Time to gut airline pricing.

July 1, 2009 8:31 PM

Arline prices: a specific question (Italy, late September, how to minimize cost?). And a general (airline econ 101).

I'm flying from the SF Bay Area to to Milan in late September. How on earth do I minimize my cost on this? Particularly, when do I buy? The fares seem to fluctuate wildly from day to day, with almost a hundred dollar swing in the lowest price in a single day recently. And, based on previous experience, not always upward either -- sometimes they go down, or sales crop up. So I feel like there should be some smarter strategy than just "buy as early as possible."

And, generally: can someone explain to me the economic reasons behind airline pricing? It seems almost impossibly random from day-to-day, airline-to-airline, source-to-source -- far more random than any other competitive industry I can think of. I can understand how if they dealt in extreme commodities like stock, gold, or oil, these fluctuations can happen (and I suppose some of it may be tied to fuel prices). But I can't come up with a theory for this happening in an ordinary service industry.

I'm curious about this in two ways: first, out of general econ-wonk interest. The airline industry seem too competitive for the simple paranoid explanation -- that they randomize prices in order to extract more surplus from consumers -- to work, because randomized prices are basically the same as increased prices and shouldn't be sustainable under heavy competition.

Second, to make buying airline tickets feel less like a gamble! If I wanted to gamble with my money, I'd go to Vegas (preferably not by plane). I don't.

This question may be beyond even mefi... I kind of wonder if anyone has ever just put a crapload of airline pricing data into a stats program with things like calendar date, fuel price, weather, destination, etc., and just thrown regressions at it until it says "ow." If not, perhaps I will one day. But first, I turn to you.

posted by paultopia to travel & transportation (6 comments total)

i don't have any definite answers, but you might find some of the information available on Airfarewatchdog's blog helpful (and even their FAQ section - scroll down to "the best time to buy").
posted by gursky at 9:04 PM on July 1

Kayak.com is your friend. It's really intuitive and allows you to set date ranges and search across dozens of airlines and ticket sites. If you're even slightly flexible on your dates you should be able to cut serious cash out of your ticket prices, and even if you're stuck on a particular date you should be able to find the cheapest possible ticket.

No idea on the economics, sorry.
posted by Happy Dave at 1:25 AM on July 2 [1 favorite has favorites]

1. FlyerTalk is a community of frequent flyers who have attempted to answer just these questions.

2. In your situation, you have a few options.

- look at Air Canada, which is often surprisingly inexpensive for Americans connecting to Europe
- consider flying to another airport in northern/central Italy (Rome, Venice, Pisa) and taking a train
- book two separate tickets with a cushion of time in between...

• San Francisco-Dublin on Aer Lingus, Ryanair to Bergamo/Milan or Aer Lingus to Linate
• San Francisco-JFK on Virgin America or Jetblue, then Meridiana to Rome, then a train
• San Francisco-Los Angeles on any airline, then Air Berlin to Dusseldorf, then another airline to Milan

You probably aren't going to get the best deal booking on one airline on a single ticket. Be aware that while flying carriers registered in the European Union gives you more protection as a consumer, there are stricter weight allowances with baggage than we have in America. If it's just a short trip, you should be fine, but if you're moving there with a lot of stuff, perhaps the single-ticket option is best.

Check out all the options on SkyScanner, be willing to take a while to get there if money is tight, and don't worry about immigration issues - it's totally legal to "enter" a country for the sole purpose of checking in for your next flight or switching airports or whatever, then leave.

Finally, don't feel like an idiot who didn't get THE BEST PRICE POSSIBLE if you choose to have a better connection, not switch airports, or arrive in the center of the city. You can price those options out easily and see what's ultimately the best value for you.

Good luck!
posted by mdonley at 3:54 AM on July 2

Re: Econ - Google Yield Management. That's why the prices fluctuate so much. I asked an AskMe question and got a recommendation for a book on how the math works. I have it on my list to read but have not gotten to it.

To keep it short- There is a ton of math and a ton of data that goes into it - I'm guessing it'll be impossible for you to back into their formulas though because while you can get loads from the FAA you can't get yields. Suffice to say there is a lot of math that goes into the pricing - it is nothing even remotely close to random. The businesses functionally live or die by their ability to maximize their revenues - and given what a terrible business airlines are they have an even harder task.
posted by JPD at 4:19 AM on July 2

The fares seem to fluctuate wildly from day to day, with almost a hundred dollar swing in the lowest price in a single day recently.

That's not much. I've seen +/-500.00 from day to day on $1000 flights. You really have to just keep checking. You can check yourself or using one of the many 'watcher' services such as Kayak, Orbitz, MSN Travel, etc.

Then, the best plan for your sanity:

(1) Decide what a "good" price is yourself, right now.
(2) Check prices every day.
(3) When you find the price you wanted, buy the ticket.
(4) STOP CHECKING, or you'll freak out when a cheaper one shows up.
posted by rokusan at 7:56 AM on July 2

The pricing into Rome is often significantly more competitive than Milan, so you might also find that flying into Rome and take the high speed train to Milan is your best option.

The math of yield management is interesting and complex. The basic idea is that they try to maximize revenue by timing the price levels based on what they know about the buying habits and price sensitivity of customers at various points in time. They also track how many seats they have sold against what their models predict by time period and they will raise prices when they are filling up faster than they expected and lower them when they are filling up more slowly.

Additionally, there is fairly obvious collusion between airlines that ventures very close to the line of price fixing. When one airline is priced even a few dollars less than another, they will capture more business. They gambit with each other frequently where one airline raises their prices and waits a few hours to see what other airlines will match them. If more airlines do so, the new price sticks. If they do not, the original airline drops back to the original price and the new price fails. This happens in both directions.

This guy in MAA Online says "In mathematical terms, the (idealized) problem of finding the cheapest airfare between two given locations is actually unsolvable, and even if you specify the actual route or the flights, the (idealized) problem of finding the lowest fare is NP hard, which means it could take the fastest computers billions of years to solve."
posted by Lame_username at 8:46 AM on July 2