Our terms of service are changing. Learn more.

Publication

Data Science   Opinion

How to Master the Art of Storytelling as a Data Scientist?

Last Updated on December 22, 2020 by Editorial Team

Author(s): Saeed Ahmad

Data ScienceOpinion

Photo by Sinitta Leunen on Unsplash

You are a data scientist, which is the reason you are reading my article.

You were given a classification problem.

You think you used the best model for modeling it.

But someone asks you to explain the process that how you did it?

and you are lost…….

Yes, it has happened to me in the past.

You are no different.

As data scientists, we play with data as a part of our daily jobs. We are used to extracting insights from the data provided to us and find some meaningful relationships amongst different variables in the data.

After the tedious process of deriving insights is complete, we are often required to communicate our work to different stakeholders. Sometimes we also need to pitch projects. Both of these require us to narrate our work in a manner that’s understandable and relatable to the intended audience.

Why’s Storytelling necessary?

Computers are emotionless, and they understand the programming languages so we work using a variety of those while we are talking to computers. While human beings understand natural languages and are creatures driven by emotions.

“When dealing with people, remember you are not dealing with creatures of logic, but with creatures of emotion.” — Dale Carnegie

Human beings love stories. They are more inclined towards accepting something that’s related to emotions and that’s exactly what stories do.

How to tell a story?

We do hear and see a lot of stories around us. More or less, stories have 3 things in common. You have characters, conflict, and conclusion in every story.

For a data scientist, there are 3 fundamental parts of a story:

Identifying the problem:

First of all, it’s necessary to identify the problem for your audience. You need to tell them how you are going to collect your data, what are the different sources for that, or maybe you need to check that if there are some ready-made data that you can use to experiment with.

Alongside that, you need to make sure that your data is not skewed or biased or something like that. Also, you must have documented strategies to remove bias or remove other anomalies in the actual collected data.

“Errors using inadequate data are much less than those using no data at all.” — Charles Babbage

It is necessary for you to outline a clear line-of-action for the audience at this stage so that they can actually understand both the problem and a roadmap for the solution.

Presenting the solution:

After you’ve identified and explained the solution to your audience and stakeholders, you’ll need to present the solution.

Your audience may ask you different questions about the data collection process, exploration or how did you model the problem depending on the competency level of the audience.

“Invite your Data Science team to ask questions and assume any system, rule, or way of doing things is open to further consideration.” — Damian Mingle

If they are technical they will be more focused on the engineering side of things. However, if they are non-technical they might be interested in knowing the solution which is more cheap, quick, and easy to understand.

Impact of the solution:

One of the important part of the storytelling as a data scientist is the ability to relate your solution to the final impact.

There can be various impacts of a solution. It may be some predictions that can save revenue for a company, or it can be finding an optimal path for a user to reach a certain place. It can also be an analysis of a climate change issue using Machine Learning that can help in the conservation of the environment.

“Hiding within those mounds of data is knowledge that could change the life of a patient, or change the world.” — Atul Butte

Actually, it depends on the type of problem that you are solving and the impact it can have on the life of different actors in the scenario. Actors are the people who are directly or indirectly impacted by your solution in one or more ways.

Takeaway:

In my opinion, for a data scientist, mastering the art of storytelling is really important. On a day-to-day basis, we are dealing with different problems and writing intelligent models, but alongside that, it’s also important to explain what you are doing in an easy and intuitive way.

This is where your storytelling skills come into play and you can really shine as a data scientist. Believe me, mastering this skill will differentiate you and help you stand out as a data scientist from your peers.

It will not only help you develop the business acumen to understand the business side of things but will also help you to shine as a professional.


How to Master the Art of Storytelling as a Data Scientist? was originally published in Towards AI on Medium, where people are continuing the conversation by highlighting and responding to this story.

Published via Towards AI

Feedback ↓