightlies and Firefox was on board. I’m like, “No shit.” He’s like, “Yeah, it’s Web fonts. You should turn back and do a business.”
So I said to the guys, I think this is the idea. It takes the design and standards work I always have wanted to do and applies it to a product at Google scale, because we had learned about serving billions and trillions of things. There’s a business model right here. It’s designed as a service, which is going to be a thing. It is about a service and not assets. Look at Rhapsody, and iTunes, and now there is Spotify. Instead of people owning assets, it is about a service that gives access to a whole library. Let’s do that with fonts because the standards are here. And so that was the generation of it all. We put together a pitch and boy did it resonate with people.
Talk to me a little bit about that fundraising process. Did your track record help?
Oh my God does it make a difference, but only in your first round and I should be super clear about that. The very first round of money that anybody raises is a bet on the team. You bet on the idea that the team has, but also on the team itself because you’re assuming that they can fail at that idea and do another one.
But the second or third round is all about growing the idea because you have traction. So having some reputation from the past means, let’s get on with the meetings. Having a pretty good idea like this meant we could land it around on reasonable terms. But nobody knew the size or potential of the idea. Honestly, one VC said, “Wow, so people pay for the shapes of letters.” That’s a very VC way of thinking about it. But the fact that there is a market and one that could be of scale was interesting to them. Moreover, there’s a large audience of people willing to pay for services to make their online presence look better. Every single website out there has paid for a domain name and has put some HTML together to make it look a certain way. That’s the size of the potential market.
And pretty early on we laid it out that our initial audience will be the Web designers and developers that write code by hand, are focused on standards, and want to do things the right way, who are totally eager for this service. The ancillary markets, where it really gets big, are the millions, and millions, and millions of websites that buy a Tumblr or WordPress theme, or hire somebody to make a website for them and spend $100 bucks on it or $500 bucks on it. We think that market is enormous, and that’s why we partnered with WordPress, Squarespace and other places.
So how did your vision of this large potential market affect product decisions early on?
A lot of the geekier part of our audience pushed back on the fact that we didn’t just give a link to the font to put in your CSS. Instead we linked to some Java script that would then make the font available in your font stack. The reason we did that was because from the very beginning, we saw fonts as just the first thing. You could imagine whole other areas, everything from image surfing, to icons, to color theme, to all of the design. If you host it centrally, you could keep future compatibility with all products and devices that come out. That was the whole point, future compatibility, as well as backwards compatibility.
So how big is the font part of it? It is really hard to say because it’s siloed. Fonts are everywhere. There are fonts in the seat back of a Virgin entertainment system, somebody paid for those. But the part that you can capture can get so much bigger. What is the market for the appearance of websites? So that was the big idea, to grow into a marketplace.
Right now when you log in to Typekit and you see all those fonts you’re actually browsing a shop. Even though you have access to everything, you feel these are all the wares that are available and you choose the fonts and take it to your your website and they’re going to be compensated for your choice.
So what is your vision for Typekit now?
I’ve always felt really strongly about not giving away fonts on the Web. And that you could build a business out of Web standards. But we also believed that you could sell products that were traditionally assets as a service, as a subscription. The big wave of objection from people when we announced Typekit was, “I’ll never rent my fonts.” I did not believe that to be true. Because if I could create a service that you wanted to use, you would never think about buying a font again. It’s not about switching to a new thing, it’s about, “I need that service.”
The big wave of objection from people when we announced Typekit was, “I’ll never rent my fonts.” I did not believe that to be true. Because if I could create a service that you wanted to use, you would never think about buying a font again. It’s not about switching to a new thing, it’s about, “I need that service.”
I believe that’s the case for everything. I don’t even go to iTunes anymore. I’ve accessed every song ever created with Rdio or Spotify. I think a lot less about my library, and a lot more about what I want right now. That’s so different from when I was buying CDs a decade ago and when I was buying tracks four years ago. The same is true for movies. Buying a DVD? No, I have Netflix. It’s a service where new stuff is available and I can watch it when I’m ready to.
I think all of our software is headed in that direction now. When we would look at the Typekit AMEX bill every month, there was no software on there. Maybe we bought licenses to TextMate, but we were spending a couple grand, at a minimum, on services. It’s our bandwidth bill and Amazon, but also everything else. Monitoring, GitHub, 37signals, all these services enable us to do our jobs more effectively.
I think that’s the vision for creative work, for development and design, for how we produce everything. So I won’t speculate on the future business model for Photoshop, but I would say it will include a service component to it. And that’s what we’re building here with Adobe Creative Cloud, that’s the whole premise. The master collection of the Creative Suite used to cost $2,500 every two years. Now it costs $49 a month. There’s plenty of people who say, “Yeah, but I’m not going to rent my software.” But what I say is yeah, but you’ll pay for the service. There’s value in not having to wait two years to get new versions of the software. When they’re ready you can download them online—just like app updates on your smartphone.
How did you as a designer contribute to the product development over the course of the first year or so?
Balancing the time, quality, money factor. I am really bad at process. Jeez, I struggle to even balance a checkbook. So I can’t stick to extreme programming or Agile development, but the point behind those methods is around momentum.
And that’s what I really wanted to have with Typekit because I knew we were totally on to something. So much so that four or five other companies were working on it. We believed that the company that would win was the one that was first to market. Because the decision to use the product was a decision to redesign your website. And how often do people do that? If they redesign their website, when will they redesign it again using different Web font provider? Years, you know, certainly quarters of time. We’ve got to be first. Right now.
And then the entrepreneurial aspect—we’re running out of money. We had our End of Revenue slide once a week in our little weekly review, and it was a date. And we’d put it in front of the team and say we’ve gotta ship, we’ve got to get to market right now. End of revenue is October 14 or whatever it was. And what was great was once we launched, we would see that date move further into the future, because we were charging. Every time we got enough revenue to move to a new date, it was exciting. So it was a good metric. It wasn’t all pure anxiety. It gave us some momentum. I think momentum is the motivating force, and that’s critical to the culture.
And how did you help balance momentum with the other factors, quality and money?
Quality, I would not compromise on—the product has to be a great experience. And then money, well, if we’ve already raised money, it’s a fixed resource. So instead of money being the third factor that we could compromise on, we called it scope. If you’re not going to compromise on quality and time is fixed, because we’re going to launch it as soon as we possibly can, the only thing you cut is scope. And that is one of the pillars of the lean startup methodology, which is do the least amount you possibly can to get something into the market so you can start to learn. Though my motivation was more than just learning, it was being the first into the market, and getting the momentum, and headway, and the head start.
So you’ve got to go fast, it’s got to be high quality, and then there’s reducing scope. And we were radical in cutting scope and maintaining quality. We were going to do the best possible thing but we did the least amount we could. There are things that are fixed and everything else we had to shed, and shed, and shed.
The example I give is that when we launched Typekit we charged customers from day one even though we didn’t have a search engine. We created some good partnerships with foundries, so that we had about 800 vaults in the library, but no engine to actually search them. You had to page through a file that you couldn’t sort. And we didn’t do alphabetical order because honestly fonts vary in quality and usefulness. If all your pixilated retro grungy fonts start with A, you’re library will look really bad. So we put them in order of how much we liked the font. You couldn’t search or sort it—all you could do was hit the next button.
We had almost no functionality in the core app, but what we did have there resonated with people so they would immediately say, this is awesome, but you know what I really need? That phrase, “What I really need,” is vital. That phrase sets up all your priorities. And now we are in market. We’re getting the head start over potential competitors and we’re continuously learning from our users.
We would respond with features as minimally as we could and that created this loop of us being incredibly responsive and transparent with our development process. Every four to five days for the first two years, Typekit added user-facing features.
What’s an example of responding minimally with a feature?
So the first search engine was actually a text box with auto complete. You couldn’t submit the search because we didn’t have a search engine, but we could very quickly load up an array of all files that matched when you type HEL, there you go, there’s Helvetica. That will work. That was minimum viable search.
And that got us a long way. We didn’t have a search engine for a long time after that. It turns out if you’re going to search for a font by name, we either have it, or we don’t. There’s no fuzzy logic around that. And that turns in to looking and actually being super responsive and super transparent. Here’s how everything works.
Are there other things you did internally to cultivate a company culture that respects, embraces and fosters design?
Lifting right out of the extreme Programming or Agile methodology is the morning stand-up. We had one every morning, though not as regimented in format as other people do. It really scripted the entire company to this day. We had 23 people at the time we sold the company, and all of them would either get on Skype or get in the same room together at Typekit and to go over everything that we were currently working on and that we were going to accomplish that day. We blocked it up into active development projects, so they were the only ones that got to speak, one project at a time. If anyone had a question you would follow-up online or after the meeting. If the meeting was longer than twenty minutes something was broken.
We would save 50 emails a day by just doing that. That, plus [37signals] Campfire group chat rooms and you hardly needed to worry about an inbox, if you do those two things. And that also led to momentum. The day starts at 10:05 every morning. Do five minutes after because everyone sets their alarm for 10:00 and it works out that way. Everybody starts their day at 10:00 Pacific Time, that’s the beginning
Another regular meeting was once or twice a week we’d do a product review. It was entirely open and anybody could come, the operations engineer guy or our office manager would come to it. We would look at some of the product problems we were trying to solve, super micro into the details. They were collaborative and had a lot of ground rules to them, but also, everybody is now on the same page.
For those meetings attendance was optional, but participation was mandatory. We didn’t have people sitting around with their laptops. If you were there, you were there, and I wanted to know. I stole that from an interview I heard with Obama who said, “I make sure I call on everybody at some point because everybody’s thinking and not everybody says or feels comfortable saying what they think.” So we would talk and we’d all try to think about it for a minute and I’d say to someone, “Hey, what do you think?” Those were great.
And then late on Friday afternoons, we would have the weekend review. This is before we cracked open the keg and when everybody’s brain is kind of fried anyway. We stole the idea from Evan Williams of Twitter, who did Tea Time, which he stole from Adaptive Path, which started after I left for Google.
The whole company would get together and go over what we did that week. We actually would do a Keynote presentation and make those slides nice. It created pride in people’s work and how awesome our company is. Look what we’ve made! And we put some effort into it. I would work on that from lunchtime on Friday on. “That thing you were talking about earlier, just make me a couple of slides. Here’s the template.” And we still do it today, The Week we call it. Everyone competes with one another to make their slides better. And it is wonderful. I also have one or two slides of the weeks’ metrics. So we know the things we measure ourselves on and how we are doing. It’s fun and relaxed, and it’s intended to be that way. It creates a wonderful culture. That’s a thing that I love. So I instituted those three things: the morning meeting, the once or twice a week product review, and then the weekly review.
What were some other lessons learned from Typekit?
I think we hired too slowly. We were extremely cautious and conservative when it came to adding people to the team, and that was experience, but I was wrong about my interpretation of my experience. And this might sound a little mercenary, but instead of being so conservative with hiring, I wonder if I should have been a little more comfortable with firing, with making a change. Instead I avoided that situation. Having to let people go, in the past in various managing roles, was the worst thing ever.
And when you make hiring decisions, I believe competence and culture have a baseline and then culture just trumps completely. I fundamentally still believe that. You have to invest in culture, you have to seek out the kinds of people you want in your company. I would say no to an incredibly talented engineer if I knew that that person would be toxic to the team. But I think that fear that someone would turn out to be toxic led me to be too cautious. And so we should have hired faster and should have taken chances on people.
Making these hiring decisions, you can get yourself into a space of your head, like what is it going to mean? How are we going to do this? So first evaluate them technically, or whatever the role is, look at their portfolio if it’s a designer, and see that they meet the criteria of hiring. But then, what’s your gut? Can we work with this person? If yes, let’s do it. Get in there and make it happen. Also, let’s all talk as a team first. Is this the wrong choice? Six months from now, let’s talk again and see if this is not working. So have the same kind of process available for us if we have to make a change. That’s my hypothesis now.
What was the effect of hiring too conservatively?
If we had been faster at hiring with the same amount of investment that we made in culture, I think we could have moved even faster and still have maintained the best team I’ve ever worked with in my life. Each team I’ve worked with has been the best, but this one is stunning. I love these people. We’ve only lost three people since we started Typekit. And all of them have gone and started a new company, to which I say, go with good graces. If you were going to Microsoft or a big company like that, I would lay down in front of you, no way. But this is sending your kid to college. So I’m sad, it breaks my heart, you’re amazing, but go! Do the next wave of innovation somewhere on your own.
So if you hire more people, you can do more, faster. But I know I still have three years with Typekit in my head, and we can do it all right now. I know what it looks like.
What advice do you have for people who want to found a company?
Pick the idea that you can’t stop thinking about. It comes down to what you really want to do. I have all kinds of ideas that I think would be great, but do I want to spend all day everyday doing that? Making tools for Web designers and developers, absolutely; that’s what I’m doing with Adobe and I’m super happy. So it’s partially that.
Pick the idea that you can’t stop thinking about… I have all kinds of ideas that I think would be great, but do I want to spend all day everyday doing that?
People usually say you have to network, but I think it’s more that. You have to build relationships. I can’t work the room at an industry party. I hate it. Inherently, it feels fake and I pretend to be super interested in what you’re saying, and I want to connect with you on LinkedIn. But you can create your own luck by building relationships. There are people that I really want to work with and somehow, somewhere down the line that will probably pay off. But it’s not “networking,” because that can’t be the reason why I’m cultivating the relationship.
People usually say you have to network, but I think it’s more that you have to build relationships.
Last question. How do you define meaningful impact or positive social impact?
I can think of two definitions. One is providing tools for the people who create the Web in however it manifests itself: mobile, websites, content, apps, or any of that. Because they are the most powerful people in the world right now. They are creating the future. And if I can create tools to make them more productive or to inspire better work from them, then that raises the tide on what everybody’s doing. I have a background in design, but I want to give both designers and developers tools to make them do better. And that’s whether I build Typekit or work with Adobe or give a talk to a bunch of students. All of that rolls up into making them more powerful.
But I think there’s a bigger picture too, around why the Web has been the thing that motivates to me. Very, very big picture, I think the way to be successful as humans is to pass knowledge along so that I can learn something and share it with you. That’s communication—you don’t have to be there or know me to benefit from what I’ve learned. Somebody figured out if you cooked the water buffalo instead of eating it raw not as many people die. And nobody else knew that until we wrote it down. Prosperous civilizations were literate. The more that we can share our knowledge about the world, the more successful we will be. That could be more empowerment for everybody. And that’s why I think the Web has done it. It’s a network, and it’s open content.
The more that we can share our knowledge about the world, the more successful we will be. That could be more empowerment for everybody. And that’s why I think the Web has done it. It’s a network, and it’s open content.
Before Typekit, we were making content wrong. We were sticking content in an image just to have some control over typography. That makes it harder for it to be translated to other languages on the fly. It makes it harder for people who can’t see the screen to read it. We’ve got to stop doing that and just do it right.
We need to make every idea we’ve ever had available at the lowest common technological denominator. And that’s why Web standards are important. That allows us to communicate. We’ve collectively created a lingua franca for the first time in human history. So if I can at least take out some of the hacks that cause us to do content wrong by putting typography on top of it, in a standards-based way, then it makes sense. It can remain text and be infinitely shareable, reusable, translatable, and all of that. Then we’re going to make communication better for humanity.
Typekit is the tiniest little nudge in the right direction, but it’s also not a nudge in the wrong direction. And if you do that on a broad enough scale, a tiny nudge has unbelievable exponential benefit overall.