What should an undergraduate student who’s inspired to change the world do? Should they continue their education in a PhD or found a startup? I have recently done both – at the same time – and my experience suggests the combination is better than you might expect.
I have just submitted my thesis in computational linguistics (the sub-field of artificial intelligence devoted to automatically understanding human language) at the University of Sydney. I am also a co-founder of Grok Learning, an education technology startup that aims to teach the world’s kids to code.
At first glance, there isn’t much in common between a PhD and a startup.
A Doctor of Philosophy, from the Greek philosophia –- love of knowledge and pursuit of wisdom – is awarded for conducting independent research, demonstrated through a significant contribution to knowledge in a specific field. The stereotype is deep, rigorous thought carried out over many years in a narrow field that few people in the world understand.
A startup is an organisation searching for a business model. It’s a business that isn’t yet viable. The stereotype is brash, fast-moving organisations that are looking to reshape, or disrupt, our lives within a few years. Think AirBnB, Freelancer and Uber.
So what do a PhD and a startup have in common? It turns out, quite a bit:
- Both start with an idea, the vision of which often sustains and inspires student and founder, and unites their team in a common purpose
- Both require financial (and social) sacrifices to achieve that vision
- Both are risky with uncertain rewards. The path is littered with failed startups and unfinished PhDs
- Both involve the search for answers in unknown conditions, where supervisors and mentors can guide, but only your exploration can (hopefully) uncover the solution.
This last point profoundly shapes PhDs and startups, and determines who excels in them. According to US author and entrepreneur Eric Ries, a startup is an organisation trying to deliver a new product or service under conditions of extreme uncertainty.
This sounds a lot like my own PhD experience and that of my peers. The very popular Lean Startup methodology that Ries champions applies the research methods familiar to students in empirical disciplines, such as science and engineering, to growing a business.
The focus of the lean methodology is learning which ideas work and which don’t through efficient, rapid experimentation that iteratively improve the startup’s product-market fit, finding a product that people want to buy at a price the startup can sustain.
A/B, or split testing, involves experimenting on your customers by changing one thing at a time and keeping a control group, and recording the impact of the change. That’s the scientific method!
A minimum viable product involves developing just enough of the product to enable these experiments, sometimes with almost no other functionality at all. This reminds me of every experimental setup I’ve seen, held together with spit and string as the student races for a publication (or thesis) submission deadline.
And if the experiments continue to produce negative results, knowing when to pivot –- a major course correction to a (hopefully) more fruitful line of exploration –- is one of the most painful decisions for a student or founder (and their advisors) to make, especially when substantial time, energy and often money has been spent exploring a particular direction.
Finally, students and founders are both all too familiar with the idea of a runway –- the length of time before funding (from savings, investors or a PhD scholarship) or enrolment runs out and the student or startup is no longer achievable.
Into the unknown
While everyone asserts that written and spoken communication skills are critical, students are often surprised that we must promote and position our work in the competition of ideas and actively develop our own academic profile, but marketing is literally life and death for entrepreneurs.
Of course, not everything is parallel in research and startups. A key difference is the timeframe and the urgency it demands. Many startups are born and die in the time it takes to finish a masters (18 months), let alone a PhD.
In a startup, the goal is to learn fast, doing experiments on a daily basis that could substantially improve your product. In research, you might take months to set up a single experiment.
So, while PhD students are well suited to the challenges of exploring the unknown, we may need to work on being more dynamic and responsive (their supervisors might agree too, but industry might say that of many academics).
Universities could encourage a flourishing startup ecosystem and energise their PhD programs by intermingling PhDs and startups, and providing founders with APA-like seed funding (during or after their degree) and considerable enrolment flexibility.
Partnerships that integrate researchers and entrepreneurs in residence could create dynamic hybrid accelerators and postgraduate programs. In-house financial, legal and administrative support would help startups begin, and a lower-friction approach to protecting IP would allow students (and academics) to turn ideas into ventures more often.
Finally, universities need to recognise the value of startup creation as an academic output.
There is much to do, but for now I’m taking my PhD experience and leaping into the unknown.