Data Science Writing is Different
Don’t Take Advice from non-DS Writers
May 11th, 2020 | 4 min read

Do you get discouraged when researching how to get started writing on Medium? Are you finding all the articles are telling you to “write every day”, “publish 3 blog posts a week”, or “write for a minimum of 2 hours a day”?

When I first decided to write on Medium, I read that same advice and thought that I didn’t possibly have enough time to write every single day. This led to me putting off writing for months, discouraging me from starting my writing journey because I’d “never had enough time”.

If you’re thinking about getting started writing Data Science content on Medium, don’t get discouraged. What you should realize is that the inherent nature of Data Science writing is different — there’s groundwork that needs to be laid. This isn’t our full-time job, and there are hundreds of hours of research and coding and design that need to be done before a single sentence can be written. It took me months of stalling and then a few articles that I eventually “stumbled upon” for me to finally figure this out.

Not Following the Advice

When I finally got around to writing, I forgot about all the advice and just wrote when I felt like it and when I had the time. In fact, my first article came to me during a client meeting where I was so fed up with being told to “just look at the data” that I immediately got off the phone and wrote the whole thing in one giant sprint. It didn’t take days of sitting down and writing at the same hour for the same amount of time — it just came to me.

The same thing happened for my second article; I was researching some techniques for a side project and came across this algorithm and decided “I’m going to figure out what this thing does”. I realized there wasn’t a ton of information out there and saw the opportunity to write something that wasn’t well documented for those that might need the information.

My third article was a bit different — it didn’t come internally but from conversations I was having with friends and family. During the early months of Coronavirus, I kept hearing two sides of the same argument over and over again: “the media is overreacting” and “we’re not taking this seriously enough”. I thought “well which one is actually true?” — “can I quantify just how much or how little the media is overreacting?”. I set out to do just that. I gathered the data, did the analysis, and wrote the article — all in 1–2 weeks, which by the “advice” standard, would have been too slow.

And now this post — an article that I’ve been thinking of writing for quite some time. If I had followed the advice of the typical “How to be Successful on Medium” article, this post would have been written several weeks ago. But to be honest, I just didn’t have the time. It took until today when I read another one of those articles and felt like I needed to say something from the perspective of someone writing Data Science content. I opened up a new draft and wrote this post in one sitting on my phone.

Why Writing for Data Science is Different

The first few pieces I put together didn’t come from a routine of sitting down every day at 6 am and writing for two hours straight. They came to me when I was working on projects and continuing my learning journey. As I continued to write, I began to realize that writing for Data Science is different in two ways:

  1. Not everyone has the time or energy to sit down and write every day. If writing is your main source of income then yes, you should do that. But for others (especially Data Scientists) this is a side hustle, a portfolio builder, or a creative outlet — and we shouldn’t feel obligated or stressed to write every single day.
  2. Most importantly, it’s important to realize that Data Science writing is inherently different. We aren’t lifestyle bloggers who sit down every day and talk about how we feel or what we’re doing this week or how to better improve ourselves. Our writing is technical — it takes days and weeks of research or analysis (or both) to write a 5–10-minute piece. Don’t put the pressure on yourself to publish every day — or even every week. It’s okay to start slow. Focus on what matters — the project you’re building or the concept you’re learning. Writing is just the final piece where you present your findings. It’s an important piece, but the other pieces take time as well.

Get Started, Not Discouraged

I hope this article didn’t come off too sour. I hope instead it inspires you to not get discouraged by all of the “advice” out there, telling you to do more. Most of those writers don’t need to sit down and study the objective function of an algorithm or write 1000 lines of code in Jupyter. Get started when you want to get started, stay the course, and don’t listen to the noise — the writing and the consistency will eventually follow.

Thanks for reading! If you want updates on what I’m writing, side projects I’m building, or articles I find interesting, feel free to join my newsletter — Aspiring Data Scientist.

This article was written by Tom Sharp.
Consider supporting them so they can create more quality content.