The other day I found myself thinking about Amadeus IT Group, which has recently acquired Sky-Link to gain access to AI capabilities. I must admit that I have been around for many years and have experienced the entire journey – from the pre-GDS era to today’s AI world. There is no doubt in my mind that Amadeus and its investors (private equity funds) are extremely capable. But the real question is whether monopolies – or near monopolies – are good for industries and society. 

Amadeus was originally built on mainframes, EDIFACT messaging andsegmentfees, a technological architecture developed more than 50 years ago. Back then airlines had internal teams  teams capable of developing and using technology themselves. But weak airline finances in the 1990s “forced” the airline owners of Amadeus to accept a very attractive offer from private equity firms BC Partners and Cinven. 

Today Amadeus is publicly listed in Spain. 

And yet – in the middle of cloud computing, APIs and now AI – this “old structure” still plays one of the most central roles in the global travel industry. That is actually quite impressive. But also thought-provoking. 

If you ask Amadeus about monopoly power, they will immediately point to their competitors: Sabre Corporation and Travelport, as well as TravelSky Technology in China. That is true. But TravelSky is significantly cheaper than the others, and Amadeus’ global market share is far larger. 

The Old Engine of the Travel Industry 
Global Distribution Systems were created in the 1960s and 1970s, when airlines began digitizing reservations. The systems were built on: 

• mainframe computers 

• EDIFACT messaging 

• booking terminals at travel agencies 

When a flight was booked, a segment was created. And every segment generated a fee.At the time, this made perfect sense. Computing power was expensive and the infrastructure required enormous investments. What is remarkable is that the model has never really changed. Even today – more than half a century later – the segment fee remains the core of the distribution economy. 

The Toll Road of the Travel Industry 
Segment fees function in practice as a toll road for the travel industry. Every time a booking passes through a GDS system, the airline pays a fee – typically between USD 4 and USD 12 per segment. Multiply that by hundreds of millions of flight segments every year and you get one of the most stable revenue streams in the global technology sector. The most interesting part is not that the model exists.  
The interesting part is that almost nobody in the industry seriously challenges it. There is constant talk about disruption, and new companies regularly appear trying to sell travel in new ways. Yet most of them eventually end up working with Amadeus. 

The Quiet Consolidation of the Industry 
Over the years Amadeus has also strengthened its position – not only as a distribution system, but as core infrastructure for the travel industry. Through acquisitions it has gradually collected different parts of the value chain. Examples include: 

• Navitaire – reservation systems used mainly by low-cost airlines 

• Skylines – software for airline operations and planning 

• TravelClick – technology for the hotel industry 

As a result Amadeus now provides technology to: 

• airlines 

• airports 

• hotels 

• travel agencies 

• online booking platforms 

A large share of the world’s travel transactions therefore passes through Amadeus systems at some point. This may not be a monopoly in a legal sense. But it is very close to infrastructure. 

The Patience of Capital 
Part of the explanation lies in the capital behind the company. Large investors and private equity funds have been willing to think very long term. This has allowed Amadeus to: 

• invest heavily in technology 

• acquire competitors 

• expand its platform 

• maintain the existing revenue model 

The strategy is actually quite simple: Control the infrastructure – and the market will follow. 

AI – The Next Wave 
Now the travel industry faces another technological shift: AI-based travel agents. 

Systems that can: 

• search across suppliers 

• compare prices 

• assemble complete trips 

• transfer customer data automatically 

In principle, an AI agent could build a trip directly between traveler and supplier. Without traditional distribution systems. And without segments. 

Or Maybe Not 
History shows something interesting. When new technologies appear, they often do not destroy infrastructure. They become part of it. 

They are integrated. 
Acquired. 
Or embedded into the platform. 

And the infrastructure becomes even stronger. 

The Golden Cow 
The segment fee has already survived: 

• the internet 

• online travel agencies 

• metasearch 

• mobile booking 

• NDC 

The question now is whether it can also survive the AI revolution.  Because if an AI agent one day can book travel directly between travelers and suppliers, a very simple question emerges: 

Why should a system designed in the 1970s still collect segment fees? 

Of course this would require suppliers – airlines, hotels and others – to invest and build their own distribution and sales structures. 

A Changing World 
The world is going through massive changes. AI companies themselves are likely to move toward control and monopoly-like structures. I have heard many arguments that only the largest players have the strength to compete in the race for the customer – and that monopolies may therefore be inevitable. 

Personally, I have helped build not only a unique product but also the entire operation of a corporate travel agency. We used all channels – including Amadeus. And like many others in the industry, we complained whenever they made changes. But I have always believed that large organizations often create more pseudo-work than real progress.