In this work, we will present a revenue management model based on a Nested Multinomial Logit customer choice model that does not destroy ancillaries value. We will introduce a two stage booking process where we take into account two different times, in the first one, a customer can buy a bundle, and in the latter one, a customer can add some ancillaries, that were not originally included.
Lastly, we will show how the assortment problem can be easily solved, without having to resolve it for each pricing problem.

Preliminary results using synthetic data, show a significant growth in revenue when using our model. We can see that offering ancillaries near the departure is crucial to increase revenue for the airline.