I’ve seen a handful of stories lately on the long-true fact that airlines sell schedules they don’t actually ever plan on flying. It was worse during the pandemic when schedules changed frequently… and up to a month or less before departure. Now we’ve settled back into a more traditional pattern which still leaves a lot of uncertainty for travelers. Today I’ll show you exactly what’s happening with one airline in particular, United, since it seems to change more frequently than most.
Your standard legacy airlines usually sell tickets on a rolling 331-day schedule. In other words, you can buy a ticket for 331 days from today and not beyond. Tomorrow, that window will be the same, but it just ends one day calendar later. That means that as of today, you can book flights up through early January 2024.
As an aside, low-cost airlines tend to operate differently and have set booking windows. Southwest is the largest of these. It was selling for travel through August 14, but this past weekend it extended its schedule through October 4. The schedules at these airlines tend to be more accurate since they go available for sale closer to departure. But let’s get back to the legacy airlines, and United in particular.
Below you’ll see a chart of United’s 2023 departures by day as filed this previous weekend. I’ve divided it into three separate categories.
Obviously, the first month on the chart is completely accurate (barring operational cancellations) since this is calendar year 2023. Those dates were already flown. But between now and the end of May, United’s schedule is largely accurate. You can see the bump up in March for spring break, and then it settles a bit. When I call this “accurate,” I mean it’s mostly correct. There will be some changes as we get closer, but for the most part, airlines try to get largely accurate schedules about 100 days in advance.
United, however, appears to have gone to a multi-tier model as we covered in Cranky Network Weekly this past week. Between June and Labor Day, United has introduced a modified placeholder schedule. This is different than the wildly-inaccurate schedule that’s posted post-Labor Day. It’s closer to being accurate, but it most definitely is not there. It will be whittled down week by week, some week with bigger changes than others, until it’s flyable.
In the summer schedule, you can see the different pattern to peak and off-peak day flying as compared to what’s scheduled earlier. It’s all part of the process that United seems to prefer, even though it results in multiple schedule changes for people who book further in advance.
This chart alone doesn’t show the depth of what changes, so we dug in deeper this week in CNW to show just how much different it is. I’ll bring some of the charts as created by Visual Approach into this post. Let’s start with a look at departures by narrowbody fleet type.
Here’s where you can see some pretty big shifts. In the June schedule, 737-700/800/900 flying jumps up dramatically in the schedule, but there is not massive number of new deliveries. You can see the 737 MAX goes down a little. Some of these will change, but there will also be flights canceled as we get closer in.
But then you can see what’s going on in September where it’s completely divorced from reality. For some reason, United removes nearly all of its MAX flights and has a very high level of A319/A320 flying. In some cases this will just be fixed by switching aircraft types and nothing else, but with each aircraft type also comes different block times and seating configurations. Changes are very likely.
If we go down further, here’s a deeper look at the regional fleet, and it’s even crazier.
The wild swing in flying by different airlines is not real. In particular, we know that Air Wisconsin is leaving United and will be flying for American instead. You can see that flying start to leave the United schedule until June when it inexplicably jumps back up. Then it really jumps in September. Air Wisconsin will not be flying for United then, but United is selling flights that it says will be operated by them.
At the same time, you can see how Mesa flying jumps up until June when it settles back down. Mesa will be putting its CRJ-900s that fly for American now into service with United. At this point, none of the CRJ-900s are in the schedule, but United has instead just filed a lot more Embraer 175 flying on Mesa, presumably ready to move airplanes around when it knows more detail about the transition. (My bet is that the initial move will take Embraer 175s from Houston and send them to Dulles, then that will be backfilled by CRJ-900s in Houston.)
It may be hard to believe that these inaccurate schedules can exist. I asked United for comment, but did not hear back. My guess, however, is that from United’s perspective, it probably doesn’t think it can put forth a correct schedule much further out, but it wants something in the market. And it presumably also figures that it will be able to provide travelers with a close-enough option to what they booked once the schedule settles. But that’s an opinion United has and not a fact that everyone can agree upon.
The number of changes can be mind-numbing. For example, I’ve been keeping track of a roundtrip from Newark to Jackson Hole that was booked for a Cranky Concierge client last June for travel this month. The changes have been frequent (shown in bold and italics).
- Original Booking: United 321 Lv Newark 1130a Arr Jackson Hole 234p, United 2155 Lv Jackson Hole 330p Arr Newark 947p
- June 2022: United 321 Lv Newark 1130a Arr Jackson Hole 234p, United 1120 Lv Jackson Hole 330p Arr Newark 947p
- July 2022: United 278 Lv Newark 1130a Arr Jackson Hole 234p, United 1120 Lv Jackson Hole 330p Arr Newark 947p
- August 2022: United 278 Lv Newark 10a Arr Jackson Hole 104p, United 1120 Lv Jackson Hole 155p Arr Newark 812p
- August 2022 part 2: United 2092 Lv Newark 10a Arr Jackson Hole 104p, United 2267 Lv Jackson Hole 155p Arr Newark 812p
- September 2022: United 2092 Lv Newark 10a Arr Jackson Hole 104p, United 1552 Lv Jackson Hole 155p Arr Newark 812p
- October 2022: United 1056 Lv Newark 906a Arr Jackson Hole 1212p, United 1552 Lv Jackson Hole 113p Arr Newark 730p
- December 2022: United 301 Lv Newark 906a Arr Jackson Hole 1212p, United 1552 Lv Jackson Hole 113p Arr Newark 730p
- January 2022: United 301 Lv Newark 906a Arr Jackson Hole 1212p, United 1552 Lv Jackson Hole 112p Arr Newark 729p
Are we having fun yet? Most of these are insignificant flight number changes, but they are still remarkably annoying. I think every traveler takes a deep breath when they see a “Your Schedule has Changed” email roll in.
But buried in here are the significant schedule changes. The first one in August isn’t more than 2 hours so it wouldn’t be eligible for a refund if the traveler couldn’t make the schedule work. Let’s forget that 90 minutes seems pretty material to me. The October schedule change actually would be cumulatively over 2 hours, but by then, the traveler may have already adjusted plans knowing they had no other choice after the first one happened. At least this isn’t on American, since American requires an unreasonable 4 hour change before allowing a refund.
So what we have here is a mix of regular monthly schedule changes that make minor shifts along with larger changes that make bigger adjustments as they get the schedule closer and closer to reality.
For United, I’m sure this makes sense, but for customers it’s a pretty miserable way to be handled. At least now you’ll understand better how it works.
17 comments on “Only A Third of United’s Schedule is Accurate Today (A Breakdown)”
Receiving an email when your flights change by a minute or two gets old fast.
For schedule changes that aren’t significant (say, < 10 or 15 minutes cumulative total since the last "schedule change" email), I think it would be better to just hold off on sending an email and to do an email/app alert a few days before the flight or when travelers check in.
10 or 15 minutes is a huge impact for those of us that live really close to the airport, or take public transit. I plan on leaving my house for domestic flights exactly 2 hours before wheels up, and align that with the train schedule. 15 minute change could be to difference between making the flight or not.
Not to argue or belabor the point, but… Do pax really need/want an email every time the schedule changes by a few minutes? I get with the “bigger” schedule changes, but at a certain point United is basically training its customer to ignore its schedule change emails.
Notifying the passenger of the revised (“final revision”, so to speak) departure time a few days before travel should allow the passenger to still get to the gate on time.
I largely agree with you. Frequent notifications are misleading and not helpful.
However, there are just so many customers and obviously not everyone wants the same thing. I suppose the airline can have a setting in user profile that asks “how frequently do you want to be notified for schedule changes?” But that is just asking for more trouble than anything else. We have all heard stories of people showing up at airport only to find out their flight had left hours ago due to schedule change that they were not notified. (And we all know airline IT is not capable of rolling out this kind of feature).
Small schedule change of a few minutes can definitely have big impact. On top of those who utilizes public transportation as mentioned by @SFO-BART (e.g. all of sudden you can’t catch the first/last bus of the day and need to find alternatives), it also have big impact on connections. A legal connection meeting MCT may suddenly becomes ilegal, vice versa.
I’m confused about why the far-out schedule isn’t a better guess. I get that the airline doesn’t want to put a lot of thought into something that far in the future. But from these graphs it looks like their placeholder schedules are clearly impossible. Why not put out something that’s a closer approximation of reality?
Cynically, the answer might be bait and switch: schedules are pruned down, not fleshed out. So airlines get some benefit of offering customers better timing than will eventually materialize.
More likely they are “feeling out” demand that far in advance so they can eventually switch up larger and smaller aircraft to more closely match demand for certain routes. And of course when you start switching up aircraft, schedules are bound to need tweaks as well.
grichard – It is all based on past schedules, because that’s the easiest thing to do. The problem is, past schedules have been a lot less accurate as of late.
That regional schedule in particular has a huge impact on me. I do a lot of regional flying on United, and given that those flights are daily or less, it has a significant impact on my work schedule, which I have to adjust in order to be able to get back home after working. The only consolation that I have is that my work schedule is even more chaotic than this, so I can’t really do any reservations confidently more than a month out. But trying to hit two moving targets at once adds more stress to my life.
I remember checking a CO flight years ago from EWR – LAX I was going on & noticed how every so often the schedule would shift by a few minutes here & there. In the end the flight was at 6:45AM VS 7:00AM. I found it a bit interesting & now I understand how it works.
How is the breakdown between domestic and international? I thought international schedules were a bit more stable.
Ron – Yes, international schedules tend to finalized further in advance, but we continue to see summer schedules tweaked regularly for international. In those cases, it’s usually more about 10 to 15 minute tweaks, but it’s more than that too.
I think it’s important to distinguish between the schedule and capacity. The farther out schedule is managed at a capacity level. As in overall capacity is up/down X.X% Multiple things can change the airline capacity goal but the principal influence is industry analysts since how closely the airline adheres to the analysts’ expectations affects the stock price (which ultimately affects executive compensation). The schedule is a realistic schedule in that it’s probably just last year’s flown schedule, capacity adjusted, but it’s also certain to be tweaked.
So if the schedule is up say 3.0%, but analysts expect it to be up just 2.8%, then the planners pour through the schedule to get it down to 2.8%. But then perhaps the economy’s looking better and analysts raise their expectations to 3.2%. Then planners add stuff back to the schedule to get it up to 3.2%. This also happens at Intl/domestic breakdown or even certain regions.
The planners put serious thought into each market as they shift capacity, but they’re not flowing the total schedule either to make sure it works. That happens about 5-6 months out in the scheduling department. The planners also are not concerned with who the regional carrier is as long as they have the fleeting the way they want it. Again the scheduling department will sort out the regional carrier mix in a closer in schedule.
As a scheduler at a major airline, I can confirm.
Bravenav – Even that tie to capacity guidance doesn’t flow through to the end of schedule. For example, last month United issued guidance that FY 2023 capacity would be up in the “high teens” percentage over 2022.
https://ir.united.com/static-files/26c1e428-b195-4144-bee4-14e59375bf99
But if you look at capacity as filed today, FY 2023 is currently 23 percent above 2022.
As for regional carrier assignments later on, that’s fine from an airline perspective, but it doesn’t take into account customer impact which can be significant. Maybe not as significant in normal times (though some people do have regional airlines they won’t fly), but United today is in one of the worst examples of that since Air Wisconsin is leaving and Mesa is adding bigger airplanes. It will presumably add a not-insignificant number of flights that will go from single cabin to three cabin. It will require a lot of maintenance on a traveler’s behalf to keep up with that. It wouldn’t matter as much if, say, United’s auto-reaccom system automatically put someone into First Class if they were on an international premium cabin ticket. But nobody does that.
The real question is how much this matters to the DOT which just made some comments about airlines publishing unrealistic schedules. There are clearly costs to reaccommodate passengers from flights that won’t operate or significantly move (beyond what automation can do) but the government seems increasingly focused on extracting a pound of flesh from the airlines and unrealistic advance schedules seems to be their complaint du jour.
UA has more challenges right now in scheduling in advance due to its regional carrier and MAX 10 delivery issues (the MAX 7 is still not certified for WN) but the government might squeeze more on UA and other airlines that publish schedules that are reduced as UA is doing.
There are times when these changes have worked in my favor. At the time of original booking, I’ll choose a reasonably priced (usually PE, sometimes F) seat on a flight that departs at an acceptable time. I would have preferred a different flight with an optimal departure time but that seat is often far more than I am willing to pay. DL is notorious for this. I’ll keep an eye on schedule changes and be ready to rebook to the flight I really wanted at no additional cost to me. I feel like I am gaming the system but in a way airlines deserve it by subjecting us to unpredictable pricing algorithms.
One aspect doesn’t seem to have been mentioned. When I book, I’m paying money now with my credit card for a string of things – flights, hotels, and other stuff. Maybe my itinerary involves 2 or more different airlines on separate tickets, perhaps there is something else in the itinerary also. I have to take on the risk that everything works… and if I lack confidence, I will wait until closer to travel date before booking.
It’s really not in an airline’s interest for people to delay booking until later – not only does it mean some people decide not to book at all… but it also makes an airline’s business more volatile and less predictable… much like people get nervous booking far ahead with financially troubled airlines. That confidence-to-book-in-advance is important… lose it and it will have a significant impact on long term business