View transcript : Trainline Fireside Chat with Barclays

Company

Trainline

Date

30 January 2026

Time

14:00 GMT

Corporate Participants

• Pete Wood – Chief Financial Officer, Trainline

Host

• Andrew Ross – EMEA Internet Equity Research, Barclays

[START OF TRANSCRIPT]

Andrew Ross

Good afternoon in Europe and good morning those of you based in the US. It's Andrew here from the Barclays team. And as part of our series of fireside chats with the EMEA internet C-suite, I'm delighted to be joined this afternoon by Pete, who is the CFO at Trainline. We've got an hour. It'll be me asking the questions just to make sure that we keep it flowing, but clearly we want to make sure that if any of you guys on the line have a burning question, that we can make sure it's asked and I'll feed it in.

So hopefully you can see a button that says "ask a question" on your screen and there you can type it in and it will appear in front of me. Or if you don't want to put it on there, then send me an email at [email protected] or also send me a Bloomberg.

So let's get going. I guess the first thing to say, Pete, is thank you very much for taking the time.

And I wanted to ask the first question, I guess inevitably about GBR Online Retail and was hoping maybe you could recap the process to get that up and running, the timeline and realistically when we as consumers may be able to actually start to use that product.

Pete Wood

Yeah. Well, first of all, thanks for hosting me today, Andrew, and good morning or good afternoon to everyone, depending on where you are.

So we have heard a bit more about the process of standing up the GBR app and website. We're really in a, what's called a market engagement stage. It's pre-procurement, and this is an opportunity for the GBR team to explore with a wide range of potential suppliers, what they might offer based on some loose information, if you like. And it helps them develop what may then turn into a procurement. I mean, that's our expectation, but it hasn't been announced. And we might expect if they continue down this path to get started in the next couple of months.

Thereafter, and the timeline is somewhat speculative, but let me share at least how we are thinking about it internally. We think a procurement of this sort of scale would likely last maybe nine, maybe 12 months or so. It won't be super fast. And then post-award, this gets you to

January 2027, there could be a period of build for, let's say, another 12 months or so. Our expectation is that this won't be a fully off-the-shelf solution that is stood up in short order.

There is a huge amount to be done by GBR in consolidating the different experiences that exist out there today and deciding which feature sets they want to fold in and which they will sunset or replicate. And then from a scale perspective, these dozen or so apps and websites today account for about 15% of the market, but no one of them is particularly big, say 2% or so. So, there's a job to be done on the foundational layers and managing and preparing for something that could scale up as well. That's our best estimate.

The government has also announced that over time they would then expect to shut down some of the existing or all of the existing websites and apps, which again, they said it would happen over a 12-month period or so. Altogether, you've got a three-year arc all in on that if you put that together.

Andrew Ross

So it sounds like it may be well into calendar 2028 until we're actually able to understand how the product looks and have a play with it as consumers.

Pete Wood

As I said, that's our best estimate. There clearly are ways you could stand something up more quickly, but I painted one scenario that we consider internally.

Andrew Ross

That's helpful. And then I guess one of the logical follow-ups is whether you, Trainline, could participate in the tender to power GBR Online Retail, and appreciate there's some nuance in that question, but what do you see as the puts and takes as to whether that might be logical for you to do it and how sizable you think that contract might be in the context of your preexisting white label business?

Pete Wood

Look, we'll certainly be interested and take a look with a view to participating is the kind of posture we have as a starting point. Difficult to say more without knowing exactly what they're after, but we have a capability to develop solutions apart from our mint green B2C solution. We have our white label businesses today. We've got scale in our platform already and the security that comes with what we've developed already, and a fairly rich feature set as well. So I think we have a good starting point, and then we'll have to look to see what actually lies within the procurement, what are their expectations and their asks and how that might fit with us.

And in terms of scale of contract, again, I think it's a watching brief here. It's too early to say, and you really have to understand the scope of what they're after to determine what sort of size contract this might be.

Andrew Ross

Okay, that's helpful. And then I'm sure you spend quite a bit of time thinking about the outer year forecast in your internal budget. What are the kind of scenarios that you're thinking about in terms of how things might play out when GBR retail is up and running? And I guess there's an initial question around type of market share that it might take and what you do to your own proposition in advance of that. And there's a second order question around costs and what you may do on that side, but whatever puts and takes as you guys think about that within the finance function?

Pete Wood

Yes, there are clearly a series of assumptions you make, but again, to paint a picture of how it may unfold, I think the day that it gets launched, you actually just have another competitor in the market at that point. They won't be sunsetting anything on that day, most likely. And so there's probably a period of running where there will be some interest and people might start to try it out and start to build a feel for what it looks like. As that unfolds and as they work through that, what they said will be a 12-month period to shut things down.

On the website, you get a redirect most likely from some of these existing brands that are out there, but the app landscape is more challenging. You kind of reach a point where the app that you've been using for a while just stops working. Then as a customer, you have to go to the App

Store and download something new. Of course, GBR will be pushing hard and marketing and pushing a message out that seeks to capture those customers at that point, but you can expect us to do the same.

Then the intervening period between now and then is really an opportunity for us to further develop our solutions. As you saw pre-Christmas, we launched the way to train in a series of features that are to help you travel. Buying the ticket is part of the experience, but actually supporting customers through their expectations and disruptions that might unfold along the journey, as well as post-sales is something that we think we can use our data sets and our engineers to build and gap the experience versus where we expect GBR to start. So we'll also be in the market trying to win those customers and our expectation is that we will win some customers at that point of switching.

Andrew Ross

So your intention is actually to be on the front foot and maybe spend a little bit more in anticipation about to try and win users, and you can see that as a potential opportunity to maybe increase momentum in your business initially?

Pete Wood

Exactly that. And as I say, I think we've got a couple of years to continue to work towards that, but yeah, that's our posture.

Andrew Ross

Okay, cool. And on this topic, there's one that's coming in from an investor here around whether there's anything around the way that your site monetises or your app monetises or is laid out, but you may change in advance of the launch of GBR Online Retail. And I guess the angle there would be around booking fees or advertising, as an extension of that, maybe just remind us as to how big each of those two are as part of your monetisation today and how you think about those over time.

Pete Wood

Yeah, certainly. We have been increasingly active in that space over the last few years. If you go back three or four years, almost all of our engineering effort was focused on acquiring customers and building scale, focused on selling more tickets. And then over the last few years, hotels and insurance were the two most natural verticals for us to begin to experiment.

We've not disclosed the quantum externally, but in aggregate, if you think of a more or less 8% take rate for our UK consumer business pre-transaction fee, there's four and a half percent today, which is commission, and the balance is a combination of fees, of other verticals and of advertising as well.

So that's a journey we’ve started and we are leaning in a bit more in that space, both to further optimise those two verticals, but also to see what else might fit. Typically we'll try to get early signals and if we get good signal back, we'll continue down that path. But if it doesn't look like it's going to work, then we'll pause and pivot into something else. That's again, how we operate.

Then on the fee front, you may have noticed that in November, we started to experiment with a

SplitSave fee. This was to test and see how the customer reacts to having the fee and the very tangible value they're getting from a SplitSave very closely aligned, and that's in place of the booking fee.

So we'll continue to experiment with monetisation in various guises along all of these fronts, and you might expect the mix to change where we see their wins.

Andrew Ross

That's interesting on the SplitSave fee, I assume you haven't seen any big delta in conversion since you made that change. Consumers are fairly relaxed on that?

Pete Wood

Yeah, the experimentation continues. It's a fairly new space to us, and I'd say that the early results are positive, but we'll continue to test and think about the consumer proposition more fully before we make any particular change over time.

Andrew Ross

I guess one area that I think will matter where there's a lot of devil in the detail is on the specific details regarding to the level playing field. What is the latest in terms of progress of your discussions that I'm sure are happening behind the scene with policymakers around this code of practice that was referred to at the end of last year and how you ensure that this level playing field is enforced?

Pete Wood

Yes. As a reminder, the announcement of the code of practice came with a couple of bits of information. And this against the backdrop from a policy perspective of an open and fair marketplace, and one where the GBR recognised that the private sector does innovate in this space and that there is value to customers there. That's the starting point from a policy perspective. They've separated within GBR the retail function from the administrative licensing part for third-parties. So that is a good step forwards and some clarification that we won't compete directly with the team that also grants us our license. So again, that's a helpful clarification. And finally, that the ORR will be the independent regulator, which is not something we have today. If we have grievances, then we don't really have a route to resolve them other than to make the argument behind closed doors.

So that's a good step forwards. Of course, there are more questions, but until the code of practice has got momentum and things start to get crystallised within that, I don't have more to share on that at this point. And just as a reminder, our expectation is this runs for a year as our best estimate.

Andrew Ross

Okay, understood. So how does the worst-case outcome look then if the code of practice doesn't give us clarity and there are still certain features that maybe you are unable to access, but GBR Online Retail can, or there's still some dispute around certain aspects of product.

What's the worst-case scenario there as you see it?

Pete Wood

Yeah, I don't think that the government has an intent to skew the pitch where they don't allow us to sell certain products. That's not the signals that we're receiving. And of course, we want the certainty of having that codified in the code of practice. I think the worst case for us lies around ambiguity, and that's something we all feel today, and the thought that that could persist in some way is probably the worst case for us, and for investors, for us making capital allocation decisions, frankly, for the team that works here as well. So, I think that's where our concern of worst case lies. But look, as I've said, we've got the right signals coming in at the moment

And from the government, as I just outlined. There are also other bodies such as the CMA who have made the position clear that as far as they're concerned, they expect a certain set of regulations to allow the private sector to compete with the government and that continues to be our expectation that they will get put in place.

Andrew Ross

That's helpful. Let's move on then from GBR and I guess talk about everyone's favourite buzzword at the moment of AI. It's clearly a big topic. We can come at this from lots of angles, but maybe from your perspective, just remind us about how you as an organisation are thinking about AI and how it fits in your business model today. And then maybe I can test you out some of the more potentially disruptive features that investors are talking about.

Pete Wood

Yeah, certainly. There are three strands to how we think about it. The first being productisation, how can we leverage this technology for the benefit of our customers and some of the features that we launched in November with the Way to Train contain that. Then, if you think about how in the future we might manage disruption, AI is a much more experiential toolkit for us to develop something for customers than traditional forms of engineering. So that's one. The second would be the wider ecosystem and how that might unfold and where we might have opportunities or risks in that space. And then finally, the productivity aspect, which is running at different speeds for us, we've had certain wins in some places like customer support. And there's an ongoing journey from a technology and a coding perspective that honestly, is changing week by week.

It's a high paced environment, as we all know.

Andrew Ross

So maybe then let's just touch on some of the consumer side and what features have actually changed on the Trainline app as a consumer that are enabled by AI that are up in the market.

The travel assistant is an example of that, but you guys feel like you are on a front foot about developing things that are genuinely useful to me powered by AI. So it'd be helpful to recap that in more detail so that it's clear and also give us some line of sight maybe as to the types of things that might come over time.

Pete Wood

Yeah. And the AI assistant is exactly one of those that we have had for six months or more or more at this point out in the wild and really is a tool that allows you as a customer to converse with Trainline. And it's a combination of wider world experience, but also very specific to your ticket, for example. You can ask questions about the terms and conditions, you can ask about when your train will arrive, and it answers back in a text form. It can also actually do things for you. If you're able to have a refund, then you can ask it to process a refund and it does it for you.

You don't have to then go elsewhere in the app or to bring up custom support. So that's a good step forwards for us. I think more broadly, things like prediction, machine learning, and AI are very close cousins.

And travel forecast is something we launched in November to give a more nuanced view to customers about how late they might be. There's a huge amount of data that we hold that we can better leverage about maybe it's how widespread disruption is on the network or whether there's a particular problem down the line that you haven't yet experienced. And that can be in the first order, just useful information to bring to the traveller’s attention. Then to speculate where this might go, you could imagine that you get to a point of actually supporting customers with their options. "Hey, maybe you should consider switching at Rugby. It looks like there's a delay down the track." And you might be able to route round it a different way or, "Look, it's just a widespread problem at the moment, best go to the pub and come back at an hour. Adjust slightly." I think that is a very interesting space that we support customers on very real problems that they face from time to time.

Andrew Ross

In terms of where it could go, do you envisage a world where we move away from, I guess the standard set of filters we've all been used to where we put in saying we're going from London to

Manchester at 12 o'clock with our railcard and we want first class or standard class and then it comes up with a bunch of options and you click through, into something that is more like a conversational search booking funnel or down the road with my own personal assistant device that books a train for me. Is that a credible path?

Pete Wood

Yes, I think there is definitely a set of use cases that could deserve different treatment depending on what your needs are. We tend to think in bookends, if you like, there's the commuter that is doing the same journey that they did yesterday and essentially knows almost everything about their journey already and just wants to get to the ticket very quickly. Then you've got the tourist who maybe has never been to Europe, maybe never been on a train and doesn't really know what their itinerary is.

I think particularly for that upper funnel part of the experience, that kind of tourist, clearly we're seeing people use these tools to develop and explore their thinking and to build their understanding before they then go and specifically do something. And that's where we see generative engine traffic growing the fastest is this inbound traffic that is switching the quickest.

It's still a percent or two, so it's a pretty modest point for now. Nonetheless, that's where we're seeing it shift. I think that text based may make more sense for that end, but much less likely for the commuter end.

Andrew Ross

Okay. I guess one of the things then the market is debating is what the, I guess simplistically what the future of the marketplace is. And in the old world, we've had this intermediary layer that's connected a fragmented supply side and put it in one place for the consumer. And now we've got these new powerful LLMs that we're all engaging with. So do you think there's any risk of disintermediation to your business model at any point where, I don't know whether it's going to be ChatGPT or Gemini or whatever personal agent I have down the road, can find some way of accessing train content and selling it to me directly without that kind of intermediary layer?

Just remind us as to how that all works, how hard that would be for them and why your model is protected there.

Pete Wood

Yeah, I think the hypotheses that we see developing momentum are that this will become another source of qualified traffic. But of course, if you go a long way out, then no one really knows how this will eventually develop. I think for the near and medium term, the announcement that OpenAI partnering with Booking.com to bring some of that into their flow, followed by the announcement more recently that they want to test and learn their way into advertising, I think point to at least this being the next stage of development. And look, Google has demonstrated that there's huge value to be created if you can get the right model and attention to, on the supply side, monetise that traffic. And from our side, we'd be willing to pay the right amount to get that traffic and to find new customers and serve them. Actually it could be helpful that the duopoly of Google and Meta properties today that serves a lot of our paid traffic could have a competitor in that space.

And look, we are good at converting our traffic, we're good at SEO and GEO, which further allows us to be able to bid a bit higher in Google, and perhaps if that's the model they take into OpenAI into that space too, because the efficiency of your conversion is something that can wash through the whole ecosystem. So that's how we're thinking about it and exploring it. Then from a disintermediation and disruption perspective, there are a series of other things that we have to do today that would make it difficult for others to disrupt as a single agent doing everything. We have a license to sell tickets. We have to have bonding in place to manage the risk on the cash that we hold for carriers. We're accredited. There are certain ways we have to formulate the ticket. And we're not a meta search player, remember. We actually do deliver a ticket. We help people on their journey. We have a customer support system. So, I think there's an awful lot to do for what is a relatively modest commission at the end of the day, and that fundamentally becomes also part of the moat. You do all of that work, but to what end at the end of it all? So that's how we see it, and that's our hypothesis of how it plays out.

Andrew Ross

Yeah, I agree. I think that makes a lot of sense. So I guess for logical follow-up from that is you kind of touched on some of the early vertical players who have launched apps into the ChatGPT ecosystem in the US. I guess no one quite knows when that's coming to Europe. I assume it will do at some point. So what's your perspective as to whether Trainline should be operating an app within a ChatGPT environment or any other horizontal LLM environment over time?

Pete Wood

I think if you play this out, you're better to be at the front of the queue and willing to test and learn and experiment. That's our approach to it, it's not really knowing all the details right now, it's to partake if we get the opportunity. And we've got the right attributes. We are Pan-European, we’re not just one country. We're a tech player, which should make it easier for us to integrate and iterate and work with another tech player. So yes, if this opportunity or if and when this opportunity emerges, we would look to work with OpenAI or even Gemini or whoever as that unfolds.

Andrew Ross

Okay. And then one thing that's come on here from an investor is to ask you to recap where you guys are at in terms of, I guess, customer acquisition costs. So let's say the Google funnel in general. And I understand that's a bit different in the UK where you have very high app usage than maybe in the international segment where you're maybe more reliant on Google as a source of traffic. And there's actually quite a lot of things going on in this question around some of the changes Google's made with prioritizing its own train product, and now we've got the change to generative answers and some de-prioritisation around traditional paid and natural search. So just recap to us what all of this means for your traffic acquisition costs and how the strategy evolves as part of that.

Pete Wood

Yeah, you're absolutely right. In the UK, our app traffic is north of 90% and it's kind of that repeat use without needing to pay for it. Of course, there is PPC, but we have high brand awareness which supports that. And that kind of model is to mop up new customers that come into the market. Within the international business, it is more important to us. We don't have that same base at this point in time. The mix of free and existing customers is less pronounced typically and domestically. If you take Italy as an example, it's more like 70% app versus the 90% we see in the UK. And then you are, again, further right, this search results page has evolved a huge amount over the last couple of years. I won't recap everything you just said, I think you narrated it nicely for me.

The punchline of all of that is that the SEO is reduced, that free traffic that we were very good at doing and getting is just not coming through in the same way because consumers never quite get to the link and they put this extra slot at the top to monetise the fact that all the players are suffering a bit and are therefore competing harder at the top. We've had this double whammy of less free traffic and more expensive paid traffic. Now, we said in November, and it persists today, that we're seeing a lapping out of that. So it's not like the thing has gone back to what it was, but it's not a growth headwind at this point. And not to say that the Google results page won't further evolve. This battle is far from over with ChatGPT and AI's place in search will continue to evolve,

I'm sure. But right now, we're seeing that lap out and that's allowing us to reappraise where we stand.

We said in Spain in particular, we were going to seek to rebalance that profitability and growth paradigm after three or four years of going hard to build our awareness, which is now in the 30s.

That was a kind of get the thing off the ground moment where we expedited some investments, and now we're fettering that back. Of course, you'll need to wait until the results of this half to really see how that's played out. But yeah, that kind of hypothesis is something that we are continuing to experiment with in H2, and you'll see the results of that in due course.

Andrew Ross

That's helpful. So one more on AI before we move on, which is about your own cost base and productivity, which I'm sure is something that you as a CFO spend quite a lot of time thinking about. So I mean, curious to any examples you can give us as to where it's driving real efficiency within the organisation today. And actually there seems to be some debate out there amongst investors in marketplaces in general. What is the genuine incremental cost to build AI enabled features? And some dispute, is this expensive, is it not? So from your perspective, has that been a big drag on the cost as you start to build some features running on top of LLMs?

Pete Wood

Yeah, I'll start with a non-tech example of a tangible win within customer services where the AI- fuelled capabilities have just meant that we've been able to serve, particularly European customers more efficiently, and we have made savings in that space which are clear to drop through. So that's one clear win. We're also experimenting in the marketing space and that's ongoing. But look, tech is our biggest cost and is probably the most dynamic of all of this. And I think the way you posed the question, it's both true that there are savings, but it also has some new costs. And that space will continue to evolve. I think it's well known that running queries on

LLMs is definitely more expensive than traditional forums, but there are also efficiencies in how you code. And that space is moving very quickly where engineers are now able to do some of the documentation through an assistant, and they're getting the efficiency that way. I think that will continue to evolve over time as well.

We're also finding that there are different parts of our tech stack that are more open to it. So our supply connections is one space that we're particularly interested in working through how we can use AI to manage that space. And that's an incubation part for us. And then of course there are the AI products we're developing as well, which again have a OpEx envelope to them aside of the efficiencies. And look, we're learning fast, we're seeing the right signals. At a macro level, we will continue with the team size. We're not about to announce a saving or a more investment in this and within that. We will evolve the skillsets that we have and that we need over time. Of course, at some point there'll be a moment to review where we are and actually can we bank some of those savings? But for the foreseeable, we'll stick with the investment that we have and operate within that envelope.

Andrew Ross

That's helpful. I want to change tack and ask you a couple about the UK Consumer business, and how to think about the medium-term growth, and try and avoid talking about GBR online retail, I guess, and just more think about the structural dynamics of the market. I remember when you guys went public, it was quite easy. The end market grew for 4'ish over time in terms of value, online gained share, and you guys gained share of your online market. And it was a fairly easy top-down model that we could use to drove your growth. And now we've had all of this noise with COVID, with train volume collapsing, and then a big rebuild at different cadence and different bits of the market.

We've had a big step-up in online penetration. I'm not exactly sure where we are now. And then we've had a year this year with some noise around a step-up in ticket prices, having had pretty high inflation ticket prices for a couple of years. So it feels like that top-down model has got more complicated. So how do you think about that top-down framework for how your UK consumer business should grow or how the market should grow?

Pete Wood

Yes, I agree. There was a period of digitisation then disrupted with COVID, and it is a more complex picture now, in part because there are other technologies and moving parts out there, if you like. And we're in a much more mature environment, of course, too, that mass market digitisation is complete. There is still £2 billion of the £11 billion or so out there sold as tickets in the station. So the laggards are still there, and that should provide some further digitisation pressure over the longer run. But these people are hard to pick off. I can't even persuade my parents yet to switch to digital. So some of them were going to take some time. And then there are some emerging changes which complicate things further. So the expansion of the Oval Zone is something we've talked about for a long time. This is around London, and that £150 million of

[net ticket sales at] risk is still the right estimate from our perspective, where some people will choose to switch to tap and go, even where others won't because they like having the screen and the ability to use their railcards.

So that's another factor to consider. And then you're right, I think it was 30 years ago, they last didn't run a regulated fair rise through the system. Of course, that's a headwind to us and I think consensus is in a reasonable place at this point, having digested that announcement for us next year. Of course, you've got the compounded effect of Oval on top. So yes, I think a more mature environment with more muted opportunity to grow certainly next year, but some upward pressure over time from digitisation. And then I know you said we shouldn't touch on it, but the moment of change with GBR is the next point of shift, I think, and timing for that is less clear cut.

But the government has made it clear what their intentions are.

Andrew Ross

Yeah. So I guess as part of that, I guess the government would ideally like more people to use the train, and I assume would also like a digital first means of booking most tickets. So is your view that we actually might see something tangible in terms of accelerating growth in the end online market that results as part of that? And obviously the share is a separate conversation.

Pete Wood

Yes, I think there is opportunity, but as I say, the digital journey is now relatively mature. So to seek to expedite it further, the GBR team will have to consider the social and accessibility aspects to further pushing digitisation. Nonetheless, things like digital pay as you go could offer another route to unlocking that. And I know you're aware that there are trials running in the

Midlands, and that Trainline is part of that trial. We are really pleased with the results that we are getting back. However, as a reminder, this is a test to see not only the technology, but how the proposition could work for customers.

For shorter trips, that might offer another route to digitisation and might move the ball forwards again, but I think this is a few years out. There's still work to be done in how you might enable a further rollout and there are a couple of options that they need to make a decision on.

Andrew Ross

I was actually going to ask you about digital pay to go later on, but since you touched on that, let's knock it in now and then we can come back to one more on UK Consumer. So where are you guys at on the trial there? It's been up and running now for a couple of months, I think. What have you learned about what's working in the technology, what's not? What's the feedback been? I guess a bit I'm not really clear on is what happens next, in terms of when there's any decision made as to whether this is a nationwide rollout, a regional rollout, who gets what contracts? I'm not quite clear on that.

Pete Wood

Yeah, the feedback's been good, and I've tried it myself up in Leicester and it's super simple. You open the app and push travel, and then you get on a train and that's it. It's seamless in that regard. It's also offered us the opportunity to fine tune a few bits. If you get enough scale going through, you see the edge cases that need some adjustment. So we're pleased with that. I don't have certainty on a next step. There is no further announcement that you are missing. We hope the trial might get extended and maybe expanded, that would allow us to continue to evolve and iterate. That's what we are advocating for, but there are decision makers that need to get their head around what the best next move is and we haven't yet heard yet.

Andrew Ross

And there's actually a follow-up question that's coming on this. So I guess there are also other providers who are operating trials at the same time. Is there any way externally you can compare whose system is or isn't working, or any KPIs you can point to that may help investors to formulate a view as to whether your solution is one that maybe is likely to go on and win bigger contracts down the road? Or is it just too soon to tell?

Pete Wood

It is a little too soon to really tell. The way that we think about it, however, is that there's the fundamental technology, but then there's the ecosystem that sits around it. We have a highly developed team when it comes to the fuller proposition of post-service of live train times and that not everyone has fully developed that at this point, even where their tech stack does the actual permission to travel and some of those aspects. That coupled with our scale gives us a real advantage in some respects. However, as I said, until we have further clarity on how this might get implemented, it's hard to opine exactly on how that advantage might play out.

Andrew Ross

Cool. Helpful. Let me just ask one more on the UK business and we'll go on to international. So on the cost side of things, we've touched on some of this around some of your, I guess, ideology around the marketing with GBR online retail. You've already touched a little bit on AI. But just walk us through the thinking in general on how much OpEx growth is going to be needed in that

UK consumer business for the next couple of years, and your ability to dabble up or down depending on what scenarios do you play through on the top line?

Pete Wood

Yeah. Essentially we have two costs, and some general overheads. One is around marketing, and that is to some extent algorithmic where if you are clearly getting a return, then why wouldn't you keep going? And then there are some bets you take when it comes to brand, and pushing messages, and evolving the way that people think about your proposition. So there's some variability in that, but it's fairly well honed in the UK. And as I've said, in Europe, we're already beginning to adjust our posture in some ways, and we can talk more about that in a minute.

Then from a technology perspective, again, as you heard me say, we have a team size that we are comfortable with. We still have a roadmap that's oversubscribed, but in a healthy way, I think some tension there when we're making choices is always good. We want to continue to invest in the experience, both in terms of how we might evolve our monetisation suite, but also to support travellers as well, and that there is this window before GBR gets out there that we can make the most of. So yeah, we have optionality for sure on the amount of investment we put behind technology and updating our product suite, but right now we feel like it's the right investment for us.

Andrew Ross

There's still a couple on international then. So back in November, you guys had some really helpful slides on laying path, I guess, for how route digitalisation comes through in France, Italy, and Germany, having, I guess, happened to a pretty good degree in Spain. So just remind us, I guess, of the key timelines, and has anything changed incrementally since then that we should be aware of?

Pete Wood

Yeah, I think broadly the opportunity is still the right one as we articulated in the past. There have been some near-term changes in the last year, which I think most people are aware of, but in particular, in France seeing Trenitalia launch between Paris and Marseille, and run some incremental services from Lyon to Paris as well. So that's been good. We've seen a good tick up in our growth rate for those routes, which plays through into France as well as support growth there. So that's been a good step forward.

Other news, there's a company called Proxima that has now announced their brand. They are going to be Velvet, and that continues to be the next launch in France. Albeit it's still a few years away, but they are continuing. And now that they've got a brand out there, you're seeing a little bit more impressed with them.

And then finally, SNCF, are meant to launch in Italy, that has slipped back a bit. I think that's now in [FY] 2028, we're expecting that, but still expecting to launch. Italy for us, we've already got two players. That third one will be really interesting to inject some dynamism into the markets.

Again, we'll want to lean into that moment when it comes, but at the moment it's a watching brief.

Andrew Ross

Okay, got it. And that actually feeds me into a question that's coming from an investor here around how to think about your marketing costs in general in international and when you dial it up, when you dial it down and I guess what you've learned from Spain as part of doing that. But it feels like an area of uncertainty in the modelling over the next couple of years potentially.

Pete Wood

Yeah and it's the timing that's the harder bit to predict. We will get good visibility as we get nearer to a moment, but when you are building trains and getting them out of the factory, getting them through the safety certificates, that's where some of the delays can start to creep in. And so that's the ambiguity for us, is could that slip? And then sometimes it comes forwards like

Trenitalia's launch from Marseille, and other times, it can slip out.

In terms of how we have evolved our thinking and marketing, coming out of COVID, we wanted to be there as people started traveling, and we made the most of that moment, both in the PPC bidding that we do with Google, but also in the brand as well. And then, we found that it was hard to justify the brand in France where there wasn't liberalisation, so we pulled back from that. We've not found it possible to get expedited growth with a better proposition alone. It's that catalyst of liberalisation and playing the role as the aggregator is the unlock moment where you can put more brand marketing in, and that's really the playbook that we've used in Spain. We've now moved that awareness to a good level, and as you heard me say, and as we announced in

November, we're now rebalancing that. However, that doesn't mean we won't do any, it just means that we're looking to rebalance the investment to drive growth versus the profitability.

So yeah, I think if you think about SNCF launching in Italy or Velvet launching on Paris to

Bordeaux, which is the second-biggest line [in France], those are moments that we will put more brand [marketing] in. As you've heard us say before, there are seven operators that might launch in France over the next three or four years, and at some point, enough of the map is filled out that the national campaign might then be possible again, and that's the exciting market for us.

We've learned a lot in Spain, we've made good progress getting into the low-teens when it comes to share of high speed, but Spain ultimately is a more modest market, whereas France is the same size as the UK, and that's really what represents the longer term opportunity for us to really play out and drive that international part of our P&L.

Andrew Ross

That's helpful. So one thing I often debate with investors is how the long-term profitability of international might look. An investor who takes a bullish view as to where the net ticket sales may end up on a mid to long-term view then naturally feeds in, well, how might the margin structure look? And I guess Spain has now reached a level of maturity where it probably does start to make sense to think a bit more about how those margins are scaling and how the economics look. So just remind us now where that Spain contribution margin sits. I know you're not going to give a specific number. But help us with the building blocks as to how this business builds into profitability, and then how we think about operating leverage over time in your more mature markets.

Pete Wood

So the first separation I would make within international is that tourist business, foreign travel, people coming in is, let's say, a fifth or so of our International [net ticket sales] and is already very profitable. You have people who are much more inelastic when it comes to fees, we get better commissions in that space, and typically they're traveling maybe not as a single traveller and they're doing multiple trips in a given month. So that's a really lucrative business that is profitable already, but has a smaller TAM. Think of it as being, I don't know, a five-billion-plus

TAM in that space. So good headroom, but ultimately a bit more limited.

Then you've got the domestic markets, and this is where there is a higher need to be aware of on price. And so, fees is something that we haven't really experimented with at this point. We're more reliant on the commissions and the ancillary revenues that we make. So that's probably one point of difference versus the UK. And then, we're trying to balance the monetisation and the friction that that creates with getting customers into a habit with the scaling and the growth aspects getting to a critical mass in these markets and not getting in the way of liberalisation and those moments that they bring. So we're still evolving our thinking in this space. We will disclose more over time, and to see how we've adjusted the profitability and growth is something that we will share more on over time. Ultimately, it's going to be that monetisation lever, and the tension between that and wanting to scale the growth over the longer run as this liberalisation unfolds over a few years that will allow us ultimately to see what margin is possible in the longer run.

Andrew Ross

Just on the topic of international, I've had for one that's come in, which is to touch on the tragic news around the high-speed crash in Spain in January. Has that had any impact on people taking the train in Spain, to your business? Anything you see around that?

Pete Wood

Yes, tragic indeed, and yes to the impact. It's not just isolated to where they've had to shut the track, there is a concern in the Spanish market around the infrastructure and the safety. I think it's really the authorities and the regulators that need to figure out the issues that have caused these, and to provide a path to reassure the public that any systemic issues are being addressed, or are addressed, and that that is not more widely felt. So yes, we're certainly seeing that impact echo in Spain. Worth adding that we don't see that echo effect in other countries, whether it be France that borders it or other European countries that obviously have heard of the news and digested it, but we aren't detecting any reticence to travel on train in other markets.

Andrew Ross

Let's do one on capital allocation, which is something that I think investors are pretty focused on. So to your credit, you guys have been, I think, quite on the forefront of buying back your own shares, you're doing that pretty quickly. Is there anything else you can do to highlight the value that you believe exists in your stock in advance of GBR online retail being up and running? And I know there's been an investor who's asked a question about why not go harder in terms of the attitude to putting leverage on the balance sheets to buy back shares, whether it might make sense to do that with some kind of tender offer as opposed to doing it in the market, and what are put and takes that sit around that maybe, and leverage in general.

Pete Wood

Yeah. We very deliberately went out in September with a punchy message, it was we were mindful of where we were and we ultimately see through the ambiguities, and we've discussed this today in some depth. But that's ultimately where we see that our proposition, our brand strength and those assets that we continue to build on will see us through, and therefore we feel good about this bolder announcement. Then, there's the question of, how bold do you want to be? And that is, of course, a live discussion. All of these discussions, I should add, are within the context of our capital allocation framework, where we're making sure that organically we're well-fuelled, and that if there are inorganic opportunities, we will explore them.

But this is a moment where we have let leverage rise up over the last six or more months. We haven't put a cap on it, if you'd like, but to talk and give a bit of colour, I think a leverage of 1.5x is something that we would be comfortable with. The program that we've announced already will continue with and, of course, in September we'll share a bit more about where that might go over time. Equally, at the same time, we don't want to put ourselves in a particularly awkward position either. So we're mindful of trying to find the right balance, and that's, as I say, a live discussion.

Andrew Ross

Okay, that's helpful. One on B2B just quickly. I saw a press release recently about you guys expanding your partnership with Amex GBT. It kind of feels, in general, like your flow of contract wins is pretty good there. So just help us update, remind us of the latest updates there.

Pete Wood

Yeah. It's been a period of great momentum in this space. The aggregation or the fragmentation of supply in Europe has actually led to some really great opportunities on the B2B side as travel management companies in particular have got pressure from corporates to offer rail, from both shifting their carbon emissions down, but also being able to report on them, and therefore funnelling them all through one provider. But on the other hand, being faced with a fragmenting supply base. Trainline already has the supply connections, there's sometimes more engineering to be done to enable B2B fares, but nonetheless, it's really allowed us to take a step forward.

I think we're now seeing the work that started a few years ago. We put a bit more PPC in, and then the next day, you see customers turn up. But with these contracts, they take a time to negotiate, there's a technical process for it, and actually wiring all these things through takes a bit of time. But now, we've got that momentum going. Great to be signing up some well-known and big brands and to be partnering with them. The one thing I would have you remember is that, as one example, Amex GBT has been acquisitive, and so there's not one platform that sits behind this. So again, there's a bit of work to enable bits of the supply over time, but nonetheless, we see ongoing opportunity in this space.

Andrew Ross

Cool. Three minutes and two questions. So first one's from an investor, which is to ask about the competitive intensity in the UK at the moment in the context of TrainPal, also interested in anything to call out around Uber or otherwise.

Pete Wood

Yes, we track the landscape, of course, and TrainPal and Uber are two within the basket that we follow. TrainPal have pushed pretty hard, particularly on downloads. That's what we notice and track. They are taking some segment share overall as other third parties, if you like, at a modest rate - it doesn't really translate directly from downloads to transactions, as best we can tell, but we're watching there. And Uber, they dialled back their credits offer to those that aren't Uber

One members, so they are still retailing, but again, we don't see the same vigour that they're going at rail at the moment. So yeah, it's a watching brief for us. And of course, GBR is the other moving part that is moving in the shadows at the moment, but will eventually emerge onto the scene there.

Andrew Ross

Cool. And my final question is, I guess to hand it back to you actually, Pete, to ask you if there's any closing remarks to what ultimately underpins your confidence in Trainline looking forward, and anything else that you think investors should have in mind about the overall investment case that we haven't touched on in the last hour you want to particularly emphasise?

Pete Wood

Thanks, Andrew. Look, I think as you've heard from me today, we are entering this next phase with GBR from a position of strength. We've got a strong brand, we've got a large customer base, and that scale also brings with it data assets, as an example, that we are looking to leverage in innovative ways using AI. In terms of AI, we are mindful of making investments, and obviously alert to how the market is evolving and to make the most of those opportunities that come. And with an at-scale tech base and that mandate to run and build products for customers, we really see some opportunity in all of that. So yes, we're confident. There will be ambiguity. We will, of course, narrate that to you as we follow it along. But we feel good about where we are today.

Andrew Ross

Superb. In which case, all that's left is for me to say a big thank you to you for your time, and to wish you a lovely weekend ahead.

Pete Wood

Thank you, Andrew. Thank you for hosting us, and hope everyone has a good weekend.

[END OF TRANSCRIPT]