This piece on startup lessons was originally written by Qumulo Co-Founder and CTO Peter Godman and featured on TechCrunch. Here is a link to the original byline.
I was born an engineer, and an introvert. The rest is pretty predictable.
When I was three I obsessed over jigsaw puzzles. At five, Lego. I showed whatever I built to my Mom, and she gave me a stream of deliciously positive, non-objective feedback. She didn't need completed puzzles or Lego models. She would have loved anything I built.
By eleven, I was writing video games on my Amstrad CPC-464. At fifteen my BFF Eddie and I, an English kid living in Kansas and a Japanophile -- (and pretty certainly the biggest nerds in our Junior High) -- were spending much of our time building and playing games. We only had two customers -- ourselves -- and we iterated according to our collective wishes. It made the games better, and it it made the experience of building them more fun.
Fifteen years later I joined a data-storage company named Isilon. I ran the performance team, and got the job of turning the one of the world's slowest storage systems into something that our potential customers didn't laugh at. Once I got on the phone with a customer and said “can you tell me about the problem you’re having,” to which the customer responded “let’s get this straight… this is a problem you’re having.”
Anyhow, Isilon had 12-month release cycles, a huge pool of diffuse potential customers with varying needs, and a team of fifty engineers working on tiny slices of functionality. This should have been the least-satisfying job ever.
But my customers were fantastic. We measured the performance of the system with five or so metrics, and at some point we decided those were our five customers. They gave us feedback on every code change in 60 seconds! The feedback was merciless, objective, impersonal, and -- lucky for us -- directly correlated with the success of our company. These metrics were perfect customers for introverts. We put on our headphones, got in our groove, and jammed.
So, maybe it's no surprise, given this life experience, that when I started my first startup, Big Crunch, in 2008, I didn't know how to figure out what to build. There wasn't yet a Lean Startup, nor a Startup Owner's Manual. Four Steps to the Epiphany had been written in 2005, but had I known about it, I would have been too stubborn to read it. At this point, my idea of how great products got built was based on the Hollywoodish mystique of the visionary founder, a vicious narcissistic fantasy which goes something like this:
For me, introvert by nature, this mystique was convenient, because it gave me a reason to avoid talking to people I didn't know.
I think I summoned the courage to talk to about ten people over the period of three or four months, looking for confirmation that what I wanted to build, a scale-out hypervisor, was needed. Having amassed a modicum of evidence, I started talking to investors, to fairly tepid response (maybe even for 2008), and, having drifted for a while, I moved on to something different.
That different something -- Corensic -- was already funded, but lacked definition of what it was. That job fell to me. We had two constraints: 1) build the best thing we could for $1.5m and 2) use the technology that the company was founded around. Customers didn't make an appearance in this set of constraints, and ended up being constrained out of the picture altogether!
People in unhappy marriages make great actors. Startups are marriages. Corensic's outwardly-happy and internally-dysfunctional startup experience lasted until Tom Button, our helpful and wise advisor, sat us down and told us to read "Four Steps." Reading, I had the predictable "Oh Sh*t" moment as I read all of the predictable and excruciating consequences of never leaving our building: disinterested customers, lack of repeat buying, and a murky path from where we were to "big." By the time we figured out how fucked we were, we were beyond redemption. End of startup.
I often tell people about what we've done differently at Qumulo: conducting thousands of interviews across six hundred users of data storage in our first two years; holding user groups 18 months before we had a product; taking our engineers into our customers' environments, and bringing customers into ours; relentlessly trying to surface stories and spread them through our company.
It took me three years and a ton of money to learn that it’s better to be a listening founder than a visionary founder … and that:
The way to build stuff people care about is the way I did it with Eddie when I was fifteen: work with small teams and get your customer one minute away from you. Listen, test, and have fun!
Peter Godman uses his expertise in distributed file systems and high-performance distributed systems to guide product development and management at Qumulo. He is the inventor of more than 20 granted patents in the areas of file system design, distributed systems, and shared-memory concurrency. As Qumulo’s founding CEO, Peter led the company through fundraising rounds totaling $100M, the delivery of Qumulo Core, and the acquisition of Qumulo’s first hundred customers. Prior to Qumulo, Peter served as CEO of Corensic, which brought the world’s first thin hypervisor to market. Before that, he was Director of Software Engineering at Isilon, where he led its file system team through several major releases.
We are always looking for new challenges in enterprise storage. Drop us a line and we will be in touch.
Enter a search term below