Being upset with the status quo.
I think this is a common trait across entrepreneurial people. You see some part of the world that can be better, and feel severe agitation to change it. You find partners that share your agitation, and go to work.
Heroku was motivated by being sick to death of dealing with servers, operating systems, systems administration, ssh, config files, and the rest of it. I want to write my app, not futz with cranky servers.
Orion Henry, James Lindenbaum, and I were running a consulting shop named Bitscribe, in Los Angeles. Every couple of months we would finish a new customer project that we needed to deploy. Every time we had to buy servers and do all the route work of setting up an app. And every time we asked, "Hasn't someone solved this problem yet?" Eventually we decided to solve it ourselves.
Orion and I wrote a MUD (a type of text-based multiplayer online game) together when we were attending UCSD. Doing this, we learned three things: we loved building services running on the internet, we loved the unix philosophy, and that he and I made a dynamite team.
When Orion was working on starting his first company, he called me up and talked me into joining. I quit my job in the game industry to go join him on building what would become the payment gateway TrustCommerce.
We've started many businesses together since. One of those was Bitscribe, where we worked with a bunch of the people that would become early employees at Heroku. James joined us at Bitscribe, and when we saw his history as a serial entrepreneur we knew we had found a kindred spirit.
Having the right group of founders is the most important thing there is. Financial success is just a trailing indicator of the strength of the working chemistry with your partners. I wouldn't trade my lifelong partnership with James and Orion for anything.
Almost too numerous to count. Although it wasn't our first business, it was our first time taking outside investment, our first time scaling infrastructure to this size, our first time growing a team past twenty employees.
But the hardest thing is always focus. The best ventures are the ones started in very fertile territory: tons of prospective customers in pain, tons of valuable problems to solve. But you can't solve all of them at once, and if you try, you'll do all of them badly and fail. Heroku has definitely been one of these. There always seem to be a hundred interesting problems to solve for every one we have the resources for.
There's only one solution: do less. Cut scope. Build fewer features. Be hyper-targeted on a very specific segment of potential customers. If you're doing it right, you'll be saying "no" far more often than you say "yes."
We've learned this lesson over and over, but it's so easy to forget. Entrepreneurs are problem-solvers, and it takes ongoing, intensive discipline to select and focus on only the most valuable and achievable problem in front of you, and ignore all the rest.
I'm a product-oriented guy, so I'll pick a product milestone: when we shipped the Celadon Cedar stack, in summer 2011. This was a top-to-bottom reworking of the product – you could even call it a pivot – that vastly increased the power and flexibility of the Heroku platform.
Users loved Cedar in a way that even I wasn't expecting. We experienced 10x growth on almost every usage metric in the six months following Cedar's launch. For me personally, Cedar felt like a culmination, the product I had wanted to build from the beginning.
We had started a number of other bootstrapped businesses, but with Heroku we knew wanted to go really big. I think of taking investment dollars as less about the working capital it gives you, and more about assembling allies. Starting a company is really, really hard. Not just startups, but even a small business like a restaurant or a web design shop. Take every ally you can get: cofounders, incubator programs, advisors, investors.
I do believe in bootstrapping, though. Historically it seems to have been an all-or-nothing thing: get on the train with angel investment, then VC rounds, then eventual IPO or acquisition. Or, fund yourself, and labor in obscurity with limited resources for far longer than it would take otherwise.
Lately I've seen more hybrid models: people raising angel rounds and using that to get to profitability. Or getting to break-even via bootstrapping, then choosing to take some funding which can be put directly toward scaling out or going into an adjacent market. I like the hybrid approach, I hope it becomes more common.
Have a vision. Don't pander. Seek a big market opportunity but display hyper-focus in the first iteration of your product. Make sure you have the right team.
We've stayed very independent and autonomous. So not much at all. Big props to Marc, George, and Parker for giving us this latitude.
Stay focused: cut scope, ephemeralize.
I believe that the platform-as-a-service market is still in its infancy. Heroku is the market leader right now, but the game is still anyone's to take.
I see a parallel between the state of PaaS in 2012 to the state of LAMP in 1998. There's a gigantic opportunity ahead of us, and much more innovation to be done – in the product, in the technology, and in the business model – than lies behind us.
Most people don't realize it yet, but all software that runs on the internet will run on PaaS by the end of the decade. Whoever dominates that market will define software as we know it. If Heroku has its way, this future will be all about open standards, the unix philosophy, and a joyful developer experience.
I'm pretty addicted to being able to have such a huge impact on the evolution of software development. It's hard to imagine doing anything but this in the coming decade.