American’s Operation Had a Very Rough Summer


You’ve been waiting for part 2/x of our series at The Air Show podcast on what’s wrong with American… and now it’s here. Today we talk about the brand and product, and it’s not all bad. But also, it’s clearly not all good either. Come and have a listen as I get to serenade my co-hosts once again, this time with the song “Happiness Runs.” If you know why I sang that song, well, you have a good memory. If not, you should listen in to find out.


For many years we’ve heard American talk about how running a good operation is the most important thing it can do, and it’s how it’ll win the day. I don’t believe that personally. I think a good operation is just a basic requirement to play the game, but that’s not the point here. For an airline that has so much strategically invested in running a good operation, you would think it would always make sure that happened. This summer, however, it did not.

American ran a miserable operation this summer with fewer than 70 percent of flights arriving within 14 minutes. I took a look at Anuvu data to see exactly what was going on. This chart tells the tale pretty clearly.

American + Regionals Operations and Arrivals Within 14 Minutes

Data via Anuvu

On June 5, American’s summer schedule went into effect. It remained in place until August 5 with the only real break being on July 4. It can be very clearly seen above that on-time performance dropped dramatically when the operation was increased, and then it returned to better levels in August. That’s not a surprise. Airlines always want to push their operations when demand is highest. But the results aren’t usually this bad.

With July looking to be the depths of the poor operation, I decided to focus on that month to break this down further. July wasn’t great for anyone, but it was certainly worse for American than all but the usual basement-dwellers.

July 2025 Arrivals Within 14 Minutes

Data via Anuvu

As if this wasn’t bad enough, American also finished next to last in completion factor, just barely edging out JetBlue for worst by completing a very low 96.4 percent of flights.

Perhaps more surprising is that Spirit is number one, right? But just remember that Spirit does not have a lot of flights. It is not utilizing its airplanes, so it can run a better operation.

This wasn’t the story of bad weather in one hub or anything like that. Just look:

American July 2025 Arrivals Within 14 Minutes By Hub

Data via Anuvu

Sure, Phoenix and LA did better, but look at the bottom. Yes the Northeast was bad, but so was Chicago. And DFW was awful. You can’t blame weather everywhere… or can you? My understanding from those on the inside is that the weather was a lot worse this summer. Storms at DFW came in unexpectedly and stuck around for longer. Chicago has had its own gate construction and capacity issues. And the Northeast was the Northeast. With less slack in the system, it was harder to recover. The thing is, there are some odd indicators that make you think there’s more to this story.

For example, look at the fleet breakdown.

American July 2025 Arrivals Within 14 Minutes By Aircraft Type

Data via Anuvu

Above you can tell that regionals did not do well, but they sure did better than mainline. When weather is bad, it’s usually the regionals that get hit the hardest, so it’s quite a surprise to see the opposite here.

And in the mainline fleet, it was the B787s that were by far the worst. Widebodies usually do better than narrowbodies. As bad as that looks, it hides the fact that it was the B787-9 that was the worst fleet, not even topping 60 percent — it was at 58.7 percent — while the B787-8 was slightly better.

Remember all those stories about the airline’s newest deliveries having mechanical issues? That could not have helped, but that is a very small subfleet. The B777s look better, but that’s only because the B777-300ERs held up fairly well at 71 percent while the B777-200ER was all the way down at 63 percent. This is more than just a problem with a batch of new B787s.

For its part, American gave me this statement:

We are laser focused on further strengthening our operation for the holidays and next summer — learning from where we fell short and making continued improvements and investments to deliver for our customers. One such improvement we’ve already made was our best-ever baggage handling performance this past summer.

So the lasers are focused on the operation and not pointing at pilots’ eyes, which is a good thing. But as I understand it, American is making adjustments in the new year. It’s increasing block times, and it’s changing bank structures a little to give more breathing room. It’s also rethinking how it serves off-peak days to see if it can use those to help maintenance catch up.

It may not be sitting still, but this isn’t a new story. It feel like Groundhog Day, just the date range and the airline name change. Weather is unquestionably becoming more unpredictable and severe, but American is an airline that has staked its reputation on being reliable. It has to do better than this.

Get Cranky in Your Inbox!

The airline industry moves fast. Sign up and get every Cranky post in your inbox for free.

Brett Avatar

16 responses to “American’s Operation Had a Very Rough Summer”

  1. Dan Avatar
    Dan

    Cranky, I think part of the problem for American is over scheduling at hubs. In April, I flew back from Rome to Tampa via Charlotte. We arrived in CLT 45 minutes early and sat on the taxiway waiting almost 2 hours for an international gate to become available. I know anecdotes aren’t data, but from flight attendant comments, it sounds like this wasn’t an unusual occurrence. I’d be curious to know if hub gate availability is impacting OTP on the wide body fleets.

    1. Bobby Avatar
      Bobby

      It seems that AA has under invested in its hub infrastructure at CLT in particular. It’s a legacy of US’s financial problems.

  2. Bobby Avatar
    Bobby

    Those who have studied Lean Manufacturing know that adding block/buffer can have downsides if not done strategically and after all other options have been addressed. Simply adding block can hide all of the embedded operational problems that may exist, which is expensive and doesn’t fix the root problem. Furthermore, adding block can produce a short-term improvement of operational performance, until new unaddressed problems arise. A root-cause analysis is the place to start – undercover the true cause of each operational disruption and fix those causes. A root-cause analysis is intellectually more difficult, which is probably why AA hasn’t attempted that first.

    Furthermore, when a thunderstorm tears DFW apart, I’m not sure that the biggest problem is that the 3-hour flight didn’t have an extra 5 minutes of buffer. Overcoming a thunderstorm is about having levers to delay out and put back together their schedule when a disruption does occur. This seems to be AA’s biggest problem as of late.

  3. Dean Avatar
    Dean

    This article feels silly. Saying its not hub specific is a little ignorant considering every hub is reliant on other hubs operating normally. It feels undisputable that DFW has had more weather than any other hub in the US this year and past. Its every freaking day that weather smacks itself over AA’s largest hub and groundstops the airport for hours. Adding in the fact ATC there has been more underfunded and unsupported compared to other mega airports in the US, thus you get this disaster. PHX and LAX consistently operate better because there isnt any weather but they will still get delayed from the domino effect DFW creates.

    1. See_Bee Avatar
      See_Bee

      It still goes back to management’s scheduling strategy – which is Brett’s point here. I’m not sure what you’re saying is true in this instance (would need data to confirm), but you can avoid the domino effect by isolating hubs with majority out and back patterns. Yes, you still need to rotate planes for maintenance, etc. periodically, but there are mitigation levers that AA could proactively implement in the schedule. Thunderstorms in DFW aren’t a new phenomenon

      For example, DL does this with their scheduling. Rotating planes between MSP/DTW via outstations is more common; however, they don’t rotate ATL tails to/from MSP/DTW at outstations for just that reason. If ATL is a mess, MSP/DTW are minimally impacted

      1. emac Avatar
        emac

        Correct, you can switch to out-back scheduling to isolate delays at a given hub. It’s less efficient for utilization.* United talks about this a lot, have increased out-back scheduling periodically over the years including at Newark.

        *Plane takes off from DFW to Seattle with a bank, waits longer on the ground to depart Seattle to arrive with the next DFW bank instead of departing sooner to hit a Chicago or Phoenix bank.

        https://jetsetterguide.com/news/united-airlines-unveils-comprehensive-strategy-improve-operations-newark-hub
        https://www.hbs.edu/faculty/Pages/item.aspx?num=51621

      2. Bobby Avatar
        Bobby

        I’d be interested, too, in the data, but I get the feeling that AA already operates most of their spoke turns as out-and-back.

        Mixing turns in the spokes can work, as long as the company has defined processes to manage it.

        1. See_Bee Avatar
          See_Bee

          Yeah, I agree; I doubt AA is doing this to any significant degree. This feels more like a system-wide ops strategy gone bad

          I seem to recall Frontier got in trouble doing this a few years ago, and Brett did a write-up on it. I realize F9 doesn’t have as many “hubs”, but LCC/ULCCs tend to do more point-to-point movement through the system instead of out and back to their focus cities

    2. JT8D Avatar
      JT8D

      But it’s possible to schedule for resilience. That is a choice management can make.

  4. David C Avatar
    David C

    Weather in DFW was definitely wetter and stormier than usual. The usual dry July and August never happened.

  5. Chris Avatar
    Chris

    There’s a reason why fliers jokes that DFW stands for “Does not Fly in Weather.”

  6. Bob Avatar
    Bob

    Part of the problem of aircraft reliability is the mechanic’s not have the time to fix issues/problems. Management has the philosophy to use temporary fixes until the aircraft has a long hub layover when maintenance can be done. The problem is these little fixes can cause major delays which snowball and the aircraft/flights are constantly delayed.

  7. tb Avatar
    tb

    When your mainline OTP is trailing your regional by 10 points that’s a Star Trek red alert level klaxon. I think they just don’t care. They continue to run the company like it’s little ole America West outta PHX. Cheaper on paper, but what’s the trailing cost of the actual poor performance day after day? In many cases, they’re using pricing power and sheer amount of lift to paper over these glaring deficiencies. Every time I fly for work I shudder to pull up the booking engine cause I know AA is gonna be cheapest and I just cannot abide their awful OTP and complete lack of service. Seems as though they are just too big to fail at this point. Scott Kirby is dead on with his assessment – UA and DL are in a class above and AA is content to just be out here picking up the crumbs.

    1. Bobby Avatar
      Bobby

      Is UA’s product really that much better than AA’s? Their OTP is only slightly better than that of AA (per the chart above). I’ll grant that UA has more Y+ seating and in-flight screens, but other than that the product seems largely the same.

      I believe that UA’s recent success is due to the backbone of their network being centered on some very wealthy places, all which have benefitted from the run-up in asset prices (except maybe IAH). Furthermore, they seem to know how to put together more competitive networks in their hubs (i.e. ORD, when comparing to AA), and they have been benefiting from profitable domestic growth made possible by United Next.

  8. Angry Bob Crandall Avatar
    Angry Bob Crandall

    Cranky,
    Is there any way to strip out the delays that are ATC related? I was on many flights this summer where the captain pushed back but said that there was a ATC delay so that they were going to wait in the penalty box for a new take-off time. And in my case except for ORD, I had ATC delays in a third to a half of the time in CHS, TPA, MIA and BNA.

  9. JT8D Avatar
    JT8D

    How were competing airlines doing at some of those hubs in the same period?

    Specifically, how did United do at ORD? And United and Delta at LAX? And maybe DL at JFK? Those would be interesting comparisons to make.

Leave a Reply to David C Cancel reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Cranky Flier