I listened in to the Back on Track Belgium conference last week, and the founder of a startup night train company said something interesting. “The reason we are against things like night train rolling stock pools,” he stated, “is that means we cannot configure the trains the way we want them“. We could not choose the colours or the upholstery for example he went on. I’ve heard this from other small players in the night train business in the past – if new trains are to be ordered, they are going to have to meet our exact specifications.

The problem is that if players in this business stick resolutely to that line, the night train conundrum is never going to be solved. And night trains are forever going to look like the one pictured: quirky, but not scaleable and not a mainstream transport mode.

I do not disagree with a potential operator’s desire to configure a product, but the question is how much, and what that does to the economics of procuring new trains in the first place.

Imagine a startup wants to run a deluxe night train between Paris and Venezia, assuming that there are plenty of high wealth tourists ready to take the trip. But that 5 years down the line it turns out that Paris to Venezia is not profitable, and the market potential for ski trains between Berlin and Kitzbühel is a better bet – and that requires more compartments, but smaller ones. Can assets designed for a Paris – Venezia deluxe train be redeployed for a more basic Berlin – Kitzbühel service? If they cannot, getting finance to make the order in the first place is going to be a lot harder.

And then imagine that to rebuild the carriages for the new lower budget purpose changes the weight balance of the carriage, and so a whole new costly approval is going to be needed. Even more risk.

And is if that is not enough, what about if each carriage for each class of passengers needs to be designed differently from the start – with larger compartments and wider windows for higher paying customers? And it is impossible to know at the start how many deluxe and how many more basic carriages are needed at the start? Oh, and each of these carriage designs might need a separate costly and time consuming approvals process.

You get where I am going here. Make the design of the train too complex and too bespoke at the start and you are unlikely to be able to get the finance for the acquisition in the first place – because the product will not be adequately standardised so as to be able to redeploy it on another route if your first attempt fails.

Starting small and experimenting is also not really an option – were a startup to order 30 carriages the cost per carriage is going to be high. Order 300 and the cost per carriage – due to standardisation – is going to be low enough to make the operation economic.

So the crux as I see it is to standardise enough. Keep all the basics very standardised – the carriage body, bogies, electronics, windows, plumbing. And beyond that – in terms of the beds, upholstery, small fittings – allow as much customisation as possible. The line between what is and what is not customisable is determined by homolgation – only things that do not need a new homolgation process are what the operators can themselves customise. Sure, that might mean that some startup might not get the windows quite as wide as it wants them, but it will get the majority of what it needs for a decent price, whereas currently the operator gets nothing of what it wants because no finance is forthcoming for a product that is too bespoke, and does not benefit from standardisation and economies of scale.

Who, I wonder, is going to realise this in the night train sector? Strikes me no operator (other than ÖBB somewhat) has understood this yet.


  1. M@trixX

    I believe we need some kind of experimentation phase, what is not feasible with a rail pool ordering a lot of the same carriages. Wast majority of today’s sleeping and couchette cars are pretty much the same as they were in the 1950s with little to no innovation and I would argue it was one of the factors leading to the decline of night trains in Europe.
    For example, we can be pretty sure hostel style couchettes shared with 3-5 unknown passengers are not very liked. They are tolerated by budget travelers, but I’m not sure there’s market for hundreds of those (new). Yeah, OBB understand that and they came with the new designs, but so far we don’t know whether those would work or not. What if people hate the couchette coffins? Or if there’s too many of the four-bed family compartments? Should we test it with a few carriages? Certainly. Should we order hundreds of those and “force” everyone to use them? Hell no! Even the OBB, backed by government and taxpayers’ money, ordered only 66 sleeping and 99 couchette cars (both closer to 30 than to 300).
    If companies like European Sleeper were sure they wanted old-style sleeping cars and the only difference would be the color of some fabric inside, it makes 100% sense to order them together with other companies and they can even be owned by some third party. If they don’t want the same thing, like french Midnight Trains with their vision of business class hotel on wheels, it makes no sense whatsoever.
    The future of night trains will be formed by those experiments and in 10 or so years, when we tried something and we know what modern night-time traveler wants and likes, then it’ll be the right time to order hundreds of the same carriages to lease.

  2. M@trixX

    Can you link some of those experiments? I don’t think there were any given there are next to no night carriages that are not over 20 years old and we’re still waiting to get wifi even on the hyped nightjet.

Leave a Comment

Your email address will not be published. Required fields are marked *