Published on January 2, 2019
During the semester, Brenda and I had the privileged opportunity to collaborate with American Express (AMEX) to create a Tableau dashboard. The main focus of this collaboration was to present their profits and losses in a comprehensive, insightful and interactive manner using Tableau.
Knowing the Credit Card Business
Prior to getting our hands dirty on engineering the dashboard, we first had to gain an understanding of AMEX’s business model. Knowing how AMEX operates was imperative for us to understand the dataset better and it allowed us to generate better insights from the data.
To give a brief overview, AMEX’s revenue comes mostly from their discount revenue which is essentially the amount charged at the merchant when their customer uses an AMEX card for payment.
Populating of dataset
Due to confidentiality reasons, AMEX did not allow us to use their raw data for this collaboration. As such we had to create our own dataset by populating certain columns with random numbers and deriving the other columns via calculation from the random numbers generated.
To provide an illustration:
|Billed Business||Discount Revenue|
|=RANDBETWEEN (500,10000)||= 2% * Billed Business|
As cumbersome as it might have been, this process gave us an even better understanding of how the business model works.
The objective of the dashboard was to give the CFO a quick snapshot of a country’s finances before he/she flies over to the country for a meeting. With this objective in mind, we began to brainstorm about the various charts we could use to show key insights that a CFO would want to know.
Initially, what we came up with were simple charts which consisted of diagrams such as profitability across the years and the attrition rate of cards across different countries.
While there was nothing wrong with such charts, they were nothing extraordinary and most importantly they do not harness what Tableau has to offer. As such, with the guidance of the finance manager of AMEX, we began to look for better ways to present the data and thought about how we could combine columns of data together into one chart to generate better insights.
An example of this would be charting the top 10 AMEX partners in terms of volume of Billed Business together with their Profits as a % of Total Profits. This aims to find out if a higher volume of Billed Business translates to higher Profits generated by a partner.
Lo and behold, it does not hold true.
After 3 months of collaboration, we were proud to present our finished product:
Key Learning Points
- Understand the data well.
- Know the objectives of what you want to present.
- Using the charts, aim to paint a story which achieves the objectives set.
- Dig deep into the data and uncover ways to generate more insights from it.
Thank you American Express for the wonderful opportunity!