Leaving Academia and Founding a Start-up
5 min read
Each year over 100,000 students enrol for a doctoral research program in the UK (Who’s studying in HE? | HESA). As it stands, there are just under 250,000 total academic jobs in the UK, which means there is a lot of competition for new academic roles that open up as departments grow, or existing academics leave (Higher Education Staff Statistics: UK, 2021/22 | HESA).
The disproportionate ratio of new PhDs to academic positions inevitably means that many PhD holders will never end up following a traditional academic career path. Many graduates are realistic about the likelihood that they will need to find work in industry, given that academic positions are few and far between, and increasingly difficult to obtain. Joining the workforce in industry can be a hard transition for some, while others find it a welcome relief from the pressure, long hours and lower salaries associated with academic roles.
Dr Phil Gooch is the founder and CEO of Scholarcy and has a fairly extensive academic track record. Like many others, when Phil completed his PhD, he ventured back into industry. Here’s what he has to say!
Can you give a brief overview of your education and work history? What were your plans at 18, and how did they change as you progressed through education and your career?
At school, I was always into science and did science A levels in Physics, Chemistry, Mathematics, and Further Mathematics. I assumed I would go into some kind of scientific job or something involving research, but I didn't really know. I just liked studying and solving problems and I hoped there might be a job somewhere doing that. I studied Natural Sciences at Cambridge, but I lost interest after the first year as I was more into playing guitar in bands. I wasn't into computers at all at the time. I switched to History and Philosophy of Science, a much more essay-led subject, which I enjoyed and graduated with a 2:1. I didn't know what I wanted to do next, and to be honest I didn't really have any commercially useful skills, as I'd effectively dropped out of lab science. At the time I planned to move to London and focus on working as a freelance musician while holding down a basic day job. That turned out to be unrealistic! However, my background in science and writing helped me land jobs in academic publishing - in editorial and production roles. The work was quite repetitive, and I thought: 'Some of this can be automated'. My computer skills were out of date, so I taught myself programming and attended evening classes. I started writing small software applications to do some of the manuscript editing and eventually to automate the typesetting of some of the monographs we published. At the same time, I was also doing voluntary work with a homelessness charity, helping hostel residents with literacy and computer skills. I found this rewarding to the extent that I decided on a complete career change and retrained as an Occupational Therapist. For several years I worked in a secure psychiatric unit helping patients regain skills and helped set up a work skills program for those adjusting back into the community.
But I started to miss solving technical challenges and wanted to see if there was a way to combine technology with healthcare. I did a Master’s degree in Computer Science, and then later a PhD in Health Informatics, investigating how Natural Language Processing (NLP) could be used to read and understand clinical notes and improve adherence to clinical guidelines. I also wondered if there was a way to use NLP to read and understand all the research papers I had to work through! During my PhD, I thought maybe a career in academia would be the next step. However, the career path of PhD, then postdoc, and then on to lecturer no longer existed, and all that was on offer was a series of short-term post-doctoral roles. Instead, I took up a role as a research developer in Digital Humanities, which I really enjoyed. I was using NLP to automate the conversion of documents in various formats into digital projects, such as the Cambridge Edition of the Works of Ben Jonson online.
It was through this transition from work, education, technical research, and software development, that I realised that I wanted to try to see if there was a market for some ideas that I'd had. Working for different start-ups such as Mendeley, RefMe, and Babylon Health, had given me the skills to build a 'Minimum Viable Product' for what became the Scholarcy Chrome Extension and Scholarcy Library.
And here we are today!
You've mentioned that you missed having a technical challenge, but what made you do a PhD? Why not just pick a job that could fulfil this?
I realised after something like 15 jobs in 15 years that I wasn't great at holding down a normal job and wanted to do something else. I'm not a competitive person, I prefer collaboration, although I am competitive with myself (personal bests etc). I had really enjoyed doing the Master’s, and it felt like the obvious next step to do a PhD, as my original PhD proposal was to build on the Master's project I had worked on. I needed something that would give me the space to focus on one thing for an extended period, something that the traditional workplace didn't offer. I saw the PhD as important both in terms of personal and professional development.
When you went back to university to do your PhD, what was it like transitioning back to education after experiencing the ‘real world’, and studying alongside people younger than you?
I'm often quite chaotic but I can also deep-focus on one thing for a long time, so the transition to a PhD where I had to be self-reliant and only needed to answer to my PhD supervisor was quite liberating for me. I had a great supervisor who was very supportive, and it was a mixed team of people in the department - all ages, some older than me, some younger. Returning to education later in life was not unusual at City University, London, where I did my PhD. I also got involved in some other projects and helped with tutorials for undergraduates and Masters students, so there was a lot of variety too.
You mentioned that the traditional academic route of finishing your PhD, starting a postdoc, and then becoming a lecturer wasn’t very attainable at that time. Did you ever feel pressure to pursue an academic position and how did you feel about leaving academia?
I did enjoy helping with tutorials and teaching. I also got to prepare and deliver a few lectures which I enjoyed too. My initial plan was to stay in academia. My supervisor was also keen that I stay - there was no pressure, but he was very encouraging. However, there weren’t the positions available when I finished my PhD, and one of the downsides of doing a PhD later in life (I was 40 when I started it in 2009) is that the road of short-term postdocs with possible future chances of getting a lectureship position, is a long one, and you have to be prepared to move both nationally and internationally.
So I was disappointed that I wasn't going to be able to do this without a lot of disruption and uncertainty in my life. I suppose I was naive to expect that I could just stay in London and find a postdoc position and then a lectureship in a few years. Unfortunately, it doesn't work like that anymore! I have friends who did PhDs straight from undergraduate in the early 1990s, and it was a much different landscape then. I was disappointed to leave academia. However, the PhD was very valuable - I would not have been able to have worked for various exciting startups without it. Between 2000 and 2012 I had been out of the publishing and technology industries, and the PhD helped me get back into it, and ultimately, create Scholarcy as a result.
How did you find the transition back into industry after completing your PhD? Did you notice anything different afterwards? Were you ever treated differently because you have a PhD?
After completing my PhD, I took a job as a Research Developer at King's College, London, in the Digital Humanities department. Most people there had PhDs so I wasn't treated any differently. Afterwards, when I worked as a data scientist, having a PhD was quite common, many colleagues had PhDs. For the roles in industry I applied for having a PhD was helpful, and sometimes required. I think that is less true these days working in AI: it's more about whether you can do the job, program, gather data, build prototypes and products etc, rather than having a qualification.
A common perception of academic positions and entrepreneurship is that you never get any time off. How have you been able to juggle your career, and family time? Do you ever battle burnout?
Yes, during my PhD it felt like I was always working on it, but I was able to take time off and time out. I wasn't doing a 'lab' PhD and also I had independent EPSRC funding, so I was able to be more self-directed and fortunately I didn't battle with burnout. However, the PhD did give me some of the skills needed for entrepreneurship; in that being self-reliant, taking initiative, taking risks, being able to negotiate, having to master new skills, are common to both endeavours.
Building and running a startup really does feel like you never get any time off, particularly in the early years. For the first few years, it was 24/7 and consumed every waking hour. I was working on building our entire document processing engine from scratch - partly out of necessity but also out of bloody-mindedness, I had a particular vision of what we needed to build, and the existing software didn't do everything we wanted. It was only when we started to become successful and assemble a team that I was able to take time off and step back a bit. So battling burnout was definitely a big issue for me with Scholarcy for several years.
Are there any other skills you picked up during graduate school that have been useful in industry, and how would you advise others in graduate school or thinking of doing a PhD?
Before doing my PhD I didn't know much about natural language processing or machine learning. I had to acquire those skills during the PhD because part of my project was to develop and evaluate an NLP processing pipeline for reading and understanding clinical notes and clinical guidelines.
So the skill of building a piece of software from end to end that could be deployed so that others could use it was very beneficial in industry. Improving my writing, presentation, and public speaking skills were also areas that I developed during the PhD, and these was useful when pitching Scholarcy at various startup accelerator programs. There are many courses online for learning to code, and now AI/LLMs are available, these can be a big help for giving hints and debugging. Presentation skills can be learned with coaching and practice.
Can you tell me about a career highlight that you’d like to share?
Apart from founding Scholarcy, one thing I am particularly happy with is being an early innovator in content automation technology. Back in the mid-90s, a book publisher that I was working for decided to bring some of their typesetting and pre-press work in-house. I built software for them that helped automate this process, so they could import copy-edited Word documents into the system - which was based around Microsoft Word and a typesetting package called Quark Xpress - and get a fully typeset layout in one click, which could be proofread, corrected, and sent to the printer. I think the system was in use for about 8 years.
Would you ever be tempted to return to academia in the future? As either an academic or a student?
I would be tempted to return to academia as a student, maybe to do a Masters in a new subject. I might consider returning as an academic as a way of sharing my knowledge about things that I’ve done since I left, e.g. as a lecturer on entrepreneurship or something.
References:
Who’s studying in HE? | HESA (no date). Available at: https://www.hesa.ac.uk/data-and-analysis/students/whos-in-he (Accessed: 20 March 2024).
Higher Education Staff Statistics: UK, 2021/22 | HESA (no date). Available at: https://www.hesa.ac.uk/news/17-01-2023/sb264-higher-education-staff-statistics (Accessed: 20 March 2024).