bchandler02

Found a thread from a few years back, but nothing recent on this topic.

Anyone else seeing ITA prices that are substantially lower than aa.com? For example, I've looked at OKC/SJU for 3-5 nights in October/November on ITA and prices have ranged anywhere from 380 to 600+ depending on what day, when I search, etc. AA.com is showing 511 for almost every day with a few rare exceptions.

Anyone else experiencing this? The last few times I've used ITA prior to this, it was always pretty accurate.

brp

Are the fare classes shown by ITA the same as the ones offered by AA? There could be a published fare for, say, O fare at the lower rate. But that doesn't mean that AA are offering any availability in that fare. Does the $511 fare class match any of the other classes shown by ITA? (I'm not familiar with ITA to know if they show the pricing for all the fare classes, but I know that EF does, so they may.)

Cheers.

lbbzman

Are you getting the same routing on ITA and AA.com? When I just did a quick search on ITA for OKC-SJU-OKC, departing 24 Sep and returning 28 Sep, I found a $380 constructed fare. But the routing is a dreadful OKC-DFW-MCO-MIA-SJU, SJU-MIA-MCO-DFW-OKC monstrosity. And then if you look at the fare details, it shows:

As you stated, when I go to AA.com, I'm presented with $511 fares, more in line with the fares shown on EF.

Cheers,
LBBZman

Edited to add: You may be able to get the ITA fare to show on AA.com by doing a multi-city search. I haven't tried, but it could be worth a shot.

Madison Guy

I used your dates on Multicity and got two $380.31 options using the okc-mco,mco-sju and sju-mco,mco-okc as the search criteria.

nbevan

Whereas for Oct 14-17 ITA gave me a fare of $469 for:

OKC-SJU QNE4C2O1
SJU-ORL ONE5B4Z1
ORL-OKC OD07ERR1

While aa.com multicity for the 3 segments gave me a fare of $511 for the same flights:

OKC-SJU QKE0B1O1
SJU-ORL QKE0B1O1
ORL-OKC QKE0B1O1

Although EF does shows O availability for the individual segments, it does not show O availability for SJU-ORL or ORL-OKC, so ITA seems to have got its married segment logic wrong?

hillrider

No, you got married availability wrong; it does not cut across booking classes. So in EF you'd be looking for OKCSJU in Q, plus SJUOKC in O.

In either cases, both ITA and aa.com use cached data; EF and actual bookings made by AA (on the phone and by AA.com) show/use real-time data, so they may not match, but if it's not in EF, it's not bookable (AA.com will fail when clicking reserve or whatever).

hillrider

deleted

hillrider

OK, so I ran the ITA scenario and it's a 4-segment monstrosity.

You will not get a 4-segment connection on EF for SJUOKC; GDSs (at least Sabre, where AA is hosted) won't give you those solutions, it will give you only a max of 3 connections.

Also turns out that married segment logic only applies to a maximum of 3 segments. So in our 4-segment construction you have the first 3, or the last 3, drawing from married availability, while the additional segment draws from that specific market's availability. However, I was unable to make any permutation work due to lack of availability in O on one of the required instances.

I was unable to make any permutation work. So maybe ITA's application of married logic on >3 segment itineraries may be a bit bunk, or it's just drawing from stale data.

This is what ITA found and I tried to book:

bchandler02

Thanks.. I see where everyone is going with this in regards to pairing and fare codes. Got real busy at work today so I'll have to play with this more later tonight, now I know what to look for.


Appreciate everyone's help.

ITA Hacker

I did a successful test booking in Sabre by booking the first return flight (AA1556) on its own in O, then booking the last three flights (AA1678 AA2290 AA2274) also in O class. Our search engine knows how this works and generates an optimal answer.

(the last three flights together only have one seat available in O)

nbevan

(EF is not clever enough to locate the AA1678 AA2290 AA2274 combination for MIA-MCO-DFW-OKC), but if that is the way ITA prices it, why does ITA show the fares as SJU-ORL and ORL-OKC, rather than SJU-MIA and MIA-OKC?

hillrider

Aha! It was probably not there when I tried to do the same (it's easy for space to come and go when it's so limited, even more so when it depends on the fluctuation in availability of three flights).

I see the availability now and can confirm that the ITA solution is bookable.

Glad to see that ITA hasn't lost its edge (kudos guys, and thanks for the post/support)!

hillrider

Pricing (faring) and married segment availability are completely different logics and don't overlap. Once you have the space (married segment availability or whatever), you can apply whichever fares to the segments independently on how they got there; origin & destinations don't come into the picture.

That's the way it works, logical or not, and its based on outgrowth from legacy system as opposed to greenfield thinking/system design.

ITA is probably the only engine who can proactively exploit this. Those guys are, in a word, AMAZING.

dstan

ITA takes inventory availability into account, although occasionally it's not perfectly in sync with AA.com. Much different than EF in any case - ITA is much more useful for flight searches - you can rank order itineraries by price (among other options).

nbevan

Thanks for the clarification!

Orbitz (but not Expedia) also finds the fare without difficulty, but clearly aa.com is not as sophisticated.