Since I began my tenure at Wayfair as a data scientist, many PhD students and postdocs have reached out to ask for advice about the transition to data science in industry. After a fruitful collection of conversations, I’ve realized that there are popular misconceptions about what it means to be a data scientist.

The reality depends on the industry and the specific company, but having made the leap myself, some light can certainly be shed on the matter to help current PhD students and postdocs thinking ahead. Read on if you’re considering a career as a data scientist in industry and find out how to prepare for it.

Myth 1: Coding is the most important skill for a Data Scientist

Many students seem to have the impression that being a data scientist means sitting at a computer all day developing machine learning algorithms. This is not the case for many data scientist positions in industry, and especially not at Wayfair. Here, we form cross-functional teams to solve business challenges, leveraging technology as we go. For example, my team members meet with marketing managers and analysts every week to brainstorm project ideas and discuss the performance of specific marketing campaigns and data science models. Thus, your success as a data scientist depends on many skills beyond your proficiency in Python, R, and SQL.

Some of the core qualities of a data scientist include:

  • Business acumen to understand non-technical challenges and perspectives
  • A knack for critical thinking and a passion for problem solving
  • Attention to detail in order to discover nuances in data
  • Understanding of the pros and cons of common machine learning algorithms
  • Excellent communication skills to explain technical findings to non-technical business partners and great collaborative spirit

Compared to coding, these skills are much harder to develop. The good news here is that since you’ve worked hard on your PhD, you likely already have most of these qualities. This is a great segway into our next point.

Myth 2: My PhD becomes worthless if I transition into a new domain

Many people assume I have thrown away my most valuable asset — a PhD in Biophysical Chemistry — by becoming a marketing data scientist. However, my doctoral training provided me with many transferable tools and skills, other than my knowledge of quantum mechanical tunneling at enzymatic transition state (if you don’t know what this means, then you understand my point).

Getting a PhD is a tremendous character-building exercise, and that’s why PhDs are in high demand as data scientists in industry.  To earn your PhD, you need to solve a series of complex problems that you’ve never seen before, that nobody knows how to solve; never mind the possibility that these problems might actually be unsolvable. While navigating these challenging times, you’ve learned how to approach new problems and ask the right questions, not to mention learning to handle stress, exhibit patience, and practice persistence.

I have interviewed around 100 data scientist candidates to date at Wayfair. PhDs often perform better at business case interviews, even though they lack a background in eCommerce or marketing analytics. It is not your past content knowledge but the critical thinking and problem solving skills that will help you foster a new career. You can carry those skills with you into any job, even if it seems a far cry from your PhD field.

Myth 3: Companies don’t give you the time to learn new things

Academics love spending time diving deep into new concepts. Consequently, some PhDs are concerned they’ll surrender the joy of learning something new at an industry job. I, too, had the same impression before joining Wayfair, but this is happily not the case.

In fact, in my past year at Wayfair I have learned more new concepts, skills, and techniques than I was ever able to in any year in academia. This is in part the result of switching to a new field. But more importantly, with data science growing and changing rapidly, we need to keep learning in order to perform well. Our data science teams have weekly meetings for project deep dives, similar to academic lab meetings, where we knowledge-share and receive feedback from our peers. There are also journal club meetings and “Lunch & Learn” lectures available to discuss academic research papers and state-of-the-art algorithms. In addition, Wayfair’s learn@work programs help all employees improve or gain a variety of proficiencies: Not just technical skills either, but also management and leadership competencies.

Myth 4: I have to complete a Data Science bootcamp to get into the field

One of the most common questions I have been asked is: “Should I enroll in a data science bootcamp?” For the courses that teach you the basics in coding and machine learning, the answer is no. You should be able to pick up these skills by taking courses on Coursera, doing Kaggle competitions, and/or beginning to work on data science projects. Pro tip: Even seasoned data scientists Google for solutions whenever they are stuck.

More sophisticated training programs such as Insight Data Science can help refine your current skill set if you are just about ready to apply for an industry data scientist job.  However, Insight programs are only seven weeks long, thus, it won’t transform you like your PhD training would have; it’s mostly helpful for those who have worked on data science projects previously.

So, how should I prepare for a career in Data Science?

You will probably want to consult other data scientists in industry for more diverse answers to this question, but alas, this is not why you’ve read this far.

First of all, ensure you have a good understanding of basic machine learning algorithms. For this purpose, I strongly recommend Andrew Ng’s Machine Learning course on Coursera. The classes are taught in Matlab, but Python implementations are available on GitHub (like this one). For Deep Learning, you can learn from Michael Nielsen’s eBook, or Geoffrey Hinton’s Neural Networks for Machine Learning course on Coursera (or the recently launched Deep Learning Specialization, also led by Andrew Ng).

Next, start working with real-world data. Find a problem that truly intrigues you. Think about what questions to ask, how to source data, and what tools and algorithms can help answer those questions. If the problem is too challenging at first, identify attainable short-term goals for the project, either the first step towards a solution or simpler, relevant tasks. Build a few MVPs (minimum viable projects) to showcase on your resume and LinkedIn profile.

Once you have developed strong core qualities and experience with real-world data, programs like Insight Data Science can help you launch your data science career much faster. In addition, the social and professional network with other Insight fellows and alumni is invaluable. That said, there are other paths that can take you to an industry data scientist job, such as a co-op/internship positions. Wayfair has its own Immersion Program at its Boston HQ that provides opportunities to learn about industry data science for PhD candidates and postdocs a year before their intended completion dates.

If you really enjoy taking on new challenges and love diving into data to retrieve insights and tell stories, your career transition into data science will be a thrilling journey. So, get ready to expand your horizons, hold on tight, and enjoy the ride!