Tracking Airfare Prices

Air fares to Europe are up significantly this year, as I recently discovered when my girlfriend and I started planning a trip to France to see relatives. To try to get a sense of whether there were any deals to be had, I started manually checking prices every day and tracking them in a Google Docs spreadsheet.

I did this for about two weeks. It was laborious, but it was fascinating in that it let me decrypt just a little bit of the arcane logic that goes into the fluctuation of ticket prices. There’s not a tremendous amount of pattern recognition that you can glean from a sample size as small as fourteen days, but the airline industry’s pricing models and schedules are so opaque and inscrutable that even seeing real prices tracked over a short amount of time — watching how they rise and fall — is instructive.

Of course, I realized too late that there are probably Web tools that can automate this kind of search for me. I hunted around a bit and found Yapta, which I’d never heard of before but does more or less what I’m looking for. Yapta layers a tracking service on top of a Kayak-powered booking engine. That’s fortunate because Kayak is my preferred travel booking site and so the search methods were therefore very familiar to me. Yapta returns what are essentially Kayak-flavored results, and you can click on any itinerary to start tracking its fare fluctuations in your Yapta account. The data presentation is rather lackluster in that there’s no graphical charting of pricing trends, and if you’re tracking multiple itineraries, you have to click through to each to see its full pricing history. But Yapta does automatically update and record the prices, saving me a lot of manual labor.

There may be other, more powerful tools out there that do this same thing as well or better than Yapta. (If you know of any, I would welcome tips.) I’m not sure if this is something that a lot of people know that they need, but it would certainly seem to be something a lot of people would use if it were packaged elegantly and if it were better integrated into the booking process.

More importantly, though, what this brought to mind for me was how lopsided the data collection dynamic is in ticket pricing. Over the course of the several weeks when I was actively checking prices, looking for a deal, I was handing over a nontrivial amount of behavioral data to the booking sites and airlines I was patronizing — not just where and when I want to go, but also my preferred carriers, routes, price tolerance and more. Clearly, that would be more than enough data to gouge a customer if a company wanted to, though I’m not accusing these sites of doing that. I’m just saying that the early promise of online travel was that it would allow for pricing transparency, that increasingly sophisticated booking tools would let consumers find the best possible deals. As it turns out, the situation we have today is that those who set the prices know more about how we buy tickets than we know about how they price tickets. That doesn’t seem very empowering to me.

  1. I use Farecast for this purpose. They claim to follow billions of price fluctuations in order to build a model to predict future price changes. Farecast was acquired by Bing and is now featured as a part of Bing’s travel search (do a flight search flight and look for the fare prediction link).

    As to the data collection, I asked about this on Quora a little while back. Got some insightful answers but no inside info giving a definite answer I was hoping for: link. I agree that the data processing demands are significant, but I still assume that some demand signaling must be happening.

  2. I tried using Farecast via Bing, but I think there wasn’t enough data for the particular itinerary I was interested in (it’s a multi-city route) for it to give me any predictions. What I like about the Yapta approach is that it’s tracking real data. But of course the disadvantage is that it only starts tracking as soon as I start asking it to do so; until then there’s no historical data available. That’s a big drawback.

  3. I like to check for good days to travel with Matrix Airfare Search (

    You can’t book there, but you can specify that you want to leave on any day in a range of a month (say anytime from tomorrow until a month later) and for a range of days (say 7-9 days) and it’ll show you a calendar with the cheapest fare on each day. Then you can drill down and see the available connections (across many (all?) airlines) etc.

    You’ll need to go somewhere else to do the actual booking, though.

  4. Hi Khoi, one option I use frequently is, which started as a hobby project by a couple of guys from Scotland and has grown and grown in popularity, beyond recognition. I have sourced and got some amazing deals through this site (ie London-Milan GBP 14.00; London-Barcelona GBP 22.00 and lots more).
    Hope this is of help.
    PS: Not aware of price tracking function.
    Best Fernando Sertage

  5. Having worked for a few OTAs I can assure you price targeting based on an individual’s searching behavior is WAY outside the realm of possibility even if they wanted to.

    It’s illegal, it would be very easy to prove, and at the end of the day they don’t care. The are mostly interested in selling you a car or a hotel with that flight, as the flight is the loss leader.

    The web has given you transparency… into the spaghetti bowl/Escher drawing of years of arcane airline pricing strategy.

Thank you! Your remarks have been sent to Khoi.