Data Story Telling - what, why, how?

There are a lot of resources where we talk about Data Scientists, Data Engineers or Data Analysts. There's a lot more when we talk about the difference between the said 3 roles. This newsletter doesn't focus on those aspects. Here we all collectively talk about how each one of us can be a data story teller irrespective of whatever role we're in or working towards. First, the pre-requisites!

Well, there are none! Whether you're someone who's starting out with data or a veteran Data professional, there's some way or the other in which we all can be data story tellers. So the fact that people feel like Data related jobs are for computer folks is irrelevant. Tools whilst necessary, are a means to get our motives fulfilled. Be it that you use excel, power bi or any tool that remotely allows you to view and understand data you're more than qualified to get into story telling!

So what is data story telling? Data story telling is a means where people irrespective of their domain or industry can understand the depths as to how a feature or company is performing with the help of data.

For example, let's say your organisation introduced a new feature where it allows the users to shop for shoes alongside clothes through your e-commerce application. You as a data oriented person have the insights as to how your DAU (Daily Active Users on the platform) and MAU (Monthly Active Users on the platform) looks like on a fairly regular basis. Now, if you simply observe the change in numbers in terms of those 2 key metrics, you'd understand as to how your company influenced traffic on it's e-commerce page, before and after the announcement.

Let's look at another similar example. In case of Apple, It took the company 48 years to reach a valuation of 1 Trillion $. Shortly after, within the next 2 years the company had doubled it's valuation to 2 Trillion $. The rate of growth in terms of company valuation before and after hitting the 1 Trillion $ mark is a story that a data story tellers looks into. With statistical backing as to the growth in terms of Percentage, Actionable events in the company that stirred the growth in valuation is the statistics that enable us to understand whether our actions are headed in the right direction or require any changes.

Now you may ponder over as to Why do we need data story tellers? Wouldn't a data scientist dig up the exact same statistical reports and help with analytics?

Well the answer is a little complicated in this scenario. Justifying as to 'why' you might need a particular person/feature is largely dependent on the impact that it brings to your organisation. Speaking in terms of the difference between a data scientist and a data story teller is a thin space that I believe is Improv comedy.

A data story teller observes the key aspects that a colleague from say, the business domain would definitely relate to. The key then lies in presenting the insights in a manner with which people feel at ease in.

A professional in data visualization, might build the most innovative charts in the form of Sankey Diagrams, Scatter plots or more. But they are generally difficult to be understood by the masses unless explicitly explained.

The point being, I'm not asserting that we should hold on to bar graphs or line charts with our lives just because they are easier to understand by the masses, but a good data story teller combines the complexities of data representation in a manner that we all feel familiar to.

To learn more about how a data story teller connects to familiar environments and makes data understandable to the masses you can check this TedxBroadway by Ben Wellington where he goes to analyze the data in New York city to great depths and connects it to basic features that we always come across in our daily lives as parking tickets or citi bikes.

Lastly, the How? How does one, become one with the art of data story telling? Well, ease out business folks as this is no Computer science.

Ideally, if you're on the path to becoming a data scientist or analyst, suck it up and prepare for a life where you focus on statistics and programming. But even for beginners, we are all familiar with excel (even if you haven't used it, you'll get the basics within 30 mins of using it) With the help of simple excel functions as mean, median, mode you can be a data story teller.

Let's assume we have a table containing data of air pollution in the city of Delhi within the last 5 years. So the first coloumn contains the date and the second contains the air pollution number (generally known as AQI or Air Quality Index). Let's see how with a simple 'mean' or average function you could tell a story about the city of Delhi!

Let's look up the date on which Delhi Govt. announced complete lockdown due to the COVID-19 Pandemic. It's 19 April 2020. We segregate our air pollution data into two halfs, one where the date is before 19 April 2020, the other after 19 April 2020.

I did some analysis on the same and here are my findings :

If you look into the graph for the city of Delhi, the air pollution has decreased drastically from when the Govt announced the lockdown of the city. The same is also true for other cities throughout India.

A data scientist would leave it at that, the statistics are drawn and there is a decrease. But as a data story teller we'll go one step further.

So what do we know, there was definitely a decrease in the use of automobiles which are a major source for air pollution, maybe that was a factor that improved the AQI?

Secondly, Even after all the trouble we had to go through (and still have to go through) due to the pandemic, did it actually end up helping our environment? So yeah! Corona does help the environment!

The statistics of air pollution are as boring to me as they are to the next person, but the influence on environment due to the pandemic (and a drastic one at that) is sure to grab a few more years.

The association of the story, made the data even more insightful and easier to remember whilst going for business decisions or evaluating features. The key highlights of a good story teller are the impact the stories have in the longer run.

If you're looking into more data story telling content do consider subscribing to my newsletter! I publish weekly on LinkedIn and the best part, it's for free!

If you want to have a deep dive look into my analysis or want the dataset to have a similar analysis yourself, feel free to reach out to me on LinkedIn (Gyanesh Samanta) or Twitter.