For today’s dashboard challenge, we were tasked with creating a dashboard from US airline data. This was the same data that the DSAU20 cohort used for their 2nd interviews except we had one day to complete the task.

Data/Story Preparation

As it was interview data, the data was rather clean except for a few duplicate data which was cleaned using Alteryx. The premise of my story comes from the fact that I have heard multiple anecdotes of US airline companies overselling tickets and kicking passengers off. This was very prevalent in the bigger companies such as Delta Airlines and United Air. In April 2017, an infamous incident occurred when a doctor was physically harmed by United Airlines and kicked off at random due to oversold tickets. This cost the company 1.2 billion dollars in market value.

My story was to analyse the big 3 companies and see if I can get any insight into how much flights get oversold as well as other metrics. I also supplemented the data with Twitter data to see if there were any patterns in negative sentiment and compared the topics between each airline.

Tableau Process

This was one of the most difficult tasks as I felt that nothing clicked when exploring the data. Everything felt very shallow in terms of insight as I would find random outliers in the timeline data that could quickly be explained by a news event such as covid. I found it difficult to go that one step further due to the time constraint. In the end, I decided to use this day’s dashboard to practice my set actions which drilled down the timeline to other metrics such as load factor, number of destinations, number of departure airports etc.

I also realised that the data couldn’t support my initial aim of finding overselling and decided to shift my direction as it was too late to change the story. As a result, this made my twitter data irrelevant and so I decided to omit that as well.

Final Dashboard and what I learnt

My final dashboard is presented below.

I’ve learnt from this and yesterday’s dashboard that I need to improve on delving deeper into the data to give my story more oomph. I remembered that this was a recurring question that came up during the data schoolers meet & greet and I think it will come with time and experience.

Thank you : )

Nam Nguyen
Author: Nam Nguyen