It’s a common question I receive on social media, and from friends offline too: “can you recommend me a platform to book rail tickets Europe wide?”
And the answer is, no, I cannot. Not universally.
The answer to the question is normally: “can you tell me roughly where you’re going? So then I can recommend a tool or tools for you“.
And I say this having gathered a hell of a lot of experience in European rail, and having tried a hell of a lot of different tools.
But why we need to fix this problem is obvious: if we want passengers to take the greenest mode of transport – the train – we cannot allow cumbersome booking systems or lousy route planning to put them off. Sure, railways might have a slew of other problems too, but can we not at least fix the aspects we ought to be able to fix comparatively easily?
So what would the fix look like? A rail planning tool to win out over them all. Utopian perhaps, but this is what I’d like to see exist.
It starts with a good route
Planning a railway journey has all kinds of parameters that are more complicated than a flight. Do I want speed? Do I want to avoid changes? Do I want a scenic route? A route via some place or another? Does I have some need that is only fulfilled by some type of train?
Luckily in Europe for rail we have a pretty good route planning tool – bahn.de*
I can – reasonably reliably – plan routes in most of Europe with it, and it will serve me up good quality results according to my criteria. It can plan me more complex routes with more reliable data than Google Maps, and it can plan me more complicated routes than some other national planners can (it’s better for France than SNCF Connect for example).
It is not flawless – some data is missing from it (Rodalies, Cercanías in Spain, Italo in Italy, some private operators in Romania etc.) and sometimes data from some countries is a little out of date – this is because most data in the far flung corners of Europe comes from UIC MERITS and some railways are slow at updating their data. But generally speaking, for most cases it is good enough. Supplementing MERITS data with GTFS data from National Access Points – that is behind the MOTIS project.
Booking a ticket – where it all breaks down currently
While I might be able to find a route with bahn.de, it is far from certain I will be able to book a ticket there – there are many parts of Europe where I simply see no price listed at all. By contrast third party platforms like Trainline and Omio can sell me tickets for more operators, but often with sub optimal routes.
Take the route Bordeaux – Bilbao for example. bahn.de routes this correctly – showing I have to take the Euskotren to cross the border. But no one can sell an online ticket for that few kilometres. So the correct result here is for a site that tells me “here’s the route Bordeaux – Hendaye – Irun – Bilbao, we can sell you Bordeaux – Hendaye and Irun – Bilbao and buy Hendaye – Irun at a ticket machine on the spot for a few Euro at a ticket machine“.
Or the route München – Wien – Budapest. This is a simple one to plan as it is a direct train across three countries, but the price varies wildly depending which state incumbent operator’s site you use – DB, ÖBB and MÁV have different pricing systems. So I need a platform that says “here is the connection, and we’ll sell you the MÁV ticket because that is the cheapest“.
Getting to these sorts of results is more complex than it might seem, and it needs an enlightened approach from a ticket sales platform – akin to what the air portal Skyscanner did in its early days. It could tell you when and where Ryanair flights would go, but not show prices for them – because Ryanair would not pass that data to third parties. But by offering the customer a fair and complete experience, so Skyscanner built trust. I would prefer a platform that tells me “take the Euskotren even though we can’t sell it” than “here’s a ticket for a less good route we can sell” or “no connection“. Another inspiration here is MoneySavingExpert in the UK – where the determination is to be trustworthy first and foremost, and then work out where this can be monetised – their approach is explained here.
This approach is going to be curtailed by look-to-book ratios, and the commission for ticket resales that railway companies will offer to third parties. SNCF for example refuses to show Trenitalia tickets on its sales channels, but imposes a strict ratio of API calls to sales on its third party ticket resellers – this essentially means SNCF is refusing to offer price comparison itself, and mitigating against anyone else building such comparison tools. And in some markets the commission a platform receives for re-selling a train ticket cannot cover that platform’s costs – although we might eventually get new EU law to fix this issue.
And this has led to platforms having to impose service charges to cover their costs. And that brings us back to the core problem – if using a third party platform is going to cost a customer more, they will use the railway company’s own site again next time. And if – in response to the question at the start of this post – a platform cannot offer me the best price, then I am not going to recommend it to others. Perhaps this charge could be more like a tip, more of a nudge – where by default the service charge was on, was explained transparently, but a customer could choose to not pay it? Or maybe a state owned operator expands its offer beyond its national boundaries?
Last but not least I think there would be some inspiration from the John Lewis slogan “Never knowingly undersold” – someone, somewhere, will have a way to get a train trip cheaper than any tech tool can show. Offering some sort of reward for anyone who can beat the price – if they explain how they did it (if it is legal) – could create a feedback loop to improve the model.
Live running data, with a feedback loop into route planning
A bunch of route planning apps integrate live running data – allowing me to see whether my train is on time or not. And there are even some fledgling European collaborations to share this data. But we are not yet at the stage where any ticket re-seller can offer a passenger live running information for their whole trip across multiple countries.
In an ideal world that data would then be gathered, and fed back into the route planning process. This is the sort of thing that Theo Döllmann is trying to do with Bahnvorhersage.de – to use machine learning and historic delay information to better assess whether a connection will be reached or not. A warning “our data shows your connection is unlikely to be reached – we suggest you depart 10 minutes earlier” would be an amazingly helpful service.
Honesty in the case of disruption and missed connections
At the moment if my journey comprises trains from multiple different operators a passenger can be left stranded and/or out of pocket – because the current legal framework is incomplete, and informal agreements between operators weak (explained here). A good booking platform would take a two pronged approach to this: seek to avoid the disruption in the first place (see the point above), clearly and honestly explain passengers’ rights to them (this route has passenger rights, this one does not etc.), and last but not least draw on Virtual Interlining from the aviation industry – essentially an insurance product that picks up the tab in the case of a missed connection. And this is necessary for now, pending these problems being solved – hopefully – in the forthcoming EU Regulation on ticketing.
So this is what I reckon we need. A platform (or more than one, doesn’t matter) that can manage to do all of that. The best routes, the best prices, reliable live running data and clarity about passenger rights. And as tickets could be sold on such a platform, commission there provides a possible funding stream (although enough to cover all the costs of the rest of this? I don’t know – business development is not my thing).
But how do we get there?
Who is going to get us there?
Comments – either here or on social media – very welcome, and I will update this post with links to sources and other services when I can. Consider this very much a work in progress!
* – other readers might be devotees of ÖBB Scotty, SBB’s planner, or ČD or idos. That’s not really the issue here – there are some tools that do the planning job well enough, feel free to keep on using the one(s) you are happy with.
