Our AI-Powered Customer Scoring Model's Assistance in Predicting 60% of Ticket Cancellations

Client Challenges

• Since the client has included all the major tourist destinations in the world in their offerings, it had to manage a huge number of bookings.
• Along with its growing footprints in multiple geographies, the number of bookings also saw exponential growth.
• Though bigger order values boost the profit considerably, they also bring with them order management problems.
• With increasing bookings, the client also saw an increased number of planned or last-minute cancellations. This included both individual and group bookings.
• Considering that it’s very difficult to transfer the last minute canceled booking to other customers, especially of the long tours, those bookings went unsold. Consequently, it affected the bottom line of the client.
• To avoid the revenue leakage, the client was in need of a reliable, easy-to-use and flexible model which can identify the customers who are likely to cancel the tickets.
• Its absence robs the client of data-backed prediction of high-risk ticket cancellations and thereby the accurate demand forecasting.

Our Solutions

In the preliminary examination of our client’s problem, we found that they had no reference data to predict the cancellation. They didn’t gather the analytics insights like what people with which age group, demographics, and income group have canceled the tickets more in the past. Also, they were without any analysis about which season and which destination saw more cancellations. Resultantly, this made it impossible for the client to create the “most-probable-cancellation-scenario”.

To overcome it, we began with data preparation and data analysis and created a phased customer scoring model including Profiling, Modeling, and Measurement which were tailed by data analysis phase.