Data Visualisation: Empowering Business with Effective Insights
Walkthrough of my experience with TATA Group’s Virtual Experience Programme
For those facing difficulties in building a portfolio to land a job in Data Analytics, let me share with you something I recently discovered! Virtual Experience Programmes are programs curated by real companies around the world and your job is to help them solve their hypothetical business problems. This is an excellent way of showing recruiters that you are able to apply your skill sets to real-world situations, which is what they are essentially looking for when hiring. I recently tried it out and I chose TATA Group’s data visualisation programme! I highly recommend this programme for beginners trying to break into the world of data analytics. For those interested, check out The Forage, which lists tons of virtual work experience programmes by a plethora of companies from many industries too!
Now, let's dive into TATA’s virtual experience programme!
The virtual experience programme starts off by giving us background information on the task and here is what it says:
An online retail store has hired you as a consultant to review their data and provide insights that would be valuable to the CEO and CMO of the business. The business has been performing well and the management wants to analyse what the major contributing factors are to the revenue so they can strategically plan for next year. The leadership is interested in viewing the metrics from both an operations and marketing perspective. Management also intends to expand the business and is interested in seeking guidance into areas that are performing well so they can keep a clear focus on what’s working. They would also like to view different metrics based on the demographic information that is available in the data. A meeting with the CEO and CMO has been scheduled for next month and you need to draft the relevant analytics and insights that would help evaluate the current business performance and suggest metrics that would enable them to make the decision on expansion.
We are given 4 tasks to complete for this programme and they are:
- Farming the Business Scenario
- Choosing the Right Visuals
- Creating Effective Visuals
- Communicating Insights and Analysis
Without further ado, let us go through task 1!
Task 1: Farming the Business Scenario
In this task, we are required to anticipate 4 questions that we think will be important and relevant to the CEO and CMO separately. Thus, we would have a total of 8 questions, 4 from the CEO and 4 from the CMO. These questions would aid in guiding us when we are curating our presentation to them.
We have been provided a dataset of the store’s sales information and from this dataset, we have to decide what insights we can garner from it and what information will be relevant to the CEO and CMO. The dataset looks something like this:
This is just a snippet of the dataset, the actual dataset has over 540 thousand rows of data! When anticipating the questions that the CEO and CMO will ask, we will have to look at what are the interests of both of them.
The main difference between these 2 is that the CEO will be interested in mainly the revenue of the company and the sources of revenue while the CMO will be more interested in detailed product and customer insights as this would allow him/her to devise marketing strategies to boost sales. With that in mind, here are the questions that I have come up with for the CEO and CMO.
CEO
- Which are the highest and lowest revenue-generating regions
- What is our overall revenue in the year 2011?
- Are there any seasonality trends in our revenue?
- What is the revenue breakdown by product, customer ID, and country?
CMO
- How many repeat customers do we have, and what percentage of sales do they represent?
- Are there any correlations between UnitPrice and Quantity sold?
- Which products are generating the highest and the lowest revenue?
- For the repeat customers, are they ordering the same product as before or a new one?
As seen from above, these are potential questions that the CEO and CMO can ask and it will be very helpful to anticipate these questions as they can be used as a guide when we are analysing the data and formulating our presentation. Now, we can move on to task 2!
Task 2: Choosing the Right Visuals
For this task, there are a set of questions that will be asked to us and we have to select the best graphs that can help the CEO and CMO to easily visualise and gain insights about their questions from the visualisation. This task is relatively simple as it is just 5 Multiple Choice Questions. Here is one example of the question that you might be asked.
The CEO of the retail store is interested to view the time series of the revenue data for the entire year. The CEO is interested in viewing the seasonal trends and wants to dig deeper into why these trends occur. This analysis will be helpful for the CEO to forecast for the next year. Which visual would most likely help the CEO analyse the data?
Looking at the question, we understand that our visual chosen has to represent the revenue by the months, allowing the CEO to view the time series of the revenue data for the entire year. An obvious visual to choose in this case would be the line chart. Line charts are very useful when dealing with time series data. The time would be represented on the x-axis and the revenue would be on the y-axis. This chart would allow the CEO to understand the important changes in revenue at a glance, gaining insights on the seasonal trends of their store, which will allow him/her to have a better understanding of the store’s revenue and help forecast revenue better. We will now look at task 3 of this programme.
Task 3: Creating Effective Visuals
This is where we start to get our hands dirty and create visualisations for the CEO and CMO! In this task, we are given 4 questions, 2 from the CEO and 2 from the CMO, and we have to come up with a visual to answer each question, meaning that we would require 4 visuals. Here are the questions given to us.
Question 1
The CEO of the retail store is interested to view the time series of the revenue data for the year 2011 only. He would like to view granular data by looking into revenue for each month. The CEO is interested in viewing the seasonal trends and wants to dig deeper into why these trends occur. This analysis will be helpful for the CEO to forecast for the next year.Question 2
The CMO is interested in viewing the top 10 countries which are generating the highest revenue. Additionally, the CMO is also interested in viewing the quantity sold along with the revenue generated. The CMO does not want to have the United Kingdom in this visual.Question 3
The CMO of the online retail store wants to view the information on the top 10 customers by revenue. He is interested in a visual that shows the greatest revenue generating customer at the start and gradually declines to the lower revenue generating customers. The CMO wants to target the higher revenue generating customers and ensure that they remain satisfied with their products.Question 4
The CEO is looking to gain insights on the demand for their products. He wants to look at all countries and see which regions have the greatest demand for their products. Once the CEO gets an idea of the regions that have high demand, he will initiate an expansion strategy which will allow the company to target these areas and generate more business from these regions. He wants to view the entire data on a single view without the need to scroll or hover over the data points to identify the demand. There is no need to show data for the United Kingdom as the CEO is more interested in viewing the countries that have expansion opportunities.
We are also told that there are some errors in the data and that we must clean the data before beginning our analysis.
Data cleanup
Before you can begin the analysis, make sure that the data is cleaned properly. You have noticed that the data contains some returns to the store which are provided in negative quantities and there are unit prices which were input in error. You will need to perform the following steps to clean this data.Create a check that the quantity should not be below 1 unit
Create a check that the Unit price should not be below $0
Please note that in order to apply the checks that have been mentioned above, you would need to use conditional formulas where the logic would state that if the conditions are met then the tool should exclude the data from analysis. You can also use data transformation methods to get rid of the bad data. Both these methods are provided in the resources section. Once this is done, the data will be good to be used for further analysis. Please note that this data should be cleaned up before attempting any question.
Firstly, I opened up the dataset in Tableau and used filtering conditions on the data to filter out errors in the data as mentioned above. Fortunately, Tableau makes it really easy to do this step! Following this, I created the visualisations for each of the 4 questions and you can take a look at it over here! https://public.tableau.com/app/profile/arthur.chong/viz/Forage_16898185433570/Question1
Now we can move on to our last task!
Task 4: Communicating Insights and Analysis
Now that we have curated the visuals to answer the CEO’s and CMO’s questions, it is time to present our findings to them! In this task, we have to develop a script and record a video presentation, presenting our findings to the CEO and CMO based on the questions they have asked. The video should be approximately 5 minutes in length. This task is arguably the most important task as it tests our communication skills. We could have derived amazing insights and created beautiful visuals, but if we do not communicate our findings to the different stakeholders effectively, they would all be useless!
Below is my take on this task!
And that’s the end of my take on this virtual experience programme for TATA! Thanks for reading up to this point!
Connect with me!
LinkedIn
Email: arthurchong01@gmail.com
