13 November 2009

The OpenGDS - A New Dawn

Reprinted with kind permission of The Beat:

The Beat ~ a travel business newsletter
New York City
11/9/09 2:32 PM

For many of us who have been involved in travel distribution, it has been an open secret that the global distribution system technology at the core of distribution was if not obsolete at the very least obsolescent--a case that has existed from at least the 1990s and, in my opinion, from before then. This was driven both by the business model of the legacy GDS as well as the strictures of the GDS core software based on IBM's venerable TPF [Transaction Processing Facility] operating system.

Over time, the pure TPF core was supplemented by auxiliary processors, Web front ends etc. IBM, too, started to move further and further away from its core business of application software and hardware to a more services-oriented model. In recent years we have seen IBM pull back from providing functionality common in TPF to being a pure operating system and coupled hardware provider. This caused many TPF users (not just airlines but banks as well) to migrate to more open systems such as Unix and Java-based programming for both cost and functionality reasons. There was considerable grumbling over the most recent upgrade to the Z class processors and the associated zTPF operating system product from IBM. The reason was that the "upgrade" created little or no additional capabilities for the user community.

The traditional GDSs were happy to pick up the development of new functions as they tried to differentiate themselves in the fiercely competitive market. Amadeus for one at its inception never included fares and pricing within the core TPF environment, and that gave them a certain flexibility in their fares products. Sabre had a heavy investment in DEC Vaxes (remember them!). So the move away from the holistic TPF based environment was an inevitable progression when the Web and its associated technologies came along.

Finally, we are beginning to see the results of these efforts in the GDS functionality. Last year Sabre in its briefing sessions on the new profile system outlined a vision of the open architecture which will see a flattening out of the capabilities of the GDS. Last week The Beat reported on what has been another open secret--that Travelport is doing the same thing with its now named Journey Manager and associated universal desktop and new profiles. Amadeus is likely not to be too far behind here.

I will not be presumptuous to describe how each player is now going down the path, but I believe it is now appropriate to state that the legacy GDSs are finally moving to what I call the Open GDS model--a novel concept when stated, but it becomes obvious when you reassess the world through open eyes. It is quite interesting to see how each of the major players is now clearly inching towards this concept and in some radically different ways. However, there are some common characteristics.

• New front end user interfaces/experiences and API/Web services for access
• Unbundled, open ports for supply access
• Advanced customer file management including PNR and Profile services
• New application frameworks

Coincidently, the emergence of credible GDS alternatives using software-as-a-service-based architecture (what used to be called "GNEs") are finally here and, more importantly, in production. Companies like Farelogix have become core to this shift to pry open the GDS model. This is now leveling the playing field with different supply players such as American Airlines, Emirates Airlines and Lufthansa--amongst others--taking legacy GDS Lite or Free approaches to both the commercial model and the technology. Other software providers such as Datalex and OpenJaw now are providing major functions that previously were the exclusive purvey of the legacy GDS world. In fares, too, we are seeing a next-generation of players independent from the GDSs, such as ITA Software and the Microsoft-Worldspan jointly developed ePricing capability.

I would be remiss if I didn't also state that the large online travel agencies have had all this capability for quite some time. So, the ability of an OTA to connect directly to the supply chain and in effect support their own versions of the Open GDS is really already in place. Orbitz before its acquisition by Cendant had already deployed its direct marketplace solution. The OTAs' white-label booking engines are clear examples of this.

While much of this is not new, it is clearly not a revolution but an evolution of the GDS model. And it comes none too soon. However, there are benefits and challenges that come along with this new world order.

The Open GDS model shows that responsibilities for distribution are now more evenly matched at different ends of the value chain. We will now see differing models emerge commercially to match this new open environment. The one-size-fits-everything model is, as I have written before, now clearly on the way out and the coffin has several nails in it already. The Open GDS model is a natural, and in my opinion unavoidable, result of fragmented content. I want to make clear that this was not a chicken and egg story. Fragmented content was always there. The GDSs just didn’t like to accept that fact.

Both ends of the distribution value chain now need to recognize that there are new realities. The suppliers now need to adopt better tools for distribution management, a role that used to be exclusively the domain of the legacy GDSs. Now no more. For the intermediaries and sellers of travel products, multi-source supply is not a nice-to-have but a must have, and in most parts of the world already a fact of life. New systems to support a more bilateral versus multilateral model of supplier relationships must be implemented, and quickly. This cannot be just a bolt-on capability; it must be at the core of the intermediaries' capability.

For the traditional GDS companies in the middle, these are new challenges--not least of which is how to demonstrate that their value remains consistent to suppliers and sellers alike. It would seem that the legacy GDSs are getting poked and prodded from all sides and seemingly all at once. They have fought a good battle, but now it's time to stop the rear-guard action and welcome the 21st Century. The next round of GDS/airline PCA (participating carrier agreement) negotiations will change the commercial landscape of the distribution market. The world will be a very different place in 2013, when the re-contracting season will come to an end. That season is now upon us and the battle lines are drawn.

Welcome to the new dawn. Welcome to the age of the Open GDS.

2 comments:

Anonymous said...

Ah, yes. Just like when the live agent was going to be replaced by The Information Superhighway. Are you one of those people who "predicted" that all workplaces would be paperless by 2001, too?

Your arguments are not based in marketplace realities, they're based in more of the same hand-wringing, Pollyanna future fantasizing which consistently pollutes the discourse of our industry.

GDSes are the most cost-effective aggregators of worldwide travel content, allowing travel buyers and sellers to easily compare and book CRS inventory. If you can prove out a survivable, cost-effective way to aggregate 750ish airlines' inventory and a year's worth of room nights from most of the hotels in the world; it'd be an interesting use of blog space.

Go for it, man. Paint us the business case of the Wild Awesome New World of Open Whatever.

Professor Sabena said...

Professor Pollyanna - I like that!

Seriously, though - your points are all valid if a little off. The GDSs are the current cost effective aggregators of SOME content. The leakage of content out of the legacy GDSs over time has reached critical mass. If the #2 airline brand on the planet doesn't participate at all and now many of the top 50 don't participate fully then there is a problem. I would suggest that there is at least 50% of the world's hotel inventory is not bookable via the GDS and a further percentage is only partially available. As for the 750 airlines - I would beg to differ on the number. But the issue is more related to the depth/quality of content and the problem of what to do with the lack of full content.

The market is voting with its wallets and its investments. This is not a question of replacing the GDSs entirely. But their role is changing and the drive comes from the availability of lower cost/better technology and the over pricing of the product by the legacy GDS companies. This is not an either/or - it is an and. Of course if the GDSs were to lower their fees then the Airlines could rethink this. But I believe that this is already too late. The airlines are sometimes slow to move but when they do it becomes inevitable. The lack of commercial functionality is palpable in the legacy environment. Just look at fares.

So yes - there will be a change. Even the legacy GDSs are starting to adopt the OpenGDS model.

Cheers