For years we’ve heard Southwest talk about how its reservation system was preventing the airline from doing all sorts of things. International flying and codesharing are the big two that have been publicly discussed lately, but there are more things that can’t be done because of its ancient system. Despite these issues having been raised for years and year, we’re still at least three years away from a new system coming online. At Media Day last week, I sat with Bob Jordan, EVP Strategy and Planning at Southwest, to better understand why this was taking so friggin’ long.
But first, how about a little history? Southwest’s reservation system is known as SAAS. This system actually evolved from the old Braniff Cowboy system. Now, Braniff was an airline that had style but absolutely no network planning abilities. After some insane route expansion after deregulation and a whole host of other issues, Braniff collapsed in May 1982. I was four years old at the time. When Braniff died, Southwest ended up carrying the torch on a version of that reservation system, so yes, it’s old and it’s not built to handle an airline in today’s industry.
Bob explained that while it’s problematic today, for the longest time it was perfect for Southwest. Southwest was a simple airline and as the major user of the SAAS system, it could make any changes it wanted as long as it wanted to invest the time and effort. It was effectively a customized reservation system, one that Braniff agents probably wouldn’t recognize today. But as Southwest became more complicated, so did the reservation system. As Bob explained:
As we started to evolve our business model, we really put our first big strategic plan together in ’06. We looked at, well, we need to think about our boarding, think about new products, think about codeshares, think about international, think about having the ability to do complicated fare rules, stays, and all that. . . . You have to sit back and re-evaulate.
And so Southwest did re-evaluate, but it kept coming to the decision to continue modifying this patchwork of a reservation system to make it do what it wanted to do. Despite the fact that there was a mounting problem with being able to handle everything, Southwest stayed the course. But it did have a project together to try to figure out what to do in the future. That project just kept getting delayed.
The other thing that’s happened in the last 3 years specifically, the project has been re-prioritized. Fuel prices spiked, the economy went into the toilet, and so in light of those facts, we interrupted a lot of things and said, we need to pursue reservations but we need to stop and build some things that will add revenue production now, like Business Select, like EarlyBird check-in, and a bunch of revenue management things under the covers that are being pursued that are generating a lot of money.
So the reservation system problem continued to fester. They continued to modify SAAS to be able to handle more of these revenue projects, but the system just couldn’t handle everything. Yet the project to really address this issue kept being pushed back until this year. Though I thought there was a decision to get a new system long ago, Bob says that’s not the case.
The question became whether to continue to build on SAAS versus whether it would be available on Sabre, on Amadeus. That discussion started early this year and we came to a decision early summer that the right thing to do was move instead of continuing to build. Then it became, move where?
Having finally made the inevitable decision to move to a new system, now Southwest had to pick which one. And that still hasn’t happened. After the AirTran acquisition was announced, Southwest said it had narrowed the finalists to two vendors, and I can’t imagine AirTran’s vendor is one of those. AirTran uses Navitaire’s New Skies system. That system has more capabilities than the previous Open Skies system, but it is still limited. That’s why we’ve seen airlines like JetBlue leave Navitaire for others, Sabre in JetBlue’s case. They need more complexity, and Southwest needs that now too.
But what’s the timing now? It’s going to be a long time. Now that Southwest has to integrate AirTran, this might delay the new system even longer so that they can work to integrate AirTran first. And then there’s the problem of AirTran’s existing international business. How will they handle that in the meantime?
We’re working on that. We’ve got timing options. International works on Navitaire and we could pursue an option that basically continues to use Navitaire for international but connects SAAS to that, or we could pursue an option where we move the international business first to the new platform.
In other words, it could be a mess for awhile. But once the decision on the vendor is made, hopefully soon, and the project is ready to commence, we still have a long way to go.
It’s probably an 18 to 24 month project once you start, probably 24 month range. We’ve got probably up to a year’s worth of work to complete our requirements gathering and planning before we begin to execute so it’s in the 36 month timeframe.
What we now see at Southwest is an airline that will continue to be hamstrung by its reservation system for years to come. Had it made the decision to switch earlier or kept the project at a high priority, then I’m sure this would be much further along, but now Southwest has to live with electronic handcuffs for a lot longer. And that can’t be good for the business, but, well, they set the priority so they apparently feel comfortable with this timeframe.
Photo via Flickr user ajmexico/CC 2.0
Wow– who knew?
Still, dumb question time here.
I remember Braniff (and flew on them quite a bit, since I wasn’t 4 at the time they collapsed), and vividly recall them having international routes, multiple fare structures, and codesharing.
So, how come Southwest can’t do all that with the same system?
Of course, I am not arguing that Southwest doesn’t need an upgrade. And nor do I want MS-DOS back on my computer!
On Braniff, you bet they had lousy network planning. That was due primarily to Harding Lawrence’s ego and his “all-or-nothing” overexpansion of the late 1970s (anyone remember their forays all over Asia with 747s?).
Former Braniff captain John J. Nance’s book “Splash of Colors” about the airline goes into this. It is a great read.
I think this has been answered below from various sources, but first, Southwest took the pieces it needed and has modified since then. This would be unrecognizable to Braniff people. And second, international travel has changed dramatically over the years, so what worked in 1980 wouldn’t work today without significant changes. Southwest has finally (mercifully) decided that it’s time to stop building on top of the existing system and get something new.
Interesting…JM beat me to it…BI and Cowboy managed an extensive Latin American network, plus and insane growth plan into Europe and Asia. BI also had one of the first codeshares (then called ‘interchange’) with AS to ANC and FAI. I also assumed (yea yea, i know the cliche) that part of the FL prize was it’s technology…but Bob Jordan says that is not the case. Hmmmmmm…..
I’m sorry, but the ‘offical’ story here does not add up in my simple mind 0_o
Well done (as usual), Brett! Hope you had a good time at Media Day last week!
It’s there own fault, WN should have changed years ago when they first stated to expand more and more across the country. When they had to start issuing boarding passes instead of those number cards, that’s the day they should have started upgrading their system.
They snooze, they lose. Lucky for them, there’s still time on the clock.
SAAS is a “domestic only” retrograde simplification (“anti-development”) of the original Cowboy.
Does anybody know why the system can’t handle international destinations? All I can think of is that all airports have a 3-letter code and it doesn’t matter where they are from the computer’s point of view.
Its not an issue as to the 3-letter code, etc. Its more a matter of meeting government requirements. When they’re flying international they have to collect and provide to customs things like passport numbers, and a bunch of other information. I don’t know the full details on what they need to provide, but I presume thats the core of it.
There system might not be good enough to handle all the IATA fare construction rules to price international travel or handle all the many taxes that apply in the various countries. It’s not as simple as just adding a few simple domestic taxes like the USA has.
Even the taxes on domestic fares are getting more complex. Because they have to be broken out (Excise Tax, Segment Fee, Passenger Facility Charges, September 11th Security Fee), and legacy res systems like Deltamatic can only store two lines of taxes, so everything gets lumped into XF. To see what that all means, you have to go to the Fare Calc line to see the breakout.
IBM wrote Deltamatic for Delta back in the 1960s, and it always amazes me how it manages to function in the 21st century with all the additions that have been piled on top of it over the years.
Don’t blame the inadequacy of SAAS on its age. SABRE is much older.
Yeah, but SABRE is better maintained than SAAS it seems. Anyone know the vendor for SAAS? or did Southwest buy them long ago?
After BI bankruptcy, Cowboy ended up with American Airlines which renamed it Sabre Automated Airline System (SAAS). It was offered as an internal reservation system to smaller carriers but by 1999 all of them had migrated to SABRE Multihost except for Southwest. AA then spun off Sabre which sold the SAAS business to EDS who was then bought by HP.
Not technically correct – I joined Sabre in July 2000 a few months after AMR spun off Sabre. The thinking and business plan was to go after new contracts with carriers that had a problem with AA/AMR owning the business. In a few months Sabre added a lot of employees and we started working on bids for AC, CO, HP and some smaller airlines I don’t remember. When the business didn’t come in – massive layoffs ensued with a new business plan – to be a software company/services provider and to see off the entire infrastructure (including the realtime TPF systems and the z/OS sysplex) was going to be sold to either IBM Global Services or EDS – EDS won.
Would updating their system allow Southwest flights to show up with accurate prices on other travel booking websites, like Kayak?
I really hope it works with Kayak. With Sabre or Amadeus this should work. Are these GDS’s really the only two choices? Doesn’t HP/EDS also have a GDS option, and is doing something for American? Or Radixx?
Oh they could offer their fares on a GDS if they wanted to, in fact they used to.
Basically Southwest doesn’t want their fares to line up against others, because they’re not as competitive fare-wise as you’d think. Its the same strategy Kohl’s uses, they’re not in the mall, so its harder for people to compare prices.
The Kohl’s around here are all in malls since that’s how they grew by going in when other stores closed. I don’t shop there but did read their clothes are cheaply made.
That could almost refer to some airlines, go in where others fail and offer a cheap product to get business.
Southwest actually does offer its fares via Global Distribution Systems now for travel agents to use. This is primarily for corporate travel agents since often companies don’t give their employees a choice of how to book. BUT, Southwest will not sell fares via online travel agents, so it won’t show up in Kayak. That’s a conscious decision and is not a technological issue. I wrote more about this over on BNET today:
Don’t know if this is true, but didn’t realize until I came across it:
“Southwest Airlines has entered into a marketing agreement with Apollo/Galileo and travel agents are now able to book reservations directly on Southwest.”
Yes, that is true as I mentioned in the response above. Travel agents can book Southwest, but the airline still does not participate in any other online travel agents for consumers to use directly.
Braniff… my very first flight was Braniff ATL-TPA in May of 1980. I was 18 years old. I was graduating from high school and so was my cousin. I used my own money for the trip to go see my cousin graduate. If I remember correctly, I booked the fare by phone and it was about $50. I paid by check as I checked in at the Atlanta Airport. They new terminal had not been open but a few months. The Braniff aircraft was old. Instead of the luggage bins we are used to, the plane had racks that looked like the ones on Amtrak. Everything was yellow from the years of cigarette smoke. I remember that even at 18 I was not impressed. I also remember my mother made me wear my good clothes for the flight.
How times have changed.
Very cool. Took the Flying Colors to and from TPA (my hometown) many times!
There’s *never* a good time to do these things. Do we really think there will ever come a time when someone goes “you know, things have slowed down, we don’t expect anything interesting to happen to change things for a few years so maybe now let’s keep this as a 2-4 year priority”? That’s patently absurd. So either (a) it’s never going to happen or (b) they’ll just have to figure out how to migrate to a new system while continuing for a period of time to evolve the old system. This isn’t rocket science.
SWA had a big internal IT project in the late 90’s to replace SAAS called NewRes. Its failure caused SWA to stay on SAAS and the CIO at the time to lose his job. It was so traumatic internally that the Systems department changed its name to Technology and outside consultants (The Feld Group) brought in to run it along with the above quoted Bob Jordan. Technology reported to him during the years that the switch to a modern res system was a low priority.
I don’t buy the argument that SAAS can’t do codeshares. It did them with ATA just a few years ago. If not SAAS, Southwest had *something* that made it work.
From my understanding, SAAS is hosted by Sabre.
ATA carried the SWA code; not reciprocal.
I seem to recall that you could book through WN for ATA flights as well. I don’t now for sure since I didn’t actually do it. I’m pretty sure I remember being able to pull it up on southwest.com.
My understanding was SAAS was maintained and operated by Sabre (they bought the BI Cowboy system?)
This is more a question of cost vs. benefit, WN does not want to invest additional resource to the ‘old’ system. WN are smart business people, speculation – they may have a great deal now and know where their next best value will be, or have a target on spend/cost that the GDS’ will need to meet.
WN have changed the paradigm before, have time, choice and are flexible to try a new alternative as well. (Possibly test & analysis of the Navitaire New Skies system with FL operations?) May be their best value option to date…all systems can be enhanced/integrated with today’s technology. Of course a good SLA, price, support, maintenance and reoccurring fee structure could get – keep the WN business. No handicaps ahead short or long term — just more options (at a cost.)
Yeah, it’s operated by Sabre. I was at World Airways at the time of Braniff’s demise first demise in 1982. Cowboy hosted at the time World, Laker Airways (remember Sir Freddie’s airline?), New York Air, and I want to say People’s Express, plus some others I forgot about. When Braniff went kaputt, all of the airlines hosted on Cowboy moved to CCS/SHARES in 30 days..crazy!! CCS was the wholly owned subsidiary technology arm of Continental. Interesting tidbit: CO had bought Braniff’s check in system several years before and plopped it into SHARES, so at least check in was kinda similar.
AMR picked up Cowboy from Braniff (which was outsourced to EDS, so maybe technically they picked up Cowboy from EDS?). And of course, later on CCS merged with System One when Lorenzo bought Eastern, and dropped the CCS name.
International operations 20 years ago are nothing like they are today, heck even 30 years ago. We now have SecureFlight, which requires we send passenger data to the TSA the moment you book, and we have to be able to receive a response from them; we have to send our APIS – your passport info, citizenship, to CBP at least 30 minutes before departure now; we have to do Positive Person Bag Matching; and much much more automation than it was in 1980, when the flight crew would simply hand write a Gen Dec (General Declaration for customs) and then a passenger manifest to customs.
And regarding Southwest’s distribution, its expensive to sell seats via the GDS. You have multiple charges per FLIGHT and per BUCKET for updating inventory, as well as the messaging involved with booking that flight.
In my latest analysis for another airline, we found out it cost us over $18.00 to do a GDS booking (for one passenger)- that included our hosting fees, ARINC fees, GDS fees, ARC fees, revenue accounting fees, merchant account fees, etc. And this was off a $98 roundtrip ticket.
Even with WN’s volume, you can’t avoid most of those fees.
And international taxes are no problem, those are easy to add into any system, and the IATA MPM pro-rate intl fares wouldn’t apply to Southwest as they aren’t in MITA nor do they have any IETs – Interline E-Ticket Agreements.
SWA’s home grown dispatch and planning (SWIFT) is also in need of a major upgrade. This is also holding SWA back from international operations. SWIFT needs replacement and/or major surgery to manage a mixed fleet (not to mention ETOPS).
Great information in this post, CF. Looks like the Airtran merger will affect the timeline, too. From the recent Q3 earnings call transcript at seekingalpha.com : (Gary Kelly) “So, in any event, that the attributes associated with a new reservation system likely won’t begin to come online for probably 18 to 24 months at the earliest, and then all up, I can’t imagine that we’ll be completely converted to a new reservations system given this new AirTran initiative for three year would be my guess.”
SAAS from BN Cowboy? I was involved in alot of CRS work back then, doing various airline startups. When we started Discovery Airways in Hawaii, SAAS was our CRS vendor, and we worked with SABRE’s team in Dallas. At the time SABRE was a wholly-owned part of American Airlines, not Braniff.
SAAS was (is) an acronym for “Small Airline Availability System”. It was indeed geared to small airlines, and could not display carrier availability to/from points outside the US. While they had the ability to handle complex fare rules, and worked directly with ATPCO (Airline Tariff Publishing Company), the system was otherwise very limited in it’s functionality.
Covered above. SABRE bought BN Cowboy in the early 80s in time
for it to become SAAS in 1989/90 for Discovery Airways.