In late 2011 I moved from life as a Computer Science PhD student to my current role at AetherWorks. Going from a town of seventeen thousand people to a city of eight million was a big change, but going from a life in academia to life in a start-up was much easier than you might think.
This post is a response to two sentiments that I’ve often heard expressed, which I’ve paraphrased below:
“It must be a major change going from a PhD to the world of work.” – The “transition” problem.
“A PhD doesn’t prepare you for life in the world of work.” – The “relevance” problem.
Clearly I don’t think this is true, but I think it’s surprising how easy my transition was, and how useful my past experience has been.
Is there a chasm between the roles of a CS PhD student and a start-up founder? I don’t believe so – in fact, when writing down the stages of each role, the job descriptions are strikingly similar:
Find the Problem: In the initial stages your focus is on finding a problem. You search for a gap in existing work that is ready to be exploited.
Initial Research: You spend the initial months of your work scoping out the identified area, examining existing work and figuring out what related areas have potential relevance. In this stage, you work to find out where the problems are, and start to think about the best ways to fix them.
Development: Once you have established the problem space, you start developing your idea. This should be fast, because you want to show the initial results of your efforts as quickly as possible. You start developing your MVP.
The earlier you get your work out, the quicker you receive the feedback that is critical to your progress: Where can you take the idea? Has it been done before? Does it need to be refined?
Pivot: At some point in these processes you realize that your initial idea wasn’t quite as original as you thought, or you find a related area and a problem in much greater need of a solution. You pivot.
Titles: While this is happening you keep a healthy disdain for your job title, because it seems to mean very little. When you work with such a small group of people, you have to wear many hats — the developer, system administrator, network engineer, and barista are all the same person.
Collaboration: You spend days with a very tightly-knit group of collaborators, working through ideas and combing over documents. When the work is presented one person might get the credit, but it’s a team effort.
Self-Motivation: When you are one of only a few people working on a project, a week of unproductive work can seriously damage a project’s velocity. You don’t have a boss looking over your shoulder, so you need to motivate yourself.
Networking: You don’t want to develop your idea in a vacuum, so you go out and meet others in your field at conferences and meet-ups. You learn the importance of a good elevator pitch.
Every PhD is different, but I find it hard to reconcile my experience with the view that it produces students with an intense focus but few transferable skills.
There are periods of confined study, but as with most jobs, there are opportunities for personal development. My ‘job’ happened to be the work of a PhD student, but that didn’t prevent me from developing other interests.
I tutored, lectured, and presented where possible. I even designed coffee mugs and indulged inpointless side-projects, all of which produced a well-roundedness that has served me well in a start-up. The notion that a PhD creates people that aren’t prepared for work outside of academia is not an inherent truth. As with any job, it is what you make it.
The primary difference between a PhD and a start-up is the end goal. In my PhD I created software with the goal of evaluating it and publishing results, so there were bugs and usability issues that I wasn’t focused on fixing. Conversely my current job is to make the most stable and usable product possible, which leads to a much stricter view of development, tighter schedules, and a more stringent QA process. The end goal of a start-up is to create a useful product that people will pay for, whereas the end goal of a PhD is to create something that advances the state of the art.
Some of the skills I developed during my PhD have proved to be incredibly useful for life at a start-up. I improved my time management and learned what it takes to motivate myself to do my best work, even on days when no-one is looking — skills that aren’t as necessary in a deadline-driven undergraduate degree, and which aren’t necessarily easily obtained in an entry-level job. It also made me a better researcher, and a more analytical thinker.
Ultimately, very few jobs can fully prepare you for life at a start-up, but a CS PhD can come surprisingly close.
 Major disclaimer: what I’m going to describe is and was true for me. I had a very practical Computer Science PhD, I had to write a lot of code, and I did a lot of work on software architectures. This makes for an easier transition than many other types of PhD even within Computer Science.
 Arguably the greatest output of my PhD was a talent for making smooth lattes. You won’t find this on my resume.
 A PhD is typically thought of as a very solitary endeavor, but it’s easy to forget the countless hours of discussions and input from supervisors and colleagues. Your name is on the document, but it wasn’t created in isolation.
 This is a view that I’ve heard from many people, but sadly most of the references I could find online were from PhD student’s themselves [a] [b] [c].
 This is the second greatest output of my PhD. Sadly, my most innovative designs were banned by the university’s image guidelines.
This is a repost of a blog I wrote over on the AetherWorks Blog earlier this year.