Analyse The Airlines Passengers’ Satisfaction

1 Project Charter.

Problem statement:

Airline industry is getting more competitive due to the disruption of budget airlines, which allow customers to travel to their destination at a much cheaper price though with less service quality. If the customer experience at Falcon Airlines is predominantly negative, it will quickly become detrimental to the company’s profit due to high capital expenditure and maintenance cost.

Business case:

Knowing the customers’ preference and feedback allows Falcon Airlines to improve their standards and services on areas that matter to the customers, and strategize the emotional selling points for their marketing campaign to stand out among their competitors, hence maintaining the company’s competitiveness and customer loyalty.

Project goal:

Predict whether a passenger will be satisfied or not given the rest of the details are provided and identify which variables/features play an important role in swaying a passenger feedback towards ‘satisfied’.

Project scope:

Perform data mining on the entire journey of the customers experience beginning from website experience, online services, and departure to onboard experience, flight duration, and arrival.

2 Presentation.

Check out the notebook or PDF slides below. The PDF is viewable in desktop if the mobile doesn’t display it properly.

3 Report.

Check out the notebook or PDF slides below. The PDF is viewable in desktop if the mobile doesn’t display it properly.

4 About the dataset.

There are 2 datasets:

  1. The Flight data has information related to passengers and the performance of flights in which they travelled.
  2. The Survey data is the aggregated data of surveys collected post service experience.

Assumptions about the survey:

  1. Integrity: We assume customers are providing their honest and careful feedback for the survey.
  2. Situation & context: We assume the customers aren’t hurry to complete the survey after arriving at the destination and experiencing the service since most customers (about 69%) are business travellers as they could be busy with work. This would result in a bias outcome.
  3. Question & feedback: We assume those survey questions are what the customers care about as the customers can’t fill in any feedback other than choosing on a 1–5 scale based on those fixed questions. The outcome will not be holistic if the company assumes these are the only questions that matter to the customers. Additional research has to be done to obtain a more accurate picture of our customers’ preference and feedback.

Data limitation:

  1. The variables/features from the survey data will not be suitable to train the model to predict the customers’ satisfaction, because we will have to constantly gather the same survey data from new customers to make future prediction which doesn’t make sense since we could directly ask them to indicate their satisfaction level. However, the goal of this project is to identify the important variables/features that can significantly improve customers’ satisfaction. Hence, we are using the model to identify important variables/features based on their effect on the satisfaction outcome.
  2. Survey is not perfect method for gathering customers’ preference and feedback, hence additional research need to be done to obtain a more holistic picture.

5 Conclusion.

The following recommendations focus on loyal customers who are neutral/dissatisfied:

  1. Extend the user research to interviews with different segment of customers after performing the customer segmentation, especially with personal travellers to find out more on customer preference. We can also inquire them about the top N important features to understand why they matters to them. We can also find out what makes a customer loyal to our FalconAirlines brand so that we can also target the disloyal customers who are neutral/dissatisfied.
  2. Extend the user research to value mapping to map out and analyse the customer touchpoint and user journey.
  3. Analyse the current business capabilities for the top N important features before moving to the business transformation planning. We can then estimate the potential reduction in customer churn cost for features that the business is capable to plan and execute for that time period.
  4. We might want to pay extra attention to loyal business travellers since the customer churn cost is higher as they travel more frequently than the personal travellers. Plus, about 69% of our customers are business travellers based on the current data.