How Southwest Tanked Its Operational Performance… And Then Took a Year to Fix It

Southwest

In early 2013, Southwest realized that demand was strong. Faced with a set number of airplanes, it had two choices. It could keep flying its existing schedule or it could push itself and squeeze more flights into the system. It chose the latter, and it incredibly failed to understand how badly that would drag down the operation. Beginning in August of last year, Southwest’s operational performance tanked. Due to issues unique to Southwest and its archaic technology, it wasn’t until last month, a full year later, that on-time percentage returned to normal.

While at Southwest media days last week, I spoke with EVP and Chief Operating Officer Mike Van de Ven. He explained in detail what went wrong, why it took so long to fix, and how the operation is now back on track. This is the story of Southwest’s bad year.

[Disclosure: Southwest paid for my flight to Dallas and hotel]

An Aggressive Schedule
In January of 2013, the operations group looked at past performance and told the network team that it could make a tighter schedule work. With demand strong and fares good, the network team wanted to take advantage. When the schedule was agreed upon, Southwest was going to operate the equivalent of 16 new airplanes’ worth of flying without increasing its fleet at all.

To achieve this, the airline had to make changes in two areas. First, it cut down on its turn times. Those are the times scheduled in between flights; or the time it takes the “turn” the airplane around. At the same time, it trimmed its block times. Block time is the amount of time scheduled from departure from one gate until arrival at the next.

The ops team looked at performance during 2012, a relatively good weather year, and it figured it could make this work. Southwest had run an aggressive schedule like this before, but the airline somehow failed to realize that this wasn’t the same old Southwest. The new schedule went into place in August of 2013 and things immediately went south.

Southwest On Time Performance Versus Industry

August 2013 wasn’t a great weather month, so the team cautiously bet that this was just a temporary problem due to worse than expected conditions. As you can see above, that bet didn’t pay off.

Identifying the Problem
By November and December of 2013, it was clear that this was a systemic problem. Though the airline could operate its schedule well on a day where nothing went wrong, those days were few and far between. Weather got ugly, airplanes broke, and the airline fell behind on a regular basis. The schedule Southwest had operated in 2012 was no longer a schedule it could run. But why?

The increase in weather issues had an impact, of course. Southwest found itself flying around weather more frequently, and those tightened block times weren’t enough. But that was just the tip of the iceberg.

It turned out that a lot had changed since 2012 in the way Southwest operated, and incredibly, the airline didn’t quite realize how much of an impact that would have. To begin, Southwest had gone through a process of adding a row to most of its 737-300s and all of its 737-700s. That meant capacity increased by 6 seats from 137 to 143. At the same time, Southwest began taking delivery of 175-seat 737-800s, further adding seats into the system.

All those extra seats, however, only made it tougher to turn an airplane quickly if there were people in those seats. That was the next issue. Beginning at the end of 2013, Southwest saw its load factors increase significantly.

Southwest Year Over Year Change in Load Factor

Lastly, baggage transfers became an issue. As Southwest developed bigger and bigger hub-style operations in cities like Chicago and Denver, it begun transferring a lot more bags. It found that even when people could make the connection in time, airplanes had to wait for bags to make it.

With turn and block times under pressure, on-time performance crumbled. In general, the day would start ok, but flights would often end up being 3 to 5 minutes late. It didn’t seem like a lot, but after 3 or 4 flights, the rest of the day for that airplane would run late under the government’s definition that flights arriving 15 minutes after schedule would show as delayed. By the end of the day, flights could be 45 minutes to an hour late just due to the cumulative mini-delays through the day.

Southwest hadn’t run more than 75 percent of flights on time since last year, but it wanted to be running an airline in the 83 to 85 percent range. To get back to that level, each aircraft line would need about 25 minutes more slack in a day.

A Small Band Aid on a Major Wound
You would think that once this problem had been identified, Southwest would have implemented a fix as soon as possible. You would be wrong. Southwest’s ancient reservations system is so basic that it can’t handle schedule changes. If the airline were to make major changes to its schedule to right the ship, it would have to manually process every single change. That would have been such a herculean task that it wasn’t feasible. (Southwest’s new Amadeus Altea system is handling international flying now but won’t be ready for domestic operation for a couple more years at least.)

Instead, Southwest decided to use as many bandages as possible to try to lessen the impact on travelers. The airline created a program called “Strong Start” to put a big focus on getting every airplane out on time on its first flight. Those were precious minutes that weren’t under pressure due to block and turn time issues.

The airline also put a higher priority on making sure all maintenance issues were dealt with quickly and proactively. In other words, Southwest had to run a perfect airline just to keep its head above water under the tight schedule it had created.

Performance still struggled, so Southwest decided that all it could do was deliver the best customer service possible. It held a lot of airplanes for late connections. It compensated travelers where appropriate. Apologies flowed freely. The end result was that while Southwest sat at the bottom of the pile when it came to on-time performance, it still maintained an extremely low complaint rate with the Department of Transportation.

Finally, a Real Fix
Early this year, when time came to plan the schedule for travel beginning in August 2014 (last month) the real problem could finally be fixed. Southwest reverted back to a schedule with more slack in it to account for the airline’s systemic changes.

On its 737-300s and 700s with 143 seats, the standard turn time became 35 minutes. In the last schedule, Southwest had about 60 percent of its flights turning in under 30 minutes. In the new schedule, that dropped to only 28 percent. All 20 minute turns have been eliminated. The larger 737-800s had been largely scheduled to turn in 45 to 50 minutes, but Mike noted that the latter is more appropriate. Block times were also lengthened a bit so that the airline wasn’t always playing catch-up.

Perhaps most interesting was a decision to increase the minimum permitted connecting time. Historically, Southwest had allowed itineraries with as little as 25 minutes between connecting flights. Beginning this August, that has been increased to 35 minutes. And in key connecting airports like Chicago/Midway, Baltimore, and Denver, it’s now 45 minutes.

That means that connections that used to be legal are no longer permitted. As you can imagine, that will end up hurting revenue, but Mike insists that it wasn’t significant. Besides, fixing the bloated costs of running a delayed operation will offset or surpass any potential revenue losses.

It’s Working
Just how is this new schedule working? Quite well. The new schedule went into place on August 10 and there was an instant impact. The airline really hit its stride beginning August 24. According to masFlight, from then through September 11, it has run on-time 83.5 percent of the time. That’s still behind Delta and US Airways but well ahead of United and American which are both under 80 percent.

In the past, many airlines have used schedule padding to run an on-time operation as a corporate priority. Mike says that’s not what’s happening here. The airline is simply adjusting to a different operating environment. These changes, though not helpful for an airline trying to keep its costs down, aren’t temporary. I still just can’t believe how long it took to fix.

The airline is now working on its schedule beginning April 2015. Mike says that with more 737-800s and the possibility of even higher load factors, the airline may have to add a little more to turn times still.

Will we ever see a return to a tighter operation? In the future, Southwest may look at having different turn times during different days of the week. After all, a jam-packed Friday night flight to Vegas will take more time to turn than a less full flight on Tuesday night. But as always the airline is cognizant of the potential customer impact on not having consistent schedules, and I didn’t come away feeling confident that we’d see that happen.

For now, the airline is going to be paying much closer attention to make sure it doesn’t have a repeat of the last year. It may cost a bit more to run the operation this way, but as Robert Isom, Chief Operating Officer at American says, there’s nothing more expensive than running a bad operation. Let’s hope that Southwest’s miserable year is over, and it’s done running that most expensive kind of operation.

Get Cranky in Your Inbox!

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

43 comments on “How Southwest Tanked Its Operational Performance… And Then Took a Year to Fix It

  1. Over the past several years AA, DL, CO, UA, NW &US were increasing turnaround times at most of there hubs. It went from under an hour to as much as an hour & a half. This was do in part to an increased use of plane changes for continued travel. NW & US employed that tactic the most. Meanwhile, WN gets planes in & out in under half the time & that’s still with an operation with serious challenges.

    1. Some of that has to do with switching from rolling hubs to banked hubs. In a banked methodology, it matters less how quickly you can turn. What matters is how to get planes to land and take off at similar times so you can optimize connections, not optimize aircraft utilization. WN supposedly does not bank its hubs and routinely offers connections in non-hub cities, so it should be turning faster.

  2. “””””Southwest’s ancient reservations system is so basic …..”””””

    Sounds like the opening of a joke which it really is. They should have known years ago as they began to grow bigger and spread out across America, that they needed to improve their system. Instead of spending loads of money to repaint airplanes and change signing over the years, that money should have gone into a new computer system instead of just trying to patch it up. Very stupid on there part.

    1. Considering that Southwest is spending $0 on their new livery, your point is invalid. It is cost neutral considering that it is normal routine maintenance for corrosion control. Secondly, I imagine going from a pure domestic carrier developing into an international carrier is a difficult task considering Southwest was 3 times the size of Air Tran at the point of acquisition. Considering that Southwest will have an infusion of 50 new 737’s in the next few months, their reservation system has to be capable and have the flexibility caused by rapid growth. Luckily, the industry analysts don’t share your myopic opinion.

      1. They had to pay someone to develop and design the new livery. I’m sure that had a nice price tag on it.

        Most of the industry analysts are happy to drink Southwest’s cool aid and look past failures in leadership because they have strong management.

      2. that ridiculous! there new leary and branding will cost money in the long run. reprinting and resignage of all airport stations including double resignage at existing airtran stations and double repanting of former airtran aircraft. complete redesign of employees uniforms and distribution labor In changing and redesigning the current logo and eliminating the old from all software systems powerpoints, res. systems, kiosk software system displays, est. all have some kind of cost involved regardless if it is written off as capital improvement cost to the system. normally an aircraft is repainted every D check inspection (normally in some cases) but now to accelerate this many will needed to be done in C and D checks respectfully for aircraft that have had there fresh canyon blue paint job done less than a year ago (but yet beneficial to those planes that desperately need it as there paint shows slow signs of fading). but really I still think it will have some means of cost involved in the change. not to mention the payments to every designer team from the designer for the aircraft, new uniforms, panting of the GSE equipment, and all others that I mention. imho.

        1. Planedoc, excellent point. This is yet another sign of Southwest managing well, but leading poorly.

          They should’ve evaluated their brand needs at the Airtran merger time, and considered rebranding then, if you’re going to have to paint planes, might as well repaint them in a new livery. Same with all the ground equipment and airport signage.

          1. By their own admission, Southwest stated they had 40 different logos within the company. By rebranding to one logo, they become cost neutral as they finish the supplies of the old brands and rebrand to one. This rebranding is slated to be completed within 7 years and not overnight which is what will lead them to neutrality.

            1. They had to have spent something on this, it might’ve been near what they were already planning on, but they spent something.

              Southwest usually doesn’t do a media day and fly out folks like CF. They had costs preparing the additional videos. They’ll additional costs because they’ll need to repaint the planes they’ve been holding back on painting at a faster clip than they they’d do without doing this.

              They’ll probably end up having to refresh airports a bit before they would’ve.

              Unless you’re an accountant at Southwest you don’t know the costs and neither do I.

              Southwest is saying what Wall Street wants to hear, and as long as the costs are minimal they won’t stick out in their annual reports.

      3. Haywood – Southwest clearly stated last week that there will be costs involved with the new livery. It will, however, cut costs elsewhere to offset the costs incurred. They call that cost-neutral, but I don’t buy it. If you could cut those other costs, you should just cut them anyway.

        Also, most industry analysts that I know have long questioned why it has taken Southwest so long to get a new reservations system. Many times over the last decade, Southwest has said it was going to replace it, only to push it down the list with other priorities taking precedence.

  3. Brett, how can it require multiple years for them to change their reservation system? Airlines have completed mergers start-to-finish in less time than that. I don’t know enough to argue the point, but it just beggars belief that a system that is handicapping the company to this degree can’t be phased out more quickly.

    1. PSS migrations take a very long time and in Southwest’s case, need many customizations to fit their business model. Amadeus hosts most of the world largest carriers, which happen to be full-service international airlines with very different needs than Southwest.

      For example, prior to Southwest, Amadeus’ Altea did not support Southwest’s ‘style’ of open seating (A, B, C, 1-52, etc), they only supported true first-come, first served open seating. So that’s one change to the system that needed to be coded before Southwest went live. If you think about that and how many other differences there are in Southwest’s business model that require changes to the IT system prior to implementation, you’ll understand better why it takes so long.

    2. Amadeus is written in TPF. Not exactly a mainstream technology with lots of programmers to choose from.

    3. Grichard – It’s just a lack of prioritization. They’ve talked about this tons over the last several years, but they’ve missed target dates and then always found something else that seemed more important to them. I find it incredible that this didn’t have a higher priority, but for some reason, it didn’t.

    4. Virgin, JetBlue, Westjet all took over a year to change their reservation systems. It really is a big deal to “change the engine while the car is driving.”

      There is a lengthy proposal process (6 months), lots of system requirements to write (6 months – 1 year), business processes to adjust (included), coding (few months), and planned release which means transferring reservations, tons of new equipment, training, and adjusting schedule / loads to minimize impact. It is a huge deal that takes a lot of leadership and perseverance to accomplish.

  4. I worked in the airline industry for 4 of my 43 years in computer technology. I still cannot figure out why it is taking WN SO LONG to implement a modern system. Taking two more years to integrate their new international system for domestic use seems strange. I know there is a lot of testing and training required and that takes time, but they really need a new system if the current one can’t properly handle sched changes which should be able to be implemented quickly to respond to changing situations.

    I remember going over to a nearby terminal to check availability on our flights (and this is in the early 1970s) to determine whether to take a free flight somewhere. That was decades ago, but I am not sure that Southwest has improved its technology (at least for the mainline systems) since then.

    1. Don – This one has been in place since the 1980s, I believe, so pretty close! What Southwest has done since then it just bolt on bits and pieces; customized it as needed to the point where it looks a lot different than it did in the 1980s but it’s still extremely hamstrung. Crazy.

      I thought it was funny that Southwest kept crowing about how smoothly its transition to Amadeus for international travel was. I should hope so since it didn’t actually have to migrate any bookings. They flipped the switch when Southwest international bookings began, so it would have been pretty hard to mess that up. It’s going to be a bigger challenge to do the domestic one, but I don’t know why it’s taking so long either. They still couldn’t give a date other than saying it wouldn’t be in 2015.

    2. Consider that their choice to remain on SAAS might not be a technical one.

      Perhaps they have a sweetheart deal that gives them the lowest GDS/CRS cost per booking of any airline, bar none, on that platform?

      So, then, switching to anything else, whether it be PSS, Amadeus/Altea, etc, involves a massive increase in costs.

      What you might see, if that were the case, would be a lot of “feet dragging” until the costs of remaining on the old platform (for example, the lost opportunity outlined in this article) outweighed the costs of moving off of the sweetheart deal.

  5. This just looks bad on Southwest’s leadership, and to an extent on their management.

    After the spectacular IT failures that I’ve heard rumored about the CEO should’ve pulled a Bethune and threatened to shoot the CIO if the new system wasn’t up by a certain date.

    I really wonder if Southwest is being managed into a mess by a thousand cuts, and sticking too close to the knitting. A leader would’ve looked at the risks of not upgrading the IT system and prioritized it. They keep wanting to do the same thing as they move into new markets. They’ve failed to get a strong foothold on the Pacific Northwest because their existing formula didn’t work. They had to cut Airtran markets that worked because they would only do things the Southwest way.

    This reminds me of a retailer I’ve worked with: Walmart. You know why its so hard to drive around Walmart HQ? Because they’re all one way streets. You do it Walmart’s way or you don’t do it. They’re finally getting out of that a bit, maybe Southwest will get there as well.

  6. has anyone on here worked an 737-800 rampside on here? my gosh I don’t care what people think of there reasoning for the position of the rear cargo door on that plane but someone needs to be SHOT TWICE for putting the darn thing waaaay to the rear of the plane instead in the mid section like the 727-200. even if the door moves at least 4 stations forward of the existing location that could at least help out in loading operations of that aircraft.

    but anyway enough with my gripping I honestly believe that more focus is needed in accelerating a completely new reservation system system wide to allow for changes on the fly and do international and domestic flights as well instead of doing a whole new logo change of the airline and so on just to make itself unique. however with or without the change the airline is already unique on its own. but I will miss the old crayon blue red belly birds but at least they will perhaps have at least one painted in the crayon blue color perhaps.

  7. AS went through the exact same debacle around 2007. Thankfully they got wise and by 2009 things were back to normal. I am surprised WN didn’t look at the history of their competitors before making this move.

    1. Darin – Not at the top, as far as I know. Mike Van de Ven has been there for a long time. On the network side, I don’t place nearly as much blame. They should try to push the schedule to maximize revenue. It’s up to the ops guys to push back and realize that this can’t be done.

  8. Mr. Cranky…..Great article on Southwest’s operational problems. Good that WN acknowledges their errors and is implementing a recovery strategy. I read your complaint that it took too long for them to initiate their operational turn-around; that it took over a year to fix the problem. Yet you also state that Southwest remained near the top of the DOT ratings for fewest complaints during that prolonged operational meltdown. Since the problem manifested for over a year, why do you think WN’s customers did not complain more to the DOT? If the offending carrier had been UA, AA, or DL, the complaints would have soared. But not for WN. Why is that, Mr. Cranky?

    1. Arubaman – Because, as mentioned, Southwest decided to lay it on thick on the customer service side. They knew things were going to be ugly, they did what they could to fix it, and then just turned on the customer service charm to ease the pain. They handled it very well from a customer service perspective, though I can’t imagine how costly that must have been.

      1. I agree with CF. Two other things:
        1. Southwest has fanboys. People start from a default position of thing everything Southwest does is great and wonderful, they’ve got farther to fall before people will complain.

        2. They’re more O&D focused. If your flight is 45 minutes late, and you still get there 45 minutes late, you’re going to be annoyed, but generally it isn’t the end of the world. If your flight is 45 minutes late, you miss a connection and you get there three hours late, you’re going to be quite a bit more annoyed.

        1. Another thing I experienced is their change policy for late flights (which probably dove-tails onto strong customer service). Last year coming from SNA to SLC, I was scheduled on some last flights of the day back to SLC. Last flight from SNA to LAS and then last flight from LAS-SLC. The SNA-LAS segment was on-time, but the LAS flight, being the last of the day was experiencing a creeping delay of the sort Cranky described above. Every few minutes, I was getting a push from Tripit letting me know it would be later and later.

          At the airport, I went up to the customer service desk and asked what their policy was in event of a delayed flight. I was told that if the flight is delayed more than 15 minutes, you can change for free! Amazing! However, since the SNA-LAS flight wasn’t delayed, I couldn’t yet change my whole schedule. I was told I needed to fly to LAS and change my schedule there. I allowed as how by the time I got to LAS, I wouldn’t be able to change my schedule because there would be no other way to get to Salt Lake.

          So, I asked to see a supervisor. The supervisor was close by, looked at the issue and re-routed me through Oakland. Brilliant! I was home 2 hours before I was originally scheduled. I was proactive, but their change policies (and a little pushing) made it easy to get flights that were on-time. I think most airlines make you stick with your original itinerary unless the flight is more than 2 hours late.

      2. For all those commenting about old res systems, turn times, block times and every other thing; All airlines can get you from point A to point B. Customer service is what wins the customer. WN prides itself on good customer service and is typically the airline with the fewest complaints. Good customer service is the winner every time and doesn’t cost WN anything.

  9. Cranky – nice analysis of the operational issues. A very thorough look at the many contributing factors to the cluster that is airline operations. Having worked on on-time performance as a consultant to airlines in a past life I can attest it is both difficult to get humans from all walks of life (employees, contractors, and customers) to behave/act consistently around the world thousands of times a day, every day, in order to run a reliable operation.

    When Continental went through bankruptcy TWICE in the mid 90’s they realized that operational reliability was CORE to running an airline. Just like AA’s CEO said…nothing more expensive than running a bad operation. It hits you in cost and revenue. Who wants to fly a carrier (in particular if you are a movin’ and shakin’ business person who is really buying time by flying vs. driving, train, etc.) that is inconsistent? If I can’t predict that I’ll get there and I have to fly in a day early I increase my travel costs having an extra night of hotel and a wasted day of my time. Sorry …. next carrier please.

    Gordon Bethune said their high value pax (and all pax) wanted “Clean, Safe, and Reliable” transportation.
    SWA for a long time has delivered just that, producing incredible customer, employee and investor results.
    As they have increased complexity (different seat count aircraft, quasi-hubs, more connections, different fare class/boarding products) its harder to predict operational impacts. 25 minute turns at LGA for instance….forgetaboutit! I’ve waited that long just to taxi to my gate.

    The very difficult balance is to see if an airline can actually improve and retool process to drive efficiency or if it just as to pad the schedule. Easy to get better OTP if you had 10% more block time across the board. But in reality you take a huge cost increase as your asset (aircraft, gate, staff) efficiency goes down (increasing costs and reducing revenue). If you can “sweat your assets” with reliability that’s how you make a F*#@ ton of money. Planes on the ground don’t make money, as we all know.

    I have empathy for the network planners and ops staff at any airline. It’s a bit like being in the furniture delivery business in Manhattan (where I live), expect you make the same deliveries to the same places daily. SO much can change that is hard to predict or out of your control.

    And in the case of SWA, your biggest supplier is the FAA and they are a total deadbeat in terms of managing airspace and rely on ancient thinking and technology that does not “sweat the asset” of the sky.

  10. Cranky, do you have a love/hate relationship with SWA? It seems like you laid it down on them in this article….

    I usually fly the first flight in the morning that SWA offers, so that explains why they’re usually either early or on time.

    1. Tatum – Love/hate? How so? I write the good and the bad for every airline. Sometimes, there’s more “bad” to be written for some airlines, but it all comes around eventually.

  11. Cranky – you posted this for a while but i was wondering if you heard anything about where the capacity that will serve int’l destinations out of houston will come? if i recall correctly, southwest’s capacity has been flat for quite a while, so i’m wondering if they’ll reduce domestic flying, tweak their schedule (as you’ve noted in this post), extend the life of their existing aircraft, or induce new aircraft?

    thanks,

    Industry Outsider

    1. IO – You’re right that Southwest has kept growth in check for some time, but the airline has the flexibility to grow as it sees fit depending upon what it wants to do with its own aircraft coming off lease and on the used market. It has been taking 737-700s on the used market a fair bit lately. My feeling about international growth is that it won’t come at the expense of domestic but will be new capacity.

      1. thanks to both Haywood and you for the response. to take this further, and in light of what they did to add supply, i wonder how this will play out with the ongoing and upcoming bargaining with the labor groups. Will they go back to capacity growth and growth in absolute costs?

        1. for those who may wonder. i’m neither an investor or an airline employee. I just see this as a complex issue, that is capacity addition and labor costs, that southwest will need to address. Initially, southwest seemed to keep capacity in check because of the economic decline but in the last year or so the US economy and southwest fortunes are looking great and yet southwest kept capacity flat. so it seems to me that containing labor costs was one of the reasons for flat capacity, lately.

          1. Well, most US airlines have kept their capacity growth flat or quite low. If there are less seats in the market, it means each of them sells for more..

            1. right. In the case of southwest we know that they are building an international gateway in houston, so there’ll be more capacity in the latin american markets. the question, i think, they’ll want to answer is how does this change, from no growth to growth, affect the ongoing and future wage & benefits negotiations.

  12. Industry Outsider, Southwest has 50 newly purchased used -700’s which will come into their fleet by February 2015. Couple that with new aircraft deliveries and they will be able to expand their schedule without impacting their regular schedule. Aircraft retirements will be limited until 2017 when Max jet launches.

  13. I flew Southwest multiple times during this period and endured lengthy delays. I no longer fly Southwest. I may fly San Francisco -> Dallas next Spring, and I will probably fly Virgin America, even if they charge more and have fewer flights than “Slow-West”.

  14. Fixed? Late again, like every other time I fly out of Vegas past 5 o clock. I’m done, really disgusting, waiting for a connecting flight and well sit around for an hour waiting for a gate at lax. This airline sucks. And they just sent me a text that it’s going to be late, wow, thanks for the heads up.

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