dr3amzz / southeast-airline-survey

Southeast-Airline-Survey

Southeast Airlines needed to lower their customer churn (sometimes referred to as customer attrition). Like many airlines, Southeast believed that the best way to minimize customer churn was to have a robust loyalty program for frequent flyers. However, their customers were valuing the loyalty program less, which was one reason why just relying on their loyalty program might not be sufficient in keeping low customer churn. Additionally, customer churn is actually a lagging indicator, meaning the loss has already occurred. As such, it was a measurement of the damage inflicted. The real goal is to reduce churn by getting ahead of the loss of the customer by identifying some leading indicators, or metrics, that might help keep a customer. In other words, these leading indicators, or metrics, could help identify when a customer was about to stop flying Southeast. These insights could provide actionable suggestions as to how to avoid having the customers leave and go to another airline. Out of this reason, Net promote score was developed to measure customers’ degree of satisfaction. The concept of NPS is that customers who are promoters are good customers to keep. Such customers may sometimes even provide free “word of mouth” advertising. Customers who are detractors are really problematic in that they may actively tell their social connections not to use the product or service. Southeast Airlines is one of the top four airlines in the United States. Like the other large airlines in the U.S., customers buying a Southeast plane ticket fly on Southeast Airlines primary routes as well as on Southeast’s regional partner airlines. Regional airlines act as feeder airlines to major airlines by connecting smaller airports to the airline’s main hubs. Hub airports are always located in major cities, whereas the regionals serve smaller cities and rural areas. Like other airlines, Southeast contracted out to regional carriers because it allowed them to lower their risks related to capacity and pricing. Specifically, regional airline contracts last for a number of years, after which, Southeast can renegotiate to adjust (up or down) the number of flights provided by that partner. This enables Southeast to more easily reflect their current market conditions. It is possible, for example, that if demand falls, Southeast would not renew some of their regional contracts. On the other hand, if demand rises, Southeast can expand their contract, and bring more planes into service more quickly than they could on their own. Note that NPS was not currently used as part of Southeast’s partner airline strategy (i.e., it was not part of the data Southeast used to help decide which partners to keep, which partners to drop and which regional airlines should become new partners).

Introduction to data set Southeast often surveyed their customers, and in fact, possessed thousands of recently completed customer surveys. Southeast has been using the surveys to calculate NPS. They would increase their focus on providing good customer service when their NPS score went down. This was typically via a memo to customer facing staff, where they were encouraged to “smile more”. The survey dataset contained thousands of observations of flight segment data collected by Southeast Airlines. In this data set, each row represents one flight segment, by one airline for a specific customer. Each column represents an attribute of that particular flight segment. Each row captures 25 characteristics of the flight (ex. day of month, date, airline, origin and destination city, if the flight was delayed), 7 characteristics of the customer (ex. age, gender, price sensitivity, the person’s frequent flyer status). The row also contains a simple survey-based rating of each customer’s likelihood to recommend the airline that they just flew as well as a field for open-ended text comments. It should be noted that there are some missing values in the dataset. The data set contained information based on 10282 surveys of a specific customer, flights from 206 cities to 196 cities. These flights were from one of fourteen partner companies. Each customer belonged to one of four airline status: blue, silver, gold, or platinum. Each customer was also able to fly in one of three classes: eco, eco plus, or business. The customers had three types of travel: personal, mileage and business. The dataset also contained information pertaining to airline routes, origin and destinations. Additionally, there was information regarding departure delay and arrival delay times. This case will use the R studio to process the dataset and create actionable insights from analysis.