Photo by Cole Keister on Unsplash

Do You Need A PhD To Be A Data Scientist?

Ever wish you could get an easy answer to what is Big Data? Get the Big Data Checklist, totally free, along with weekly blog tips delivered directly to your inbox.

The PhD in data science. It seems like the Holy Grail.

But do you really need one to be a good data scientist?

Today’s post sheds light on when it’s ok to not have a PhD and when it is almost required.

I will help you decide if you should go for the PhD or if you are better off without one.

Scientific working

First of all, let me say this: Any degree, Masters or PhD, is more or less only a certification of your skills. It attests that you are able to deep dive into a topic in a scientific way.

What makes the PhD special is the level of scientific working and theoretical research required. The main difference between a MSc thesis and a PhD dissertation is how much it contributes to the research in a field.

A PhD dissertation has to contains your original ideas and findings.

This original content needs to result in totally new scientific findings. Findings that help advance the scientific knowledge in this field.

Also, although you might do some hands on research during the dissertation it is mainly theoretical in nature.

A masters thesis on the other hand, is mostly showing the current state of science. It contributes to scientific research only in a small way.

The thesis is more about showing that you are skilled in a field. Showing that you have some original ideas and that you can implement these ideas in a real world setting.

Applying vs Researching

There is a big difference between applying and researching. Take machine learning for example.

Applying an algorithm to solve a business case requires totally different skills then researching new algorithms.

A researcher is usually not going to realise the finished solution. Researchers know every aspect of the field they work in.

He knows the theory behind the algorithms in detail. He is able to modify existing algorithms to come op with with totally new and original ones.

Very often researchers are going to write papers on their findings to advance the global scientific knowledge. Check out Google research papers how this process looks

Appliers on the other hand need mostly different skills. Although they need to understand how something works the main focus is to implement a solution for a specific use case.

Applying an algorithm requires to do a lot of work around the actual core machine learning algorithm:

How is data coming in? How does the data need to be formatted? How do I need to preprocess it that the algorithm can use it? How do I need to post-process the data so that the users can make sense of it?

How can I make all the above work in an automated way to provide a solution for the problem.

You don’t need a PhD!

Ask yourself the one question: “Do I want to apply or do I want to do research?”

You can be an amazing data scientist by applying existing methods to solve problems with data science.

That’s why I say a clear No.

You don’t need a PhD to be a good “applying” data scientist.


Research on the other hand is different. You almost definitely need a PhD to be a amazing Researcher.

Doing research, creating new algorithms and writing scientific papers. That requires not only far deeper knowledge of a field like for instance deep learning.

Research requires skills for scientific working that you learn while doing a PhD.

I for instance don’t have a PhD and I never wanted to do one. I like building solutions that solve problems for customers.

I would be a horrible researcher! And that is totally ok.

How about you?


If you liked this one please share it with your peers on social media. And if you haven’t subscribed to my email list already consider doing that so I can notify you about my next article.

PS: A critical thought.

I have the feeling that a lot of people see the PhD as their path to money. When in fact the idea behind the PhD is a noble one:

Advance the knowledge of the scientific community. Help make the world a better place.

What is your take on PhD and MSc or BSc? Share it with me and the community in the comments section below!


Liked this one? Checkout these articles:

7 Super Simple Steps From Idea To Successful Data Science Project
Complete guide on how to do a data science project

— The Brutally Honest Truth About Learning Big Data the Right Way
Things you need to understand that makes starting with big data easy

Apache Spark: A Success Guarantee For Data Scientists
Awesome things you can do with Spark that every data scientist should know

How to Create New and Exciting Big Data Aided Products
How Netflix uses Big Data and Data Science to create products people love.

Comments 3

  1. I just completed my masters (thesis) in machine learning. I would ideally like to be a data scientist or a machine learning engineer (apply learning algorithms) in industry but as per my experience, most of the industry looks for people with a PhD or else good experience in this field. So it gets really difficult to find the mentioned job profile in a good company.

    What should I do more (don’t really want to go for PhD) to have a better chance in getting suck jobs.

    1. Post

      Hi Ryan, congrats on your masters!

      It’s true that companies are looking for PhD’s or people with good experience. I strongly encourage you and everybody else to start publishing your work and thoughts.

      It doesn’t matter if it’s a blog or on medium or GitHub.
      Apply yourself, learn new things and write about it.

      Post code and explain backgrounds.
      This will help identify you as a experienced professional.

      Yeah, doing a PhD only to get better chances finding a job is a silly choice.

Leave a Reply

Your email address will not be published. Required fields are marked *