July Was a Terrible Operational Month, But There Were Some Bright(ish) Spots


If you flew on an airplane in July, the chances were higher than usual that you had an operational problem. And if you flew in the Northeast US during that time, I can almost guarantee you had an operational problem. From weather issues to air traffic control problems, July was bad, and it wasn’t due to any specific meltdown. It just appears to be a general inability for the system to function when things don’t go perfectly.

I dug into Anuvu operational data to break this down, and I decided to challenge myself to see if I could pull out any bright spots. There were a few.

Let’s start with the most basic of measures, just looking at the percent of flights that airlines completed in July (including their regional operations).

July 2023 % of Flights Completed by Marketing Airline

Data via Anuvu

The shining star was Alaska, which isn’t a surprise. Alaska is a good operator, but it also largely avoids the Northeast, so it’s exactly what I’d expect to see from the airline.

But who is that in second place? Allegiant? That’s a welcome relief for travelers who have suffered through Allegiant cancellations over the last few years with no real alternatives. Again, the airline doesn’t do much flying in the congested Northeast, but that shouldn’t take anything away from this.

Southwest had a very solid performance as well, so kudos to the team over there. Same goes for American which did cancel more than the top three, but it has a lot more flying in the Northeast. Below that is where things start getting ugly.

For United, it’s a Newark story. Newark has simply been unable to cope with the number of flights scheduled when the weather gets remotely ugly. And JetBlue, well, it’s just JetBlue. Yes it has a higher concentration in the Northeast, but it always tends to run a weaker operation.

Let’s move on to those flights that actually did operate. How on-time were they?

July 2023 % of Flights Arriving Within 14 Minutes of Schedule by Marketing Airline

Data via Anuvu

Alaska is there again in pole position as the only airline above 80 percent. And then there’s the regular first place finisher Delta right behind. But then… American. American has been crowing about its improved operations recently, and really, the airline deserves credit for finishing toward the top of the pack. It’s a far cry from the old days.

Southwest and Allegiant may not cancel a lot of flights, but their on-time performance leaves something to be desired. And then there’s JetBlue which… I’m trying to keep it positive so I’ll just skip over that.

It’s one thing to look at a single month, but I do want to zoom out here and look at some other metrics for the first seven months of this year combined. Specifically, let’s look at on-time performance but then also incorporate block performance.

Jan-July 2023 Arrivals and Block Performance by Marketing Airline

Data via Anuvu
B0 = % of flights operating within schedule duration
A14 = % of flights arriving within 14 minutes of schedule

B0 means the percent of times that a flight operated within the scheduled duration, meaning if the flight was scheduled for 180 minutes and operated in 180 minutes or less, it would count… even if it ran 10 hours late.

What we can take from B0 is how efficiently each airline is scheduling itself. A really high B0 would indicate there’s probably a fair bit of padding in there. Just take a look at Delta. It has the best on-time rate for the first seven months in this group, but it also has the highest B0 by a large margin. It is buying that on-time performance by padding more than the others.

Beneath that, you can see Alaska which has turned in a truly remarkable performance. It has a very low B0 but still keeps the planes arriving on-time more than nearly everyone. Yes, yes, it doesn’t have much in the Northeast, but that is still a well-oiled machine right there.

While Alaska deserves the crown, let’s again not ignore American here. American has a higher on-time percentage than both Southwest and United with a lower B0 which means it is operating more efficiently and getting a better result. The airline has really been improving, and it deserves real credit for the operational work it has done.

Before I finish this, I want to do one more look. This is on-time performance by month for each of the big four going back to 2018.

Arrivals Within 14 Minutes of Schedule by Marketing Airline Over Time

Data via Anuvu

There are a few things to take out of this. Of course, we see the big bulge during COVID. With so few flights operating, it was easy to run on-time. But look around the edges and you get more useful trends.

Delta continues to lead the pack overall, but look at how American has really closed that gap in the last year. It is running a better, more consistent airline for the first time in a long time. Southwest, meanwhile, has really fallen toward the bottom of the pack since COVID. It may have bigger fish to fry right now, but this has to be on the radar.

Let’s end this on that positive note that I started with. Kudos to Alaska and to American for doing their best work this year. It hasn’t been an easy summer, but some airlines have done it better than others.

Get Cranky in Your Inbox!

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

16 comments on “July Was a Terrible Operational Month, But There Were Some Bright(ish) Spots

  1. I haven’t seen the B0 metric mentioned or used a ton, but I really like it when it is used in the proper context. The way that you used B0 & A14 to compare/contrast/analyze airlines’ operational performance was quite fascinating. Analysis like that is what has kept me a loyal reader of this blog for many years.

    I know you were trying to keep the post positive (and of a reasonable length), so this isn’t criticism at all, but I’d also be curious to see histograms of delays by airline, or perhaps even S-curves.

    By S-curves, I’m thinking of flight data sorted by minutes late on arrival and a line for each airline, and with the Y axis of the graph representing the cumulative % of flights, including flights that arrived on time, and the X axis representing the delay, such that one could easily pick a point on the line and say that (as an example) for 80% of the flights that United operated in July arrived <= 25 minutes late (or whatever). That said, management in my day job absolutely LOVES graphs like that, so I may be a little biased in terms of the way that I like to visualize data. :-)

    1. Took the words from my fingers… I was just opening up the comments to praise his use of block performance to really put on time in context. At an airline I used to work for as leadership in the hub, we took the block concept and applied it to hub performance in a way… we measured hub throughput… what was the change between on/in/out/off with a goal of positive numbers… the thought being the hub can drive system performance and even if the system is late, you don’t want to lose even more time.

      I like Cranky’s point on American’s B0…. that it does show a genuine improvement. You can buy performance through block, but spending that to make lasting operational changes is the better strategy.

  2. Before we crown American too much, consider the weather at the hubs. Here In Dallas in July it’s been brutally hot…been at least 105 for a week straight now for example…and almost no storms.

    And Phoenix has had very little monsoon storm activity too, and Charlotte is a bit under on storms so far also.

    Compare that with constant storms in the NE. So often these numbers depend on who is getting weather and who is not. Especially when the NE is so vulnerable to any disruptions.

    I say that as long time AA supporter also. I’m glad to see them doing better, but the weather is helping a lot. And as a resident sweltering in Dallas I’m ready to see some storm delays here for sure.

    1. The problem is that a single storm can disrupt an entire network or the entire system depending on size & or location. The only solution I can think of is the system needs to shrink to minimize such impacts by reducing capacity. First reduce short regional flying & then up gage keeping slots filled with fewer, but larger planes.

    2. Yes, when hub weather is good, all goes well.

      But it is interesting to see how AA has improved. I am surprised and impressed, compared to prior summers. And yes, my AA flights in July were on time or early.

    3. AA might be smaller than DL/UA in NYC proper, but if we’re just talking about the NorthEast (aka the ATC impacts to the NE), AA isn’t smaller than any other carrier therefore any network or on-time issues would impact AA just as much via their DCA, PHL, or NYC hubs. AA just seems to be doing a much better job compared to their past performance.

  3. Is it possible to say to what extent Alaska’s excellent A14 with a low B0 means either they allow more time on the ground and/or have a bit more slack in the fleet? (I think allowing more time on the ground to make up time instead of scheduling the absolute minimum possible turn time is a bit more honest than putting a ton of padding in the schedule between departure and arrival.)

    And I echo others that B0 seems like a very useful statistic.

  4. JetBlue… what a basket case. God willing the government successfully blocks the NK merger and saves them from themselves, then maybe the board gives the C suite a way-overdue kick to the curb. Anybody notice that Spirit’s hurting too, posted a miniscule 3% adjusted margin in Q2?

    1. As Friday’s Cranky Weekly Review succinctly put it, “Forty-six percent of JetBlue’s flights in July arrived late, with an average delay of a whopping 85 minutes which is… *checks math*… not good.”

      I wonder how much B6’s historically weaker operational reliability impacts the ability of its commercial team to sign “preferred carrier” deals with corporate clients. Certainly there has to be some internal push/pull between the commercial team and operations, especially as B6 matures to the point where it can’t rely on leisure travel and small business travel as much.

    2. Spirit’s earnings got me looking at balance sheets and I don’t feel like they are in as bad a position as is being described. They’re sitting on something like 1.2 billion dollars in cash. They can flail for quiet some time and they have the money to sort themselves out. I’ll be curious to see how the merger battle plays out between the government and JetBlue. I get the sense that JetBlue needs Spirit more than the other way around.

      1. Spirit has been sitting on that amount of cash, more or less, for a decade. Pretty good rainy day fund, and good they don’t immediately run to it at slightest inconvenience.

  5. Just as AA’s performance needs to be considered with regard to weather, so does JetBlue’s. A major percentage of their flight touch the NE, compared to other carriers. They are trapped there, which is the entire reason they are trying to grow out of it by buying Spirit’s planes and pilots.

    I would strongly suspect that if you just considered flights in and out of the NE, everyone was in the same boat B6 and UA are…too many flights for too little airspace and when it goes down, they struggle.

  6. The one that should jump out at you is Frontier. They are NOT tied to the NE, and yet they barely got half their flights on time.

    Ain’t neither of the Spirit wanna be suitors doing well right now operationally.

  7. I’m surprised that Alaska did as well as it did, since I’ve heard through the grapevine that AS is running its fleet harder and longer than they ever have before. So kudos to that team for making it happen.

    1. Alaska’s hubs are in Seattle, Anchorage, and California. All of those have good weather in July. That’s the major reason for their performance.

      1. Don’t forget Portland. It maybe less impactful than Seattle, but feeds north/ south traffic as well as east/ west all be it on a smaller scale.

Leave a 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