A critical objective of Volantio’s machine learning platform is to understand the key drivers of customer flexibility. What motivates a passenger to accept an offer? How can this be done in the most cost efficient manner?
Data from Volantio’s Yana platform (Figure 1) shows that a clear, and obvious, relationship exists between compensation amounts and acceptance rates - as compensation increases so too does the rate of acceptance.
This dynamic is brought to life every day at airports across the world when airlines are looking for volunteers to give up their seats. The primary lever being pulled is the gross compensation value.
However, Volantio’s data - driven from the Yana platform now deployed with 7 carriers globally - shows that a better predictor of acceptance is a metric we call “Dollars per Delta” (DPD), which is the compensation per hour of time between the original flight and the alternative.
For example, imagine the original flight was supposed to depart at 11 am. Passenger Glenn was offered $600 to take a flight at 5 pm. Passenger Bonny was offered $300 to take a flight at 1:30 pm.
One would assume that by offering 2x the compensation, the acceptance rates would be significantly higher for Glenn vs Bonny. However, the DPD numbers prove more insightful - Glenn’s DPD is $100 per hour ($600 / 6 hours), whereas Bonny’s DPD is $120 per hour ($300 / 2.5 hours).
As Figure 2 demonstrates, results from the Yana platform show that Bonny would actually be more likely to accept her offer, despite the fact that she was offered 50% less than Glenn (in the absolute sense), because her DPD was 20% higher.
Leveraging this data, deployed in an automated, algorithmic fashion, airlines can begin to be significantly more strategic about what they offer to customers in exchange for their flexibility. At a minimum, airlines should vary the dollar values offered as a function of the time duration to the alternative flight.
Finally, it’s important to note that the above dynamics are not the only ones influencing passenger flexibility. How much “advance notice” you give passengers is also very important, for example. In addition, results will differ from airline to airline, and will change over time.