The Tacoma, Washington-based newspaper, the News Tribune, did a great piece on how Alaska Airlines fixed its reliability issue over the last several years. In short, Alaska used to be pretty bad in on-time performance, ranking towards the bottom of the list of large US carriers. For 2009 and into 2010, Alaska has turned it around dramatically. Those of you who follow these rankings would see Alaska near the top of the list -- generally just behind Hawaiian.

Here's a link to the article:

The article raises an interesting chicken and egg question. Did Seattle become a great on-time airport because of Alaska's efforts? Or did Alaska benefit in part from something at Seattle?

From 2005 to 2010, Alaska's on-time performance went from 71% departure / 70% arrival to 88% departure / 86% departure, ranking #2 in 2010 YTD. A very impressive climb highlighted in this article.

However, Seattle as an airport had its on-time performance go from 77% / 75% to 88% / 84%, also ranking 2nd in departure performance (and 3rd in arrival performance).

Strangely enough, the article neglects to mention that Seattle opened a third runway in November, 2008. According to the Port of Seattle press release here, the runway was intended to cut average delay minutes in half, with a starting point of 10 minutes (actually more like 11 in 2008 total). The runway did help, average delay minutes went down to 7 minutes in 2009, and less than 6 minutes in the first quarter of 2010.

So who is taking credit for this -- Alaska Airlines, or the third runway? Both, obviously. But who cares, as long as the on-time performance improvement is there!

Alaska's improvements does outpace Seattle, but Alaska has enjoyed a ~30-40% market share at Seattle during this time, meaning it's hard to know for sure which one is driving the other. One stat we can look at to try and understand better is the percent of delays by cause. If you look at this stat for Seattle going back to 2005, you see the percent of delays that are caused by the airline have remained between 29% and 36%, hitting 33% in 2009 and 36% in 1Q10. But since overall delays are plummeting at Seattle, the percent of overall flights delayed because of the airline has also fallen, from nearly 8% in 2005 to under 4% in 2010 so far. Notably, this percentage has also dropped in half for NAS delays (6% to 3%) and Late Arriving Aircraft (9% to 4%), the other two big buckets of delay causes, leading credence that perhaps the runway has been a major contributor to lowering delays across the board.

In conclusion, it's hard to tease out Alaska-driven improvements, although you can't argue with results, and Alaska's on-time performance surely allows us to credit them with some major positive changes in the last 5-years.

Aside from highlighting the outcome of a better operation, the article is a great case study in airline management and how it can achieve desirable results by focusing resources on an initiative.

Here are some of what I think are the key take-aways:

1) Despite my general argument that an airline's on-time performance is dependent more on the airports they fly into and the flight time than anything else, this does show the amount of power an airline has to help their operation. There is a theoretical limit to on-time performance driven by weather, congestion, etc. But how close an airline gets to that limit is clearly dependent on doing the things that Alaska is talking about here. These actions also help significantly when delays are caused by other things since the procedures are in place to handle them most efficiently.

2) Being on-time costs money. Not only did Alaska talk about investing in big screens (several per gate), they are also talking about choreographing their crews in a way that makes them subservient to the plane. In other words, the aircraft is the constrained resource so you should manage your work crew such that it is always ready to service the aircraft (never have an aircraft waiting for crew). In an ideal world, this is the proper procedure since your planes generally cost more than the incremental people-time.

3) The state of operations is both more and less sophisticated than you might think. On one hand, Alaska talks about all these systems, tools, procedures, and technology to keep its hub running smoothly. On the other hand, just a few years ago ramp workers might not have had good or consistent info on when an inbound aircraft was going to arrive! And if you think Alaska is being innovative here for a relatively small hub operation, you can imagine what some other airlines might be doing.

This is a great lesson in the power of having a team work towards a focused goal. They set out to do something, were willing to spend a bit of money along the way, and came out with a system that works better, makes employees and passengers lives easier, and contributes to the company bottom line.

Large and complex companies have tough problems that require innovative solutions -- but they also stand to reap the large benefit of those solutions. Kudos to Alaska for allowing the public a view into what they did, and here's hoping more airlines will aim for the "theoretical" limit of on-time performance, weather and traffic delays aside.