There’s been a whole lot of talk around January’s operational performance. Heck, United even decided to put out a Super Bowl ad digging at Southwest about it in Denver. I decided to cuddle up with Anuvu’s operational data to pick it apart and see who really did well in January and who did not. It’s quite a mixed bag.
To start, I put a high-level chart together that’s full of confusing lines and colors for each marketing airline (meaning, regionals are included). Enjoy.
January 2023 Operational Performance by Marketing Airline
There is a lot going on here, so let’s start by noting some of the big picture stories.
- Frontier had a very bad January, finishing dead last in on-time performance by a lot, and pulling up at next to last in completion factor
- Delta did generally well, as it usually does, but it wasn’t even up to Delta’s usual standards
- American’s focus on D0 continues with it pushing a higher percent of planes on time than anyone else
- JetBlue canceled the lowest percentage of flights. Good on you, JetBlue
- Southwest canceled the highest percentage of flights, but its on-time performance was near the top (not D0, which Southwest always lags in, but A14)
- For an airline that created a Super Bowl ad, United is pretty middle of the pack
- It was clearly a rough month for everyone with nobody getting above 80 percent on time
That’s all well and good, but that’s the easy story. There are a whole lot of subplots going on this month, and that’s where I was hoping to focus to find some fascinating kernel of information.
Let’s start with the highest profile event of the month, the FAA’s NOTAM system failure which snarled traffic on January 11. Here’s how the airlines did on completion factor on the day of the event as well as one day prior and one after.
Completion Factor by Marketing Airline During FAA NOTAM Outage
Anyone else see an upside down widget here? Anyway, what really stands out most here is that Southwest and American canceled a whole lot more than anyone else that day. They both recovered very quickly, but that day was a rough one.
Even if a flight wasn’t canceled, forget flying on time on the 11th. This hit the East Coast hardest since it was really early in the morning on the West Coast, so Hawaiian’s 77.9 percent followed by Alaska’s 69.1 percent arrivals within 14 minutes sound ok. But then it falls off a cliff. Allegiant hit 55.9 percent with JetBlue right behind at 54.8 percent. Southwest was worst at 20.2 percent.
For the month overall, cancellation numbers really weren’t all that bad. When the worst is 97.2 percent (Southwest), you’re doing ok compared to some of the December horrors. So I decided to look deeper into arrivals within 14 minutes. I started by pulling the big four and putting them into a daily chart.
January 2023 Arrivals Within 14 Minutes by Marketing Airline
I stretched out the scale here to make it easier to see differences. On January 11, they all plunged below 50 percent as we just discussed above. The general arc of the month here shows a rough end to the holiday period followed by some nice calm. But then it just got choppy. Toward the end of the month, we see a mix of trouble at the various hubs that hit different airlines differently.
January 22 was awful for Delta. Why? Look at all that rain around Atlanta. There were thunderstorms all night long into the early morning.
And what about United on January 25? Houston is the problem child that stormy day with… more than 4 inches of rain?!
And that’s really what drove the performance here. Different hubs got hit harder at different times due to bad weather. There were no meltdowns, just responses to bad weather. Let’s take a look at that a little more.
Denver was in the news a lot, so let’s look there first.
January 2023 Denver Departing Flights — Arrivals Within 14 Minutes by Marketing Airline
All the hub airlines in Denver had a miserable month. You can argue that Southwest was worse than the rest during some of the events, but it was generally better than the rest during good times. There’s no clear winner here. Everybody was a loser in Denver.
Or how about a look at Chicago?
January 2023 Chicago O’Hare and Midway Departing Flights — Arrivals Within 14 Minutes by Marketing Airline
Southwest certainly started off worse but then it jumped ahead. American definitely had a bigger down than others later in the month. Something clearly went a little more wrong there. Looking at the regional carriers vs mainline, I don’t see any notable differences, so it was some systemic issue.
Dallas shows a similar story. Southwest had a good run above American in the early part of the month but then American caught up. And who can guess when the ice storm was? Yup, that’s an easy one.
January 2023 Dallas Fort Worth and Love Field Departing Flights — Arrivals Within 14 Minutes by Marketing Airline
In the end, it was just a bad month all around. Some airlines were less bad than others, but nobody ran a particularly good operation. It looks like nearly everything was just the result of poor weather. Such is life.
28 comments on “Picking Apart January’s Operational Performance”
Although I’m not very sympathetic towards the airlines in bad weather, I see where it can affect them.
The one stat that we will never probably get to see is time it takes an airline like AA to get a passenger to their destination after their original flight is delayed.
It is embarrassing when they say things like (we can get you out two days from now.
They cannot control the weather, but they can have significant control on the recovery.
Not so much.
These airline run their planes almost full already. When you have a significant disruption, where are the additional open seats to clear a backlog?
Let’s use really rough numbers for DFW. American runs close to 100,000 seats out of here each day. At 75% loads, that means they have 75,000 passengers and 25,000 empty seats.
Now have an ice day where they have to cancel most flights, and 60,000 passengers need to be rebooked.
There aren’t 60,000 open seats the next day. Instead, it’s several days’ worth of open seats. So they can’t accommodate everyone. And other airlines don’t have near enough space either.
That’s an explanation of a problem I already pointed out. The ice storm is the exception, the reality is the daily and werkly IRROPs.
The lack of circa 80’s interline agreements, customer service holds, regional jet expansion etc combined with the lack of spares/reserves is just sad.
They are going to keep on until they get federally regulated to make changes.
Running flights so close to full so that they have no leeway to handle cancellations is a choice the airlines make. Certainly running at lower load factors would have costs on good days, but running at higher load factors has a very real cost on bad days.
What would your solution be?
Are you going to mandate that they have to leave a certain number of seats open?
How much more in airfare are you willing to pay for this? Because if you tell them to block 25% of seats on every flight, your airfare just went up 25% or more.
And interline agreements won’t do a damn bit of good. The hub airline typically has 80% or more of seats. There aren’t enough seats on other flights to help. Plus those flights are already almost full too, and they are dealing with cancellations as well.
So how would you guarantee someone out the next day?
I’m saying that absolving the airlines of blame because flights are full is not reasonable; they choose how willing they are to make customers deal with inability to rebook by, in part, choosing how little slack to leave in their schedule and how fully they book their flights.
I didn’t suggest anything like a 25% reduction in load factor, but it’s not as simple as reducing x% of seats sold increases fares by x%. Fares are set by supply and demand, not (directly) by costs. But in general, the absolute race to the bottom, low fares at all costs mindset has made the experience worse for consumers. Fortunately that’s gotten a bit better in recent years; it’s now easier than it was for much of the 2010s to pay a bit more for a somewhat better experience.
It was others who brought up interline flights, but it’s not true that they won’t help. Sure, if a snowstorm hits Chicago, United will be as hosed as American and Chicago O/D passengers won’t have great options. But a passenger going from Indianapolis to Portland could be rerouted on Delta via MSP if MSP is less impacted, for example. Neither interline rebookings nor lower load factors are silver bullets, but they both help.
You keep wanting to blame the airlines. But the only way to assure seats will be available for displaced passengers would be to force them to block seats.
Given that demand is already high, and flights are pretty full, what do you think that would do to prices? Simple supply and demand equation, and you would be artificially reducing supply. It’s not like these airline have a bunch of spare planes or pilots to add more empty flights.
Sure you could make American switch passengers from Indy to Portland to a Delta flight through MSP or DTW. But those flights are also pretty full, and if you close down AA, UA, and WN hubs in Chicago it won’t take any time at all to fill those other seats.
Problem: the system runs at near capacity in good times. In bad, it’s overloaded and people get stuck.
Solution: none without substantial additional costs.
Alex, I don’t think you will change his mind.
Airlines now provide an ESSENTIAL service that when they fail to be able to deliver on what is now an extremely high priced service (from my airport), it has multiple costly ramifications for the passengers.
My original point that I am sticking with is that they (the airlines) better figure out soon a way to deal with this issue or the federal government will eventually fix it for them…then nobody wins.
Something like eu 261 but daily so there’s a real financial cost for not getting someone to there destination for several days.
Even though I work in a completely different field (manufacturing) I learn a lot about seeing deep dives like this on operational failures. Thanks Cranky!
Good job on KJZZ today.
So many jets coming into Scottsdale….
Yo – Thanks. They had so many questions for me before we went on air including things about the economic impact of the Super Bowl. Definitely above my pay grade! But in the end, I think the questions asked were good ones.
Jetblue stubbornly refuses to cancel flights regardless of the customer impact. Remember, a 24 hour delay still counts as completion!
Yup. That was Delta’s playbook for over a year.
But would you really have them do it differently? At least if they don’t “cancel” the flight, all the folks on it won’t have to rebook and find another seat on an already full flight.
At least when Delta did it they’d usually find everyone seats on other flights and run the plane nearly empty. It would still need to be moved for repositioning, so they just moved it under the original flight number.
Consider this….. Major hub airport has poor weather conditions, resulting in major delays – but it can still operate, but not at 100% capacity. 100% capacity is by definition set up for a good weather day and all systems working perfectly. But now, the hub airport is reduced to 60% capacity (say). So, everything backs up and that results in cancellations and from numerous delays.
So now you have a problem that has nothing to do with the airline. How would you handle that?
Or, is that a true statement? Who plans on a 100% perfect operation 24/365? Planners know that there will be disruptions. Airport planners know this, and airline planners know this. They all have their area of responsibility.
So, IMHO, it becomes an issue of how can any airline handle (the inevitable) disruptions. Blocking seats (which would raise costs) is certainly not the answer! I don’t see any good solutions here – so far!
I don’t see anyone who suggested blocking seats. I at least suggested running the system at a somewhat lower load factor than they do now; that doesn’t mean no flights will be booked full (or overbooked). It means revenue management that optimizes for a somewhat lower systemwide load factor.
Airlines know that they won’t be able to operate at 100% all the time, so booking more than they can fly does in fact have something to do with the airline, and it has very real costs to both airlines and passengers. I’m certainly not arguing that it’s obvious where to balance those competing factors. No airline truly books their operation at 100% of their theoretical maximum capacity with 100% load factors; they all leave some slack. The not-at-all simple question is how much slack in terms of schedule padding, turn times, fleet utilization, and load factor.
There are some airlines that book above 100% – simply because they have high levels of no-shows. I may be dating myself, but this was quite customary ‘back in the day’. I suspect that with greater levels of automation it’s not as necessary as it was before.
Sure, the overbooking I referred to; their booking models predict actual loads at 100% or less. Either way, with their pricing models, scheduling, etc, airlines decide how much flexibility and ability to accommodate IROPS they want. It’s airlines, not consumers, who have the data and the big picture understanding to design that flexibility.
David C, what is the federal government going to do? There isn’t even full time FAA head at the moment – they aren’t going to do much at all right now.
EVERY solution you can come up with increases costs, which means passengers will pay more. Some people who read and post on this blog think that they just can sock it to the airlines. Not happening. Whatever costs we leverage against the airlines will be charged back as fares, or possibly as tax abatements and other public funding.
You know what the only real solution is? Tell some people to drive instead of fly, either directly or by raising fares. Reduce the demand. That’s the issue – demand is HIGH, and is close to outstripping the capacity of the system. The system works okay during normal times, but when weather slows things down, there is not enough slack to handle it. So…who are you going to say can’t fly?
I also want to take a quick shot at the idea that making airlines interchange passengers. Suppose a passenger buys a ticket on Spirit between Dallas and Atlanta for $89. That flight is canceled. American and Delta charge $199 or something. Are you going to force American and Delta to take the passenger at the cheaper cost? How is that fair to everyone else on that plane that paid $199? It just doesn’t work.
Again the reality – weather is going to screw the system up, and we will just have to live with it the best we can. As a passenger, always watch the weather, plan accordingly, and ALWAYS have a backup plan, if not 2.
With all due respect, I’m not advocating for that plan, I will repeat that if it continues without some good faith effort a solution will be forced on the system. It does not have to be fair to be put on anyone.
I work in the nursing home industry where regulation and penalties mount against operators even in situations beyond their immediate control.
Again, I’m not drawing a moral equivalency between a nursing home resident and an airline passenger but rather the ability of the government to force ‘solutions’ onto ‘essential’ industries’….
I’m just telling you there was a time when IRROPs were better handled…
There was such a time, before today’s massive demand and high load factors, as well as todays much higher crew rest requirements.
CF
From your experience, when you point out Frontier’s performance as being abysmal, what factors begin to play into an airline shooting itself in the profitability foot because they simply won’t invest on operational stability?
David – It’s hard to say. At some point, airlines all realize that poor performance won’t cut it. Spirit did that when it brought in Bob Fornaro several years ago to run the airline. Frontier may eventually do the same, but it seems to be even more deeply-devoted to keeping costs low than Spirit, so it might not have the epiphany.
CF D0 A14 mean what?
CF = Completion Factor.
D0 = Depart within 0 minutes of scheduled departure time.
A14 = Arrives within 14 minutes of scheduled arrival time.
CF = Completion Factor D0 = Departures on time or early A14 = Arrivals within 14 minutes of schedule