Sometimes the lowest fare really isn’t, but why?


I honestly wish I knew the answer to this question. Sometimes I think I have it figured out and then another ridiculous situation comes along and shatters my world view. I mostly understand how fare rules, inventory allotments and various other restrictions interact in the GDS networks to produce a fare calculation. And I believe that those systems are set to show lowest fares based on the parameters you input. So why don’t they always work? I really have no idea. But it is definitely broken.

Trips to Istanbul have been all the rage lately thanks to very good fares in the market. And this past weekend a friend was looking at flying over from Cleveland for a long weekend in January; the $838 fare offered some motivation. And then, suddenly, the fare went up to $1,044. Normally I’d say this is just inventory buckets shifting or a fare disappearing (in fact, I did suggest that initially) but something piqued my curiosity and I started digging more. Running a search on ITA gave these results:

image

Not surprisingly, all the OTAs and the airline website showed a similar price. But what if we ask the system for a lower fare? ITA allows you to set all sorts of details in the search parameters so rather than just asking for United Airlines flights I specified the lower fare class which had been showing the cheaper fare earlier:

image

Much to my surprise, it comes back with the exact same flights for $200 less:

image

Part of me got excited, thinking that this could be leveraged into a best rate guarantee since certainly one of the various OTAs would properly price the cheaper fare, right? Alas, I cannot find a single one which does. I obviously haven’t checked them all, but the fact that the ITA engine isn’t automatically pricing it properly suggests that there is something bigger afoot than just one company messing things up.

This is most definitely a bizarre situation and, unfortunately, seems to have no basis in reason. Makes being an airline customer all that much more challenging. And since United pulled inventory data from their website this weekend it also means that double-checking a fare like this is even harder. After all, how would I know that there was K inventory available if I hadn’t previously known about the cheaper K fare in the market?

This is not good for consumers. Not at all.

Related Posts:

Never miss another post: Sign up for email alerts and get only the content you want direct to your inbox.


Seth Miller

I'm Seth, also known as the Wandering Aramean. I was bit by the travel bug 30 years ago and there's no sign of a cure. I fly ~200,000 miles annually; these are my stories. You can connect with me on Twitter, Facebook, and LinkedIn.

14 Comments

  1. I don’t have a good answer for why this happens, but I have also observed that ITA and other fare search engines are pretty bad at finding the lowest fares in each market, especially for international itineraries. If you don’t have GDS access (Sabre is a lifesaver for me now), I’ve found that the best strategy is to look up the cheapest fare in the market, then do a search for that specific booking class … lots of great options tend to turn up out of nowhere.

  2. Let me tell you what’s going on here. The T bucket is actually the cheapest one. I looked up on ExpertFlyer and TKWNC16S/TKXNC16S are cheaper than the KLLNC5N. Turns out the T bucket has a $500 fuel surcharge, while K has $250, so there’s your price discrepancy. So Matrix is technically finding the lowest base fare, but why the heck is the YQ so different? And is there a practical way to force a K reservation, like calling UA directly?

    1. Interesting detective work on the difference between the base fare and the total fare. Not sure I’ve ever seen a comparable situation but it makes sense, I suppose. Sortof. Still seems wrong that the system cannot figure out the total price properly when deciding what to display rather than going on base fare and then assembling the final details later.

      As for forcing the K fare, it is possible on United. On their Advanced Search page you can specify a particular fare bucket. It works with this one in my testing.

  3. I checked a few other fare buckets on Matrix and they all have the 500-550 surcharge, so maybe K was a data mistake that they swept under the rug for now. It gives me new ideas for how to track down lower fares though. If the lowest fare bucket on Matrix doesn’t match what ExpertFlyer shows, there might be something going on.

  4. Just a side effect of the algorithm used to give you those results in a short amount of time. Every single website will have something like this (albeit hopefully minor) – you probably won’t figure out why without understanding more about ITA’s architecture and code.

  5. You can book (I at least tried dummy booking without purchase) this K fare on United by using Advanced Search and putting “K” into “Enter specific classes of service”. It prices $837.90.

  6. Great detective work! I too don’t know the ultimate answer(s). I suspect… While UA SEEMS to want our business, it is in their interest to direct the self-booker to a higher cost fare bucket. IMO, restricting the bucket and availability information is akin to shooting themselves in the foot, especially for the Q&D trip that you described. In this circumstance, the $206 price difference could easily become a deal-breaker for this PAX. Does UA want to fill the seat at $838 or complete the scheduled flight with $838 less gross revenue? If UA and the others really want us to do our own bookings via their websites, I believe that they are obligated to make ALL available buckets available and without such a fight for the data. In a similar situation, and with traditional travel agent’s but a part of history, I’d rather employ a client-paid travel assistant (such as Crankly Flyer and many others) for ~~$150 or so, to find and secure that $838 all-in fare, pocket only $56 difference and know that I had a knowlegable advocate in my pocket. Again IMO, and in the example cited, UA is simply being dishonest – and I do not like that. Ten more demerits for UA. Too many more shots into their left foot and they may loose a lot of ‘legs.’ I hope someone at UA follows these blogs. If they did, they might fill a few more seats.

    1. I don’t think this one is on United so much as on ITA and the other GDS vendors and the way they’re constructing the fares. Still, it is a reminder of just how hard it is as a consumer to aggregate and process large amounts of data associated with airline fares. Winning the game seems harder than ever these days.

  7. First of all, if you really want the cheapest fare, ITA shows an AA+Turkish combo for under $670 so why even bother with United?

    This is most likely the case of a mis-coded mistake fare that they turned off until it’s fixed. Otherwise, the T fare should still show up in the results, regardless of how it’s interpreted and sorted.

  8. Seems like there’s always squirrelly stuff going on with YQ to Istanbul.

    Just booked a nice $504ai fare out of ORD for nov. Try to book it with the necessary positioning flights on the same pnr and the yq doubles.

  9. This is not limited to just ITA’s pricing logic. I checked the same in SABRE.

    I booked the itinerary from the start in K and the result is $837.90.

    Do a bargain finder WPNC entry on the same itinerary and it says T class is best at $1083.90.

    So the same result as ITA. SABRE is checking only the base fare and not checking the overall total cost.

    The challenge here is one needs to go to a tariff to determine what the best fare is in the market and the difficulty is most people do not access an airlines complete tariff.

    Had I simply booked this itinerary in Y class and done the WPNC in SABRE, I would have gotten $1087.90. Only by looking at the tariff would I have been informed of the K fare. Unfortunately, not many agents refer to tariffs anymore.

  10. I am the friend referred to in this blog post. Thanks to Seth’s detective work and advice, I plugged a “K” into the fare field and voila! I’m ticketed to IST. Thanks, Seth!

Comments are closed.