TUI and the Blockchain

Earlier in the summer, leisure travel giant TUI released a corporate video showcasing their exploration of blockchain technology as a means of managing their committed bedstock, an application they’ve dubbed ‘Bedswap’. CEO Fritz Joussen had previously attracted attention by commenting on the potential for blockchain technology to replace the likes of Expedia – disintermeditating travel distribution – and claimed that this threat to the bedbank model had been one of the drivers behind his decision to divest Hotelbeds last year[1].

It’s perhaps slightly curious that TUI would publicly showcase what presumably is intended to be a strategically vital initiative, rather than beavering away in secrecy. Perhaps they’re signalling their confidence in their market dominance, or perhaps there’s value in publicising their bleeding edge credentials to reassure their stakeholders and keep their share rating nice and buoyant. Or maybe they’re simply embracing the transparency and openness that characterised the early cyberpunk peer-to-peer movement. Who am I to say?

In any case, this sounds like industry-transforming stuff. What’s going on?

We wrote about the blockchain concept a couple of years ago. The key underlying principles are:

  1. The technology, first-and-foremost, solves the ‘double payment problem’ – the potential to re-use a digital asset in the absence of a preventative control.

  2. There is a built in mechanism, such as ‘proof-of-work’, which makes the block chain very robust against attempted subversion (‘Byzantine fault tolerance’).

  3. The need for an intermediary, a ‘trusted third party’, to adjudicate transactions is thereby avoided. Hence Bitcoin needs no central bank (or any other kind of bank).

‘Bedswap’, as far as we can tell, is conceived as a private blockchain application. Essentially that means that it is intended to be an in-house database that uses the first feature above to manage the allocation of committed capacity (beds, in this case). No overbooking, and the potential for a very granular approach to pricing and yield management.

On the other hand, the second feature above breaks down if a majority of the blockchain users are working in collusion (‘the Byzantine Generals Problem’). The distributed, proof-of-work mechanism is also extremely inefficient and slow by design, with masses of computing power devoted to (basically pointless) brute force calculation. None of this is necessary if there is a controlling party, in this case TUI, so why would you do it? It’s arguable that a private blockchain application isn’t really a blockchain at all – it’s just an ‘add only’ database. So maybe there’s less to this than meets the eye.

On the other hand, Joussen’s remarks about the potential end of what he calls “monopolistic structures” in dominating travel distribution look far more radical. Is he right?

Let’s brush aside the point that there are several of these ‘monopolies’, Expedia,, and so on.  It’s true that these platforms become strategically effective through network effects: the bigger they get in terms of suppliers and buyers, the more value they have to both sides of the network. That in itself wouldn’t be solved by a global, open-source travel (or hotel) blockchain – it would still be monopolistic in that particular sense. The assumption though is that this would be a much cheaper form of distribution, because there’d be no dominant commercial intermediary able to extract high margins from suppliers. Control of the blockchain would be widely distributed.

But that assumes that the problem that OTAs (On-line Travel Agencies) and bedbanks are solving is that of ‘trusted third party’ – making sure hotels (for example) only sell a bed-night once, and perhaps managing the payment for the service. And to an extent they do. But that’s not what they’re for.

First and foremost OTAs and bedbanks are aggregators of inventory. They make suppliers visible to customers in a curated and easily searchable way, and they accumulate customer eyeballs to give suppliers a better chance of reaching them than relying solely on their own PPC and SEO activity. If neither of these conditions were true, Expedia wouldn’t exist. And blockchain technology in and of itself wouldn’t address this issue.

Secondly, the more powerful OTAs act somewhat like supermarket chains in moderating supplier prices to the benefit of consumers (and their own margins). They are monopsonists (or more strictly, oligopsonists), not monopolists. Again, blockchain technology does not, in and of itself, address this issue. TUI in fact have a long-term plan to integrate their front-end systems with Bedswap in order to facilitate granular, ‘unique-customer-to-unique-bed’ yield management. They aren’t doing this to drive retail prices down.

It’s certainly possible that a global travel blockchain could be used to handle booking and payment. Google may be as well placed as any actor to facilitate such a thing. But in and of itself it won’t disintermediate OTAs. That would need something else.

And Google may be as well placed as any actor to solve that problem too.


Thanks to John Charnock for reviewing and commenting on a draft of this article.



[1] Tnooz, 23 August 2017