How to leave academia and get a data science job

“Graduate opportunity: PhD + minimum 2 years’ experience required”

If you look at job adverts for data scientists, you’ll often see “PhD required” or “PhD preferred”. On the surface, it therefore looks like having a PhD (and possibly also postdoctoral experience) should give you a great advantage in landing your first role. 

However, it can actually be quite tricky to get your first commercial job after academia. As well as needing a PhD, job adverts often state that prior data science experience and commercial knowledge are required. These can be incredibly challenging to obtain, especially if your research isn’t directly related to data science. Even if you do have some of these relevant skills, you may still face rejection after rejection from your job applications with no real understanding as to why. 

Why a PhD isn’t always enough

When I started my job search, I assumed that my PhD would be an asset and that I wouldn’t struggle to obtain a commercial role. However, when I read through my university’s guide on obtaining a non-academic position, it was quite negative. They stated that companies may not “know or care” about my PhD, and that they may have negative preconceived ideas about academia.  

Many of my colleagues told me that they struggled more than expected to get their first non-academic role. Little experience with working as part of a team, limited chances to discuss your research with non-technical audiences, and niche research subjects are just some of the barriers that might hinder the success of your applications. Code written during a PhD is often written as quickly as possible too, resulting in messy work that doesn’t follow the best practices for collaborative coding. 

A PhD gives you many beneficial skills, but the narrow focus and isolation often means that you will lack some of the skills needed to succeed in a non-academic role.

How your PhD can help you get a job 

It’s not all doom and gloom! An academic background gives you a wide range of skills that are vital to become a great data scientist. For example, research involves lots of problem solving and working on complex technical projects.

You can show dedication and passion for a subject, regardless of often challenging working conditions and little guidance. You’ll have had many opportunities to work and learn independently, showing that you’re self-motivated. You’ll often have great technical and critical thinking skills too. So how can you use these skills to land your first commercial role?

How to land your first commercial data science role

Stuart and Tom have previous written a blog post about how a PhD isn’t actually necessary if you want to become a data scientist.

Although academia gives you the opportunity to develop and certify your data science skills, there are lots of other ways to build expertise in these areas.

Likewise, if you have an academic background and no commercial experience, we believe that you can still succeed with your job search (and many data scientists at Peak are proof of this!)

Here are some tips to bridge the gap between academia and commercial data science:

1. Showcase and build your non-technical skills now

Make sure that people know you are more than just your research. Employers need to know you’ll thrive in a collaborative working environment:

• Teach: Teaching at university can highlight your ability to explain complex topics in a simple way. It can demonstrate people management and presentation skills, as well as allowing you to evidence your experience in helping others problem solve
• Present: Presenting at conferences, especially to audiences outside of your specific field, is a great way to show you can confidently present in a non-academic environment
• Do outreach: Participating at outreach events (e.g. public or schools) is valuable for both you and the people you connect with
• Join committees and clubs: Being a part of committees or clubs can show enthusiasm for topics other than your research
• Speak to data scientists at events (e.g. hackathons, lectures): This gives you a great opportunity to find out more about a typical data science role and how to excel with applications and interviews. It also allows you to explore how the role can differ between companies, as well as getting expert advice on anything data science related

2.  Boost your technical skills

Make sure you comprehensively understand your research, especially any coding and mathematical work. Data scientists often use R or Python, so it’s a huge advantage to know at least one of those. If you have any major gaps in your technical data science knowledge, check out our blog on how to self-learn data science.

3.   Where possible, get experience of applying data science in a non-academic setting

Be aware of the differences between academic and non-academic applications of data science. Even a research career with a data science focus isn’t likely to give you a full experience of how your work will look in a non-academic setting. Try to get non-academic experience of data science through attending talks, reading blogs, and speaking to data scientists to learn about what to expect.

4.   Make your job applications count

• Make a non-academic CV: Avoid technical language that is specific to your research, and relate your experience directly to the job advert. Focus less on your publications and instead highlight the skills related to the role’s requirements
• Look out for roles that let you learn on the job: Find organisations willing to support your growth and help you become a great data scientist. Some organisations have job adverts specifically for people leaving academia where they’ll offer additional support to help with the transition
  
• Don’t avoid recruitment agencies: Having conversations with recruitment agencies can help you understand the roles available and what organisations are looking for. They can offer you a perspective on why your application may not be succeeding, and what level of roles you’d be most eligible for
• Accept that you’re likely to get a few (or a lot of) rejections: Some jobs do require experience of applied data science, meaning the organisation will prioritise people who meet that requirement ahead of those who don’t. Where possible, ask for feedback from the company as this can help you with your next applications

5.   Make sure you’re interview-ready

Think of every interview like a potential mini PhD viva. Make sure you prepare carefully for your meeting with the hiring manager. Here are a few tips on how to get ready for the interview:

• You might be asked to explain why you want to leave academia: It’s important to ask yourself why you want to leave academia before you make any decisions. Prepare an answer for interviews that focuses on the positives of applied data science, rather than the negatives of academia
• The interviewer might ask you about your research: You may be asked to summarise your research, discuss any flaws, or explain why you did something in a certain way. You’ve spent so much time on your research, so make sure that you’re engaged with it!
• Demonstrate awareness of the challenges you might face: Be prepared to discuss your weaknesses and how you might overcome them. It’s not a problem to lack experience in certain areas (everyone will when they start their first data science role) but it’s important to show you’re aware of any weaknesses and that you’re willing to learn more once you start the job
• Show interest in the organisation you’re applying to: Although your first job is a great learning opportunity (and it’s fine to mention this), make sure that you explain what it is specifically about a company that sets them apart. Read blog posts, look at social media, and find at least one thing that excites you
• Speak to recruiters or other people from the organisation you’ve applied to: This is useful if you’re struggling to get through the first stage of the interview process. This could be through LinkedIn or company-hosted events. It can be a great way to stand out on your application, and may help progress you to the next stage

Conclusion

If you’re thinking about leaving academia, data science can be a great career choice to consider. Your academic experience can give you a lot of the necessary skills that are required, but you might be missing some of the skills that people ask for in commercial data science roles.

Try and gain experience however you can, and be honest about your weaknesses at the interview. Practice makes perfect – you probably won’t succeed in your job search straight away, but keep trying and you’ll get there eventually! 

If you still have questions about how to land your first data science role after academia, take a look at our ‘Getting into data science’ Community discussions.

Acknowledgements 

Thank you so much to all of the data scientists who offered advice to help with the writing of this article, including Thomas Richardson, Eleana Makri, Lee Whittaker, Saurabh Singh, Tom Hassall, Lara Bogatu, and Stuart Davie.