Why Tech Startups Keep Failing at Rail Innovation

Picture the scene. A startup gets ten minutes in front of a train operating company at an industry event. They've rehearsed the pitch, the slides are sharp. They spend the entire time selling something the operator has no authority to buy.

The pitch is often strong. The technology is genuinely good. What the startup hasn't done is the basic work to understand who actually maintains the trains.

In many cases, that function sits with the ROSCO or with a separate maintenance contractor under a long-form agreement. The operator in the room may be interested, but they can't sign for it. You watch the polite nods, the 'we'll take this away', and then the conversation quietly stops.

The startup leaves thinking the meeting went well. The operator leaves wondering why these sessions keep sending them pitches they have no mandate to action.

The timeline problem nobody mentions

The deeper version of this problem shows up in conversations I have most weeks. Someone contacts me because they've built a product and want to bring it to rail. Within a few minutes it becomes clear the solution would touch the safety case.

That changes everything about the route to market.

It means a longer assurance process through Network Rail, RSSB-led work, and the operator's own safety management system. If you've raised investment on an 18-month return horizon and your product affects the safety case, you’re going to need a different commercial plan. It’s no surprise then that transport and utility startups face a 55% failure rate within the first four years. When rail decision cycles are measured in years and startup runways are measured in months, the mismatch can be too much for some companies.

What politeness actually means

In the pitching scenario above, the founder mistakes politeness for progress. Rail people are genuinely curious. When a startup walks in with a new idea, the instinct is to engage, ask questions, and explore how it might fit. That curiosity is real but it isn't a buying signal.

'Really interesting, do send me more' usually means 'I enjoyed thinking about that for ten minutes and I'm happy to keep talking.' A founder who reads it as a soft yes has misunderstood the room.

Rail is a genuinely nice industry. Most people in it are courteous, considerate, and reluctant to be blunt. In some commercial industries a founder would get a 'no' (‘I’m out’, ‘not for me’) quickly and unambiguously. In rail, the same answer might come wrapped in three follow-up emails and a suggestion to reconnect after the next industry event.

Founders need to read meetings on substance, not on warmth. Budget, timing, named projects, buying authority, a specific follow-up ask: those are the signals that the conversation has somewhere to go.

What rail gets wrong about startups

Let’s not put all the onus on the newbies: the biggest misunderstanding runs in the other direction. Rail executives assume that startups are naturally trying to enter the rail supply chain. They forget that for most founders, rail isn't on the radar at all.

The sector doesn't appear in the standard list of target markets that an accelerator or investor would push a young company towards. When a startup does turn up at a rail event, it's usually because someone made a deliberate decision to explore the industry. That choice can just as easily be unmade if the engagement gets too hard.

Rail isn't competing with other rail suppliers for a startup's attention. It's competing with every other sector that moves faster.

A startup might be five people, sometimes fewer. Every meeting they take, every document they write, every site visit they attend is a direct cost against a finite pot of cash. When a large rail organisation asks for a detailed proposal, a security questionnaire, a bespoke demo, and three rounds of stakeholder workshops before any commitment, that's weeks of work with no revenue attached. You have to be very, very keen to supply to the rail sector!

When translation actually works

For a startup, a pitching session is genuinely attractive. It's a discrete task with relatively low resource demand. There's no lengthy form filling, no prequalification questionnaire that filters them out before they've had a chance to speak. These formats can be very powerful, if both sides are suitably prepared and expectations are managed by someone who has an idea of where both sides are coming from.

The Heathrow Express Supplier Challenge is a good example of what happens when someone does the translation work on both sides.

What the operator side needed was to define their challenge statements in terms that a non-rail organisation could understand: Inside the industry it's easy to write a brief that assumes shared context: ownership boundaries, asset categories, established acronyms. A startup reading that type of brief would either misinterpret it or self-select out so translation is needed up front. Equally important was setting the commercial expectation in advance: Heathrow Express’ open access status is a key fact that would reframe the opportunity for a startup. There is a direct connection between product benefit and commercial gain, supported by the ability to buy - unlike most public or contracted operators who run services under detailed agreements with the government, with specified outputs, fixed budgets, and limited freedom to adopt new products. For a startup, this is an attractive legible route to market.

The vendor theatre problem

The Heathrow Express example shows what happens when the engagement is real. The opposite is more common.

Vendor theatre is what happens when the format of innovation gets reproduced without the substance. The room looks right, the pitches happen, the photos get taken, and then nothing moves. Everyone leaves having done their bit, and the system carries on exactly as it was.

A common version is the innovation showcase organised to satisfy an internal requirement. Someone has been told to demonstrate engagement with SMEs. A session gets booked, startups get invited, presentations are heard. There's no budget attached, no buying authority in the room, and no plan for what happens next. The startups leave with a positive meeting in their CRM and no route to revenue. The organisation files the attendance sheet and the box gets ticked.

Another version sits inside formal procurement. A challenge brief gets written, but it's been shaped so narrowly that only the incumbent can credibly answer it. The technical specification mirrors the existing supplier's product. The exercise looks competitive on paper. In practice, the outcome was determined before the brief went out.

What's missing in both cases is the institutional process that turns engagement into adoption: a named decision-maker, a budget line, a procurement route the startup can actually use, and a follow-through commitment with a date attached. When those things aren't in place, the session is a performance. It might serve a purpose internally. It shouldn't be confused with the real thing.

Do the people running these sessions know what's happening? It varies. Some understand the limits of what the session can produce, but acknowledging that explicitly would undermine the exercise, so it stays unsaid. Others have organised the session in good faith, turned up wanting to engage, but haven't mapped out what would need to happen for any of it to convert into a real opportunity. They assume that if a pitch is strong enough, the system will find a way to act on it. It won't, and they're often surprised when nothing materialises.

Why neutral platforms matter

This is part of why neutral platforms matter. When a rail operator runs an innovation session, the startups in the room know they're being assessed as potential suppliers. When a tech company hosts an event, the operators know they're being marketed to. Both settings are useful, but they shape behaviour. People put on their commercial face and avoid surfacing problems they don't want a competitor or a buyer to see.

A neutral platform removes that pressure. At the Rail Innovation Group, we don't have a product to sell, a contract to win, or an organisational position to defend. An operator can describe a problem candidly without it being read as a request for proposal. A startup can ask basic questions about how the industry works without worrying that their ignorance will count against them in a future tender.

If we were a rail organisation, startups would treat us as a buyer, even if we said otherwise. If we were a tech company, operators would treat us as a vendor. Sitting in the middle, with no commercial dependency on either side, is what allows people to relax into the conversation. It also lets us be honest with both sides in ways that would be awkward from inside either camp. I can tell a startup their timeline is unrealistic without it sounding like a deflection. I can tell an operator their procurement process is excluding the companies they say they want to attract without it sounding like a complaint from a rival.

What would actually fix this

Strip everything else away and you're left with an incentive problem. The parts of the industry with the budget and authority to adopt new technology are rewarded for stability, not for change. Until that shifts, the disconnect will keep reproducing itself regardless of how many innovation programmes get launched.

The current system pays operators to run trains under contract. It pays infrastructure managers to maintain the network to defined standards. None of those payment structures reward someone for taking a risk on an unproven product from a small company. The personal incentives run the other way. If you adopt a new technology and it goes well, the credit is diffuse. If it goes badly, the accountability lands on you. A rational person, sitting in that role, doesn't take the bet.

Innovation gets pushed into adjacent activities that don't disturb the core. Pilots, showcases, accelerator programmes, innovation departments. All of those exist, and some do good work. None of them touch the operational and procurement decisions that actually shape what gets deployed at scale. That's why the industry can simultaneously run a busy innovation calendar and adopt very little new technology in any given year.

The fix would be to align the incentives properly. Make the adoption of validated new technology a measurable expectation in operator contracts and supply chain frameworks. Build the accountability for not innovating with the same weight that currently sits on the accountability for failed experiments. That's a political question rather than a technical one, and with Great British Railways on the horizon, the temptation will be to build another set of innovation mechanisms on top of a system whose underlying incentives haven't changed. That would be a missed opportunity.

Why I keep showing up

Well, the system isn’t immovable, it’s just highly sticky - new companies do squeak through to transform to the way rail does things. 

A founder who turned up to one of our events five years ago, knowing nothing about rail, is now running a company with operational deployments. A rail executive who used to sit through pitches with their arms folded now hosts their own innovation sessions and asks better questions than I do.

I see it in the conversations. Ten years ago, the word "startup" wasn't really part of mainstream rail discourse. Now it is, and the people using it have a clearer sense of what it means. Once a sector has language for an idea, it can plan around it.

I see it in the policy. The fact that conversations about Great British Railways now include questions about how to engage SMEs is itself a marker of how far things have moved. The answers aren't yet good enough, but the question is being asked, and that question wasn't on the table when I started this work.

What keeps me showing up is partly that I find the work interesting, and the people in it are good company. Rail attracts a particular kind of person: thoughtful, patient, often funny in a dry way, and quietly committed to making things better for passengers they'll never meet.

The deeper reason is that the work matters. Rail is a sustainable mode of transport in a world that needs more of those. The industry is going to have to adapt, and it can't do that alone. The companies with the skills it needs are mostly outside the sector right now, and someone has to do the work of bringing the two sides into the same room.

The system is resistant. It isn't closed. The gap between those two words is where I've spent the last ten years, and it's where I expect to spend the next ten.