Originally Posted Here
“Soft Skills” have become a commonly thrown around phrase when discussing a job search. It’s just vague enough that people don’t have to really commit to saying anything in particular, but it’s specific enough that we feel like we know what they’re saying anyway. In essence, a soft skill is something that can improve your main job-related skills, but isn’t required for the job itself. Though it seems unrelated, the best soft skill for data scientists is actually the ability to write—and write clearly.
Why Writing?
You may be thinking: I work with technology and numbers, why do I need to write? Or But my research is so good, shouldn’t it just be about the data and not the way it’s presented?
Those are valid questions, but they ignore a key component of the way humans interact with each other: through words. Having a beautiful system or perfectly clean data doesn’t mean you’ll be able to make anyone care. What will make people care is your ability to share your ideas in a way that people understand? Writing is the most important soft skill you can have because you’ll use it for everything, no matter how techy you think you are. From pitching to your boss or applying for funding or writing for a journal, people have to be able to understand your ideas if they’re going to care about them.
This is especially important when you’re working with highly technical information, as data science can often be. Your manager might not know all the elements of a great machine learning system, but they need to know that you know how to explain it clearly enough for them to understand. They also need to know that someone can explain it to the public (or, more likely, to the marketing team to the public) so they, too, know why they should care. And writing is a great way to prove you can do both of those.
Starting to Write
Writing can seem daunting if it’s not something you’re used to or feel good at. But, as they say, practice makes progress, so you should get practicing and begin sharing your ideas.
You can write for all sorts of publications, which will both get more people to hear about your great work, and build your practice of explaining your ideas. There are hundreds of blogs on Medium which are looking for contributors, you could write a LinkedIn blog series, or you could even write for us. All of these options are lower-stakes places to practice writing, so that when you have to put together that proposal, you’ll actually know what you’re doing.
Tips for Writing
First, I highly recommend reading this piece about academic writing by Deirdre McCloskey (technically it’s for economists, but it applies to all scientific writing) and this piece on writing tips for PhD Students (which, again, is just good advice for any scientific writing) by John H Cochrane. There are also some great books that help you learn more about the structure and practice of writing clearly. I like How to Write a Sentence by Stanley Fish, On Writing Well by William Zinsser, and The Elements of Style by William Strunk Jr.
Second, practice writing in drafts, and giving yourself time to edit your pieces. This means writing the first draft, leaving it for a couple hours or days, re-reading and editing it, and repeating the process. In the editing process, it can be helpful to print out your piece and correct it in pen, to read it out loud, or to have someone you trust read it and provide feedback. All of these methods serve to catch errors, places you’re being unclear, and spots that read awkwardly.
Third, and finally (and again), just keep practicing. No one’s first article is a work of art, few people’s first pitch gets picked up. But the only way you’ll get better at this important soft skill is by practicing it. Learning to write, and learning to write well, will be invaluable to your company, and to your career.