‘We Just Couldn’t Keep Up With the Volume’

In the days leading up to Christmas, the chief executive of Southwest Airlines, Bob Jordan, thought the company would quickly recover from the frigid weather that forced thousands of flight cancellations nationwide. But that changed by Dec. 25 when it became clear that the airline’s problems were spiraling out of control.

After a sleepless night, Mr. Jordan and his team decided that the only way to pull the airline’s operations back from the brink would be to cancel even more flights: around two-thirds of its schedule for several days.

In an interview this week, Mr. Jordan described how things went so wrong for Southwest. The company’s meltdown disrupted the lives of about two million customers, ruined countless holiday trips and left travelers sleeping in airport terminals. Regulators and lawmakers are calling for investigations and stiff penalties against the company.

More than two weeks after the crisis began, the airline is still trying to fully understand why its operations, long considered by analysts to be among the most efficient in the airline industry, failed so spectacularly. It has hired a consulting firm, Oliver Wyman, to study the debacle.

Mr. Jordan also said that Southwest was working with General Electric Aviation to update crew scheduling software, often referred to as SkySolver, that labor union leaders said could not reassign pilots and flight attendants quickly enough after flights were canceled or delayed, even as other airlines snapped back to normal. Perhaps the most daunting task for Mr. Jordan, who has been with Southwest since 1988 but has been chief executive for less than a year, is regaining the trust of employees and customers.

This telephone interview was condensed and edited for clarity.

With the benefit of time, how would you describe what went wrong?

We messed up and caused problems and we need to fix our problems, but it did start with this historic storm. That, coupled with super cold temperatures, caused things that we just never see.

We had jetways that wouldn’t move and we had frozen de-icing fluid. We had aircraft engines that were frozen or iced, and on and on and on. We would suddenly see an issue in Denver, and have to cancel a bunch of flights in Denver. Issues in Chicago; cancel Chicago. Issues in Nashville. You end up with this historic level of cancellations across a number of days. That then means you have a historic level of aircraft repair, which then causes a historic level of crew you have to move.

You’re trying to solve these problems. And as you’re solving them, you have more problems. More cancels, more problems; more cancels, more problems. We just couldn’t keep up with the volume — volume we’d never seen before.

The way an airline works is you want to see something and stay ahead of it. As the volume increased, we got behind.They’re solving problems from an hour ago, then two hours ago, then four hours ago.

Everybody’s talking about SkySolver — which, by the way, is actually called Crew Optimization now, from G.E. — which is an industry standard tool. It’s a great tool. It is not designed to solve past problems. It’s designed to look at a future set of issues and give you potential solutions for crew. .

There’s been confusion over “Well, your technology failed.” The technology did not fail; it worked as designed. Our processes worked as designed; they just were all hit by overwhelming volume. Now, that doesn’t mean that those things don’t need to be better. That’s part of what the Oliver Wyman work is. We’re also working with G.E. on rapidly developing a new release that would allow the software to solve past problems, so they wouldn’t have to be solved manually.

How do you make sure that these systems or just the whole operation doesn’t get overwhelmed in the way that it did?

We can’t have this happen again.

We have terrific leaders and terrific folks in our network operations center that are always watching the operation. But we put in place a whole other level of indicators, basically an operational dashboard that is looking for evidence that the operation is getting beyond what’s normal. I’m calling it hypercare.

Just in case we need it, we’ve got over 100 surge crew schedulers. Their normal job is an analyst in finance, for example. And they’re able to come help.

With the ice on engines, we did not have enough engine covers in certain locations. And that’s something we can fix right now. We had issues with de-icing. So right now we can look at our de-icing equipment and processes and procedures.

Tell me more about how Oliver Wyman is going to help. And what is the company’s board of directors doing?

We’ve engaged Oliver Wyman as an independent set of eyes to go through the event. They’re talking to folks here; they’re talking to folks on the front line in places like Denver and Chicago; they’re talking to our union partners.

The whole goal is to produce an understanding of what happened and why, to prevent this from happening and then understand where we need to increase priorities: maybe move faster, maybe dial up technology spending, add resources.

The board has set up an operations review committee. I don’t want to speak for our board, but the purpose of that committee is to understand what happened; understand how the company and management is moving through responding to the event; provide help and assistance to management as needed.

A lot of employees are frustrated. A lot of customers are frustrated. I think a lot of people want to know who to blame. What do you say to them?

At the end of the day, no matter how we got here, I’m responsible. It’s on me to deal with it and to take care of our customers and take care of our employees and ensure that this does not happen again. That’s 100 percent of my focus.

Your unions say that for years, Southwest has been too focused on investors and hasn’t invested enough in the operation. What’s your response to that criticism?

I’ve been asked a lot about whether we have antiquated technology. We spend a billion dollars a year on technology here at Southwest. And, like every company, there are always things that are brand-new, like our maintenance system, and there are always things that you’re working on.

We have a really good operations improvement plan. Crew scheduling is a big part of that. If you just look back at the last year, we put, I think, eight new versions of SkySolver in. We’ve hired 120 crew schedulers. We put leadership over crew scheduling and realigned ourselves. I’m not saying for a minute that we’re done or that that’s enough.

On the shareholder question, it’s not a singular decision. We are financially very healthy. And because of that, we don’t have to make choices between investing in technology or investing in our shareholders. During the holidays, we went to premium pay, we offered gratitude pay, we did gestures of good will for our employees, just like for our customers.

How do you regain the trust of your customers?

We’ve got a 51-year history of operating a really good airline and having terrific service. That didn’t change just because we had this event.

That doesn’t dismiss the fact that we messed up. I think the best thing that we can do is use this event as a catalyst to be even better. This is a great company, but the best thing we can do for our customers is continue to provide a great product, invest in our product. You’ve seen us do all those things — we’re adding power ports and larger bins. Keep investing in them, keep running a great operation, keep providing terrific hospitality and be even better.

Can you help us understand when you knew things were going really wrong around Christmas?

We plan days and days in advance for storms, hurricanes — all those things. We watch the operation every day; our leaders are on many calls a day just to understand how we’re doing. I’m on multiple calls a day, on a normal day. All that was going on.

When you have a big event, it typically takes an airline three to four days to recover because it just takes that long to get fresh crews into the network — that kind of thing. And until the 22nd or 23rd, it looked like we were recovering pretty normally.

You get into the 24th, and it looks like we’re beginning to have a tougher time. I think it was really Christmas Day that we decided we’re in a different place here in terms of level of disruption. On the 25th, we activated everybody here on site. We took a more radical approach that night to canceling and setting up for the 26th.

When we woke up on the 26th — actually, I don’t think anybody went to sleep — that was the point at which we said, “It’s going to take a larger set of actions to get ourselves back to normal.” You’re thinking a lot of things, but No. 1, all of the time, is be safe. If we need to stop the airline, stop the airline, because we must be safe.

On the 26th, it was obvious that we need to take more action; we weren’t going to be able to operate our way out of this. It’s really hard to decide to take the schedule down 60 percent for three days because you know how many people you’re going to disrupt and how hard that’s going to be. But until you get yourself back to normal, you’re going to continue to disrupt people.

So I had to give our terrific operators permission to take more drastic steps. And we had folks that went away and in hours they redesigned the network to then deploy on the 27th, 28th and 29th. And it worked beautifully. We were 95 percent on time on the 29th. There’s all kinds of movement to get crews and aircraft in place to start the 30th at full tilt. I’m really proud of our folks. Not to dismiss the fact that we really messed up for a lot of customers.