By day, I’m a data scientist. I spend most of my time looking at tables, graphs, and numbers; running statistical analyses; and solving business problems with predictive analytics and machine learning. While those are the typical daily tasks, you might be surprised that my job also involves quite a bit of writing.

Yes, even data scientists need to be able to write effectively.

But I don’t just mean technical manuals, white papers, and reports. This goes beyond the basics of technical writing. I also have to translate my findings into blog posts or reviews for critical stakeholders and even those who have next to no experience with the data.

Thankfully, I studied psychology in college. Through my undergraduate and graduate career, I learned how to explain my findings in layman’s terms, and because I also love writing, this allowed me to combine my analytical thinking with my creative side.

Why is Writing Important for Data Scientists?

You might ask – why is writing important for data scientists? Isn’t there someone else who takes care of that stuff? 

You’d think that all copywriters handle data write-ups, but you’d be surprised to find that copywriters also need to communicate with the data engineers and data scientists to tell a brand’s story. I admit I have a bit of a competitive advantage when it comes to these tasks. Because I studied the elements of storytelling for my fictional pieces, I’m better able to apply them to my data analyses.

For instance, as I lay out my coding notebook for my analyses, I also keep track of what I’m doing by taking notes. At first, it’s just commented in the code so that I know the transformations and calculations, but later I expand this to full markdown which involves headings, paragraphs, bullet points, and more.

Why do I take the time to do all this?

Because if the person making the company’s decisions - whether that’s the CEO, the Chief Business Officer, or another executive – isn’t engaged by my findings or doesn’t fully understand what I did, then essentially the whole project was useless.

Dos and Don’ts of a Data Write-Up

When I approach my final data write-ups, I try to keep some key points in mind. Below, I’ll go over some of my major Dos and Don’ts:

Do

• Start with a problem – If you’re analyzing data, there must be a problem to solve or a goal in mind. If you going into a project just trying to find anything significant, you’re set up to fail. Give your readers the context.
• Create meaningful visualizations – They say a picture is worth a thousand words, and that can apply even in a complex data analysis. If I want to show someone the relationship between two variables, I might briefly explain it (with a sentence or two) but then really drive the point home with an interactive graph.
• Provide a solution – What did you actually find and what does it mean? If I can’t answer that at the end of my write-up, then I didn’t do my job as a data scientist or a writer. If someone gave you a complicated math problem and you spent hours doing all the calculations and writing out the work on scratch paper, but you didn’t provide a solution, then you got the question wrong. By the end of a data write-up, there should be a clear conclusion and interpretation of the findings.
• Introduce the next steps and implications – What does it mean? Based on what you found, it’s crucial to go a step or two further. Is more research needed? Is there a new question to answer? All of these things make up part of your data’s story.

Don’t

• Rely on technical jargon – Yes, sometimes technical jargon is necessary, but depending on your audience, it could be unnecessary. If I’m trying to write a blog for my brand’s customers to understand how we came to a business decision, I don’t want to just throw out terms like K-Means clustering, logistic regression, or neural network without an explanation. If I do introduce those terms, I need to at least describe what they mean in a way that any reader can understand.
• Copy and paste statistics – This is my biggest pet peeve! Honestly, when a see a write-up that’s supposed to go to the general public and it has the stats copied and pasted from the code without context or elaboration, I almost want to stop reading myself – even as a data scientist. There’s something to be said about introducing statistics as they are needed. 
• Make assumptions about your audience – If you want to write for your audience, research your audience. Don’t make assumptions. Look into who the target market is for your piece – whether that’s the data team, business stakeholders, or the customers – and tailor your writing based on the research you’ve conducted.

Data Science Involves More Than Technical Writing

Data science and creative writing aren’t as different as people might think. I say that as someone who practices both skills regularly. Even though the majority of my day job focuses on technical communication, having a background in creative writing and employing the help of an agile copywriter can take a data science project to the next level.

If you’re looking to hire a writer who is skilled in turning your data into a meaningful message contact us today.