I want the inside on how the commercial air industry works!

October 22, 2012 4:23 PM

How is the market for a commercial air route determined?

I'm curious, how does an airline determine their routes? I flew from Tampa to Los Angeles in June, a non-stop flight at just over 5 hours and is only offered once a day in both directions, and it got me wondering.

What's the market for that? My flight was pretty full. From the business side of things, how do they determine their destinations? Obviously not every airline flies everywhere in the US, and often times only 1-2 airlines will get you to where you want to go without any connections.

I'm just interested in knowing how the commercial air industry works in this regard.

posted by signondiego to grab bag (5 answers total)

I never did air traffic routes, but when I was a A&D strategy consultant, we used to have to figure stuff like this out. It's not hard to get a general idea of demand. If 100 people fly from City A to City B, and 80 of those people then fly on to City C, then you can pretty easily assume there's a market for flights from City A to City C. Once you have that, you have to look at why people are travelling, what's driving that demand (and what would limit it), and you create a model from there. And you're frequently wrong, and have to iterate over and over to get it better. (That doesn't mean you're wrong, it's just how things like this work.)

These are fun problems to tackle.
posted by NotMyselfRightNow at 4:46 PM on October 22, 2012

I used to work for an airline and they had very sophisticated software that would analyse this information and a whole lot more. Each route, its historical usage (split up between the various levels of frequent flyers, corporate customers etc), which agents booked the flights, which agents were booking more or less of a certain route ... the sheer amount of info was mindboggling and the guys working on these systems were highly paid and knew statistics and how they could be applied to this kind of thing back to front.

For each route they don't just analyse that route but connecting ones... does that plane, and flight crew come straight back? Go on another route? How much does it cost, would it be cheaper to go another way? How much food do the people generally consume on each flight? What type of food? On which route do people buy more services? Which people are they (frequent flyers/first class/business class etc)?

Basically this is bread and butter for the airline industry and they have very detailed stats on each flight, route, agent, customer, passenger type and a whole bunch of other things.
posted by Admira at 6:03 PM on October 22, 2012

You may want to look on the Airliners.net "commercial and civil aviation" board for more discussion of this - often someone asks, "hey, why can't we make a [podunk airport] to [podunk airport] route work?" and a bunch of people in the industry comment with really interesting answers.

It was there that I learned about Lufthansa's Frankfurt-Shenyang route - Shenyang's only European flight! - which largely exists to serve the demand of German companies working with the BMW plant and other industrial firms doing business there.
posted by mdonley at 7:57 PM on October 22, 2012 [3 favorites]

In my non-internet life I'm an academic working in transport modelling. Short answer (as others have said): it's very complicated. You can get a rough idea of how many people will want to fly from A to B by considering the population and income levels in those cities, but there are many other factors as well. For the airline the ultimate goal is to maximise their profits, or at least minimise their losses, by using as little fuel and as few staff-hours as possible to transport as many paying customers as possible. So (for example) they may calculate whether a direct route between A and B would cost them less than a flight from A to the airline's hub city, and then another flight from the hub to B, bearing in mind that passengers from A to B will like the second option less, but passengers going from A to C and C's other destinations can also be accommodated. They also want to minimise the number of aircraft they need, and use them for as much of the day as possible. In the second case two planes are needed, but because they're going to and from the hub they can more easily be used on other routes as well.

Airlines will also be considering what they are competing against (either other airlines or other modes of transportation). This may mean only one or two airlines serving a route. One bigger aircraft a day is generally cheaper for the airlines than multiple smaller ones (e.g. less fuel is needed), so they will try and do that if they can. On the other side, for some major city-pairs in the US the airlines compete on frequency, typically where there are lots of business travellers, so multiple airlines will have frequent flights between those cities.

There are also a lot of calculations made around ticket price, as any one who's ever spent time on ticket booking websites will be familiar with. Generally prices are manipulated over time to try and get each flight as full as possible. This all interacts with the other calculations, and with airline business models. So you end up with an enormous optimisation problem.
posted by gnimmel at 1:47 AM on October 23, 2012 [1 favorite]

In fact taken from a post on the airliners.net forum that mdonley linked to, a WSJ article on The World's Oddest Air Routes and the reasoning behind them.
posted by jontyjago at 8:39 AM on October 23, 2012