My husband recently flew to Las Vegas to see a Nascar race. On arriving at the airport, he entered a scene of mayhem. The airline, Air Canada, had overbooked the flight by something like 25%. Desk agents were begging for people to take a later flight, but everybody wanted a seat on that plane.
Obviously, the Air Canada statistical models had predicted that on your average flight from my hometown to Las Vegas at the end of February, there would be x% cancellations, y% no-show, and z% late arrivals that could be shunted to later flights. What the Air Canada statistical model did not anticipate, was that 160,000 people had purchased tickets to a Nascar race, and had to be in Vegas in the next 48 hours; when you pay that kind of money for that kind of event, you don’t cancel. You don’t miss your flight. You show up, ticket in hand, having been promised a ride to go see the pretty cars.
What are the chances that similar scenes played out in other airports across Canada that day? Pretty good, I bet. I imagine that most of the people who flew to Vegas to see Nascar via Air Canada were put into a pretty foul mood, even the ones who eventually got on the original plane. A dollar sum could probably be calculated as to the losses in both revenue and goodwill that Air Canada experienced as a result of their inability to accurately predict the number of people who could be promised a ticket.
I won’t give you a song and dance about how a flexible identity situation could have fixed this problem – because the truth is, you have to have a flexible company first. There just aren’t too many of those out there. Still, success is more and more a function of information. You can’t just gather information, you have to distribute and process it too, and do so in a timely fashion. What if you were to ask people at registration time whether they were attending an event. Suddenly, you would have the information necessary to change your statistical model – assuming you could take that information from your reservation system and mash it up with a calendar of known large events. What if someone or something was given the capability to arrive at the conclusion that an event of this size would affect attendance percentages for a certain range of dates with respect to flights to the event location.
Pipe dream right? Current companies are just not equipped. But new companies will be. Future competitors will have less legacy baggage, and a digital native’s understanding of the value of treating every internal system like the fodder for the worlds best corporate mashup. If they can build the flexibility in from the start — just think of how far they can go. And just think how tough it will be for the dinosaurs to compete.