When the results achieved by a business are poor or less than expected, should the strategy be changed? It is a question every entity wrestles with at some time, and typically the worse the results, the more there is a demand to change the business strategy or the pricing strategy. Sometimes changing the strategy is the right answer, but very often poor results are a result of poor execution. In those cases, the business should stick with their well-thought-out strategy, but focus their efforts on improving the implementation.
By now, everyone has seen the video or heard about a passenger being dragged off a United Airlines flight. The removal was reportedly due to a common airline practice of overbooking the flight, however when an insufficient number of passengers accepted the offer to voluntarily change, the man was told he needed to leave the flight. When he refused, he was forcibly removed. Predictably, there have been calls for airlines to change their pricing strategy to exclude overbooking, which keeps fares somewhat lower than they otherwise might be. There have even been calls for new laws to prevent overbooking.
In my view, eliminating overbooking would be foolish. The airlines have done an excellent job over the years of learning how their customers behave and adapting to those behaviors. One of the lessons has been that a small percentage of customers will not show up for their flights. When those customers do not show up, and the plane flies with empty seats, the airline generates less revenue and little or no profit. Although competition ultimately determines the prices airlines can charge, when they do not use overbooking in their pricing strategies, all airlines have an incentive to raise prices enough to offset the lost revenue from empty seats.
With the overbooking component of the pricing strategy, when too many passengers arrive for a flight, the airline offers an incentive to any customer willing to take a later flight. If an insufficient number of passengers has accepted within a certain period, the incentive offer is increased. This process continues with increasingly larger incentives until enough passengers accept. In the case of United, their poor implementation included at least two fundamental problems: 1) they put a cap on incentives offered to customers, and 2) they didn’t use the timing principles of their revenue optimization models.
All the airlines have statistics on the number of no-shows and the acceptance rates of offers to take later flights. They also know how the acceptance rates are affected when the next flight is not the same day. That means they have a good understanding of how much it could cost them to bump passengers when they do not have any no shows. However, those average and predicted acceptance rates reflect a distribution, not a single point. So sometimes the voluntary bumping fee will be small, other times it will be large, and sometimes it will be very large. By setting a cap on the bumping fee, United did not allow for the rare occasion the fee would be very large, and they did not receive enough acceptances. Had they continued to increase the offer, they would have eventually reached an equilibrium point. It might have been expensive, but United could have then updated their revenue optimization models to account for such possibilities.
United’s second fundamental error was not paying attention to timing. Their study of customer behavior has shown that customer price-sensitivity changes as they get closer to their departure date. As seats are sold, the probability of the plane selling out increases and fares increase. Passengers who are late in shopping for fares invariably pay higher fares. United could have applied that principle to overbooking. As departure times get closer, customers view taking a later flight as increasingly inconvenient. Once the customers were in their seats, that inconvenience factor would have increased substantially. United should have realized that by making the decision to bump four passengers for a flight crew so close to departure time, their price for bumping would go up substantially. The airline should have made significant increases in their offers before all the passengers were on board.
I have focused this post on United and airline overbooking, but the point applies across industries. Results can vary substantially from expectations for many different reasons. Sometimes the reason is a bad strategy, but often it is simply poor execution. When you experience poor results, by all means evaluate your strategy in the context of where you compete and how you can win. But if you are just not executing well, fix that before coming up with a new strategy.
Leave a Reply
Want to join the discussion?Feel free to contribute!