Hold on to traffic in a legacy CO fortress? That does not seem to make sense... From IAH, who else are you going to fly?

I would think if anything, they would have offered this to HVF's in more competitive hubs...

But, who says logic (at least my logic) always prevails...

interestingly my last 4 of my last 5 CPUs (and 4/6 for the year) have been on UAX:

SAV-ORD (CR7)
RDU-IAD (Q400)
SMF-LAX (CR7)
LGA-ORD (E170)

sort of going against the mantra of bigger cabin = better chances.

Well, if you're sick enough of the captive hub prices, didn't like the merger w/PMUA, had a crap 2012 (like many of us did), have a tiny bit of flexibility to route through that other major TX hub, etc - why not?

UX flights typically have been among the easier upgrades, from the E500 days through UDU and now into CPU. The percentage of F seats can be fairly high (the Q400 and CR7 both have a higher percentage of F than do the A319 and A320, and on par with the 737-700). They also are often found on routes which aren't as elite-heavy, which helps.

Historically, I would agree with this statement. This year, however, things have gotten interesting. My home airport is in the midwest. It is a short hop to either ORD or DEN, which I typically use to get to either the east or west coast.

Up until this year, my CPU success rate was always higher on the short hops to and from home. It makes sense--not a lot of value in buying an upgrade, lots of UX with its higher ratio of first, and relatively elite light loads.

This year my upgrade percentage is down overall (around 70% over all UA flights, down from around 90% two years ago), but it is now lower to and from my home airport than it is for my flights from hubs to the coast. Why? I'm not sure, but I would point to a big decrease in overall capacity and many more no-first regionals, which I think tends to concentrate the HVFs on the planes that do have first.

How about COS-IAD?

What does elite percentage have to do with picking the right plane for the load factor on that route? It shouldn't matter if the plane is has 0 elites of 100% elites -- the plane chosen should be based purely on the number of pax on the route.

Well, I didn't get a CPU on SFO-LAX on a Tuesday in February. I did just get a CPU on LAX-SFO tomorrow....

That being said... no, CPU's are not improving for me. They are not getting worse though.

A bulk import tool would be nice indeed, although I am wondering if it wouldn't add noise to the data: besides the "when", you would also lose the "how" (instruments).

Alternatively, we could integrate with the TripIt API, similar to what OpenFlights does.

How do people feel about the two options?

I don't normally update TripIt with my upgrade information, but I do update OpenFlights with cabin info after importing from TripIt (at the end of my trip). Perhaps an integration with OpenFlights instead?

Ah, yes, that's even easier since the export format is documented (http://openflights.org/help/csv.html). It would be a manual process though (a sync once in a while would be required).

If you feel comfortable telling, are you in DSM by chance?

I've been contemplating adding a proper REST API to OpenFlights, but never found an application that would use it for sure. If there's interest, that's something I could look into.

DCA-IAH-SJU-IAH-DCA all 4 upgraded this past weekend. And I was 3 for 4 the weekend before on a transcon. Even SFO-EWR cleared. Improving .... maybe(?).

Cheers

That would be ideal! An automatic sync could happen once a week for example. I'll be happy to help testing it.