In my last post, I promised to share some lessons gleaned from SABRE’s 50+ years of history as a B2B transaction platform. Many of these lessons are useful for today’s B2B entrepreneurs.
Here are some key points gleaned from Sabre’s history and prospectus.
1. Sabre moved from a closed/biased system to an open decision engine.
As originally conceived, Sabre was American Airlines’ passenger booking system. It was a proprietary system which showed only AA flights. Even when it “opened” to show other companies’ flights, the Sabre display algorithm showed AA flights first and relegated United flights to the second page as competitive weapon. (AA learned long before Google that no one looks at the second page!) This systemic bias, and complaints from competitors, eventually led the US government to force AA to eliminate the bias.
Over time, Sabre became a more true decision engine, allowing flights to be displayed according to user requirements. This same lesson of the importance of unbiased search played out again in the early search wars that helped paved the way for Google. Users will put up with ads, but they hate the idea that the search results themselves are driven by the platform’s competitive issues or business model, not their best interests. (My new favorite example of unbiased travel results is from the travel site, Hipmunk, whose default ordering of flights is by “agony”–a combination of price, stops, and total duration!)
2. Serial Verticalization led to a complete solution.
Sabre started as an airline booking tool. But what the user wants is a complete trip, not just an airline reservation. Sabre now provides the ability to book hotels, cars, rail, cruises and most major elements of a trip. I’m not sure if they make as much on the non-airline elements, but they needed to enter these businesses outside their core to meet the user need.
3. Value-based pricing helped drive usage.
We can argue about whether a 2-3% net yield is high for the services a Global Distribution System (GDS) provides, but from the start the pricing was transactional–tied to value/usage. To drive participation, Sabre also subsidized the buyers and charged mainly the suppliers. Sabre may have had minimums for travel agencies (and they definitely still have volume discounts and rebates to drive users exclusively their system) but customers could get started on the system cheaply and grow with it.
4. Becoming an industry platform is a long slog that never ends.
Sabre has been around 50+ years. Sabre and its competitors have changed everything about the way we book travel, but they still represents a small part of all bookings. Consumers can still book from almost every carrier directly and there are still online and offline travel agencies–remarkably in some ways. In an interesting irony, American Airlines is trying to offer some online capabilities only through its website and not through GDS customers, including Sabre! B2B industries take decades to change and dominance of a platform is always relevant to a particular segment or portion of an industry. The industry is still complicated–it does not form a straight line from airline to GDS to consumer. Here’s an informative chart from the prospectus:
Note all the hidden ways that fees move around and how fragmentation differs at different levels in the value chain.
5. Concentration leads to antitrust concerns.
Even though there are still many paths to booking a ticket, there are just three GDS systems worldwide. Network economies are often overstated, but they do exist and exert force in favor of fewer GDSs. For the same reason, there are only a few payment networks, a few major search engines, EDI networks, etc. And where network economics exist and are exploited, they eventually draw antitrust scrutiny. In some ways, antitrust worries are simply the price of success. Sabre, for example, is still paying out a lot of money for past practices. Owners of these types of businesses have to carefully watch how they conduct themselves.
6. Core commerce activity is still king.
It’s easy to get excited about all the value-added items a marketplace can provide with the data and the applications it can add, etc. EDI networks tried to add applications, payment networks as well. Ariba and Tungsten/OB10 have been adding payment and financial services. But platforms must always remember to invest in the core commerce business. Note that 50+ years later, Sabre is generating 80% of its EBITDA from the core Travel Network.
7. Structure is related to function.
Sabre would not have developed as the industry platform it became as a wholly owned subsidiary of American Airlines. You can hear the parent company saying to the developers of Sabre “What do you mean you want to offer United reservations on the first page?!”. Congress forced Sabre to evolve and American Airlines took the subsidiary public in the 90s. Eventually Sabre was spun off in 2000. Now Sabre is back doing battle with American on their enhanced direct booking product. The EDI folks inside GE, AT&T and the network folks at Ariba all learned the same lessons. Industry platforms are just that, not company platforms.