The world’s leading publication for data science, AI, and ML professionals.

The Importance of Writing as a Data Scientist

Looking beyond just data, numbers, and graphs

Photo by RetroSupply on Unsplash
Photo by RetroSupply on Unsplash

One of the most underrated skills that a data scientist can develop is the skill of writing.

Why is this? When we think about the role of a data scientist, it breaks down to four main buckets:

  1. Being the person who understands the data the most: When working with engineers, product managers, or data scientists from other teams – you should be the data expert. This is because you the most hands-on experience with the data in your domain. Not only that, but as a data scientist you should be shaping what/how data is tracked and how they are defined.
  2. Making sense of the data: You might be ingesting millions or billions of rows of data per day. How do you make sense of all of this? This is where SQL, R, Python, and analytical skills shine.
  3. Communicating an action to take based on the data: You looked at the data and you found a clear opportunity for improvement. How do you convince others that they should dedicate their time and resources to work on this?
  4. Understanding the impact of the action: You convinced your stakeholders to build a feature to improve the product. Did this work as intended? How much did this feature improve the product?

I think that a lot of aspiring data scientists tend to overindex on the first two buckets above— they focus 90% of their time trying to be experts in SQL, Python/R, and statistical modeling. While these skills are fundamental in a data scientist’s arsenal, I think there should be more emphasis on Communication – specifically writing.


Why is writing so important?

Simply put – it’s a Data scientist’s most effective method of communication.

Photo by Jason Rosewell on Unsplash
Photo by Jason Rosewell on Unsplash

Right now, with most of us doing work remotely, we’ve had to move to a more asynchronous, less-in-person method of communication. Whether it’s writing something informal such as a Slack message, or a drafting a report that will be shared with the management team, practicing and honing your writing skills will help get your point across. This leads to impact. This reduces miscommunication. This leads to recognition.

Because data scientists are hired to be experts at data and making sense of the data, the work that we do is inherently complicated. This makes communication especially important. If we spend a week diving into a bunch of complicated datasets and writing thousands of lines of code to analyze it, we need to think of a way to summarize our work so that others can quickly understand your most important findings. What is the key takeaway? What should we do with this info? How did I arrive at this conclusion?

Why is writing more effective than other forms of communication? Writing also allows us the luxury of being deliberate in the message we want to craft. Done effectively, it’s an efficient way to get a point across because it doesn’t depend on searching for time on people’s calendars for a Zoom meeting. There is also more permanency in writing, so if you wanted to look back or search for something you said a few months ago, that should be doable. Sharing is also much easier – whether it’s a Slack message, Google Doc, or PDF. This allows your message to spread more easily.


How does one get better at writing?

This is a little tricky. Because writing is more of an art compared to data analysis which is more of a science, it’s a little harder to put structure around improving one’s writing. That being said, getting better at writing should be more enjoyable. Here are a couple of things you can do to improve your writing:

  1. Read more. I would argue that it’s impossible to get better at writing without reading more. Read anything. News articles, nonfiction, fiction, etc. As long as the author is a great writer, you’ll subconsciously absorb how to be a better writer just by reading more.
  2. Read more – with a focus. Learn how writers hone their craft. One of my favorite authors, Steven King, has a great book on how he approaches writing. Murakami, my favorite author, has one as well. While both of these authors are mainly fiction writers, their craft is impeccable and I learned a lot from reading about their process. You might look for a particular Data Science blogger who’s writing you admire and try to read their articles.
  3. Write more. Medium is a great place for this. If you’re passionate about a topic, writing about it would be a fun way to get practice.
  4. Ask for feedback. Find a mentor or manager who you think would give helpful feedback. Before sending out something, ask if they can look it over and give suggestions.

Conclusion

Hopefully this article has convinced you of the importance of Writing as a data scientist. While it’s great to be an expert at data analysis, please don’t overindex in it. If you’re spending most of your time bogged down in Udemy classes, I would hope that you’re also making time to focus on another important aspect of your career – writing.


Related Articles