Trainline plc (TRN) Earnings Call Transcript & Summary

[Audio Gap] to the Trainline Plc's fireside chat on the topic of it's tech, data, AI and its expertise around that. My name is James Lockyer, and I'm the technology analyst at Peel Hunt. But for the purposes of today, I'm simply a bystander eager to learn more. I'm delighted to be joined by Peter Wood, Chief Financial Officer; Milena Nikolic, Chief Technology Officer; and Mike Hyde, Chief Data Officer. Thanks for joining us today.

Great. Thanks, James. And yes, it's great to be here today and to be able to share a bit more about our tech platform, which, of course, underpins the whole commercial success of our business. and also to be able to share the opportunities we see with the unique data set we have.

That's great. Before we dig into that though as you would have it, you've put out a trading statement this morning, describing the first half financial performance. Perhaps, Pete, you could kick us off with a quick overview of your year so far.

Yes, sure, of course. So yes, we're very pleased with a strong group performance in H1. Our net ticket sales grew 23% for the group. Revenue for the group grew at 19%, a bit behind NTS due to some maintenance FX, but both are ahead of consensus and despite the strikes that we've got in the U.K. And if I just unpack the different segments a little bit more. So U.K., just continued focus on the world cup markets with the quick buy and other features that just make it even easier to buy tickets on the day of travel and that has continued to drive the growth of the top line. E-ticket penetration has moved furthermore as a result, so 43% for the last year, and that's now 46% overall, and Trainline is a key part of driving that forward. International Consumer was at 24%, net ticket sales growth, a little bit behind consensus, which is at 31%. The highlights of the international portfolio continue to be Spain and Italy, so these are our aggregation markets, but we have a really strong proposition. And the liberalization in Spain is really driving and supporting our growth in that market. And in Italy, we continue to see growth in the regional part. So not only winning share where there is a high speed and competition, but also where there isn't as well. And that's really underpinned by the app, which is now 60% of all the transactions that we have in our European business. We've also called out a couple of headwinds on the website. There are some normalization effects year-on-year versus the COVID affected year of '22 and particularly as we came out. So we're seeing demand begin to soften year-on-year as the half has unfolded. And we've also seen the competition for PBC auctions increased a little bit. So this time last year, there wasn't much. The train operating companies have now got a post-covid budget, and we're seeing more activity from them. And there's some evolution in the way in which Google is presenting its search pages as well. So the travel module, which was for flights and hotels now contains trains and they're experimenting with how and where that shows up. And as a result, the SEO prominence that we have has dropped away a bit. So there's a bit of a headwind there. From a Trainline Solutions' perspective, we've also had a good growth of 37% net ticket sales. There is some, obviously, growth from the business market from a lower base, which is supporting that growth. And in IT Solutions as well, I think there's some kind of ongoing recovery from COVID, which the B2C market -- our B2C proposition kind of a quicker to the punch there, but we're now seeing some of the uptake in this part of our business as well. So that's all good. We have reconfirmed guidance for the year despite the strong results. And really, ambiguity is around the ticket office closure, consultation that is ongoing in the U.K. There is a scenario where the government looks to go after this and to pursue shutting some of the opposite, and that might incite more reaction from the unions and there might be more strikes. So that's one scenario that would take us to the bottom of the range. Of course, if it plays out in other ways, then maybe will be of the range as well. And then finally, we have already stated our capital allocation policy and announced a GBP 50 million buyback.

So just on that share buyback. Can you talk us through your rationale and thought process for coming to that decision at this moment in time?

Yes, sure. So the business has come out of COVID very strongly and has continued to trade and grow as you've seen from the H1 results. And that has led to a significant deleverage of the business. And so look, we'll still look to fund our expansion and our growth, in particular, on the organic side, we might consider inorganic, but our assessment was that even after prioritizing that there was some excess capital to restart to shareholders and the flexibility of a buyback program was attractive. And given where the share price is trading today, we think there's an excellent investment opportunity here.

Great. Thanks, Pete. Now Milena, as CTO and Mike as Chief Data Officer, you both bring unique perspective to Trainline. Could you briefly run us through your professional journey, leading up to Trainline, your remit at the company and how you both fit into the organization.

Thank you, James. Yes. So I have started my career journey as a software engineer at Google and have actually done that for quite a few years across number of products like Google Search and Androids and some production infrastructure stuff and so on. At some stage, into my career journey, I moved into leadership roles. And over the next 7 or 8 years ago, well I have led teams and organizations of various sizes, most recently as Senior Director for Google Play developer ecosystem. So having joined Trainline 2 years ago now, just over 2 years. I'm in charge of leading innovation and tech vision for Trainline. This is with a goal of advancing our customer proposition and helping as many customers have most seamless rail journeys, of course, also helping partners leverage our 2021 Platform One technology and all of that in order to put rail into [indiscernible].

So hi, James, Mike Hyde. I've been kind of working in data and building data teams for -- I think over 20 years now. I started off doing that in the guys of sort of business consulting and strategy consulting companies in the early 2000s when the kind of big data revolution first happened and companies really started to acquire large quantities of data and then were sort of asking themselves what they could do with it. And it's been a super interesting journey in a sense. I got into technology after that and ended up working at Skype in London. So I built the data team, data science, data engineering team for Skype consumer, lived through the mobile revolution as part of that as Skype tries to transition from desktop to mobile. That team acquired by Microsoft. So I ended up building data science teams inside Microsoft Office and Skype eventually became Skype for business and what is now teams essentially. So still powers a very large proportion of the world's calling and communication. And as I went to Facebook. I was hired by Facebook to be -- to lead the -- all of the data science, data engineering teams for the London site. So went through 5 years of really big growth there. We went from a very small footprint in London to thousands and thousands of technical people by the end of it. I was also one of the founding members of the workplace products, which is Facebook's first attempt at a B2B market, which was super interesting. So I've been kind of building and growing products with data for a while now. I joined Trainline 2 years ago, but just very shortly after Milena in fact, really with the belief that there's an amazing opportunity to leverage data in rail and for Trainline to really see that as a strategic asset to grow the company. And also my extension to help really modernize the experience of traveling by rail, it causes the whole industry to deal with a little bit of that modernization as well. So -- so inside of Trainline, I run data science teams, data engineering, machine learning, AI, kind of like the full stack of the data teams that you'd expect to work with, be it with Milena and the rest of the exact teams on steering and growing the company.

Excellent. So across to both of you, lots of different verticals, business life cycles growing different parts of them, bringing a lot of experience to Trainline. So Milena, to frame the discussion, can you give us an overview of your tech stack given the core features and capabilities and where the consumer brand trainline.com fits into at all?

Right. So you simply put our [indiscernible] customer touch points and then a layer of back-end. So felt bit about both. On the front-end side, we have our set of mobile apps, which are very highly rated 4.9 stars across android and iOS. So we have web front dense, which are really highly ranked in [indiscernible] and search engines. And then we have APIs on the B2B side, which we integrate with our partners. So I'm very happy, which is the Polish level of quality and it's kind of the fronts that our customers and partners like very much. But the part that I'm particularly proud about is, I think our layer of back end that kind of comprise our sort of Platform One. And the reason for that is that, that's where we sold the hardest problems. And rail is complex. And I'm sure some of it is obvious, some of it is probably not quite as obvious. You're not going to -- maybe we get on talk a little bit more about it here, but it's -- there are 1,000x more airports in Europe than there are -- sorry, 1,000x more train stations in Europe than there are airports. There are interesting traffic patterns. There is -- there are no GDSs. So aggregators like ourselves or anyone looking to have that proposition that's kind of across carriers has to individually integrate with multiple train operating company APIs, which adds a lot of complexity and then kind of there is a huge difference in the domain across their shares, ticket types, seat maps and all kind of the other optionals and so on. So sort of our that kind of set of back-end layers or Platform One covers the entire customer journey from these early stages of greening up the next exciting trip to Thailand or buy the ticket for your next daily commute kind of into London through kind of the transactional journey taking in case anything changes kind of being able to change our journey as well as on the day when you're traveling letting you know about the platform, any instructions. If there are instructions kind of what will be the next best action and so on. I think the part -- so there are quite a lot of different components. We won't have time to talk about all of them. But if I spend kind of our capability is going to be that central part of the global transactional platform. And I think 3 key aspects of it would be searching journey planning capability. So again, we have that across hundreds of different carriers, which will be dozens of different technical integrations, and we have optimized that to be both fast. So if users ticket from London to Nottingham, we get to you within matter of seconds or very often under 1 second. But also those journey planning results are highest possible for it. So you can get the cheapest ticket. You can get the one with the smallest number of transitions, you can get all the times that you wanted and so on. So search and journey planning is kind of the first part of the global transactional platform. Second part is our checkout and booking flows which have been optimized for scale and currently serve more than 1,000 transactions per minute at peak times. And this is something that's really interesting where trains are different to the aeroplane if one of those things that might not be obvious is for most of the rest of the travel industry, they can build things knowing that people tend to buy tickets kind of many months in advance, whereas we have to deal with this morning commute peak or evening commute peak where a lot of users are coming to expect to be able to buy a ticket at that moment. So that kind of peakiness in those traffic patterns and being able to have the scale and robustness technical systems to serve that in a very infrastructural ways for -- particularly in the U.K., where we -- where we are kind of the largest seller, but kind of also growing important European market. So that would be the second sort of part of our global transaction platform. And then the third bid that I want to call out is the fulfillment part. So in addition to showing you the journey and then being able to purchase tickets, we let you get the actual ticket, which is in the largest percent of cases or bar code or an e-ticket in a matter of seconds. And again, it is important, because of that more in peak scenario whereas you will buy your plane ticket weeks, sometimes months in advance. You book your hotel typically much in advance. Our consumers are walking into a train station and as they're approaching the barrier gate, they're buying the ticket and they expect that bar code to be landing in a matter of seconds. So again, the sort of the latency that we kind of have to provide, yes, delays that the kind of the reliability of our platform just has to be really solid because of that. And then kind of, again, on top of that the real-time [indiscernible] to get into a lot of details. I'll let Mike talk about our data assets, the smart data features and maybe the last thing I will just very briefly call out is our customer center and customer service technology, which is kind of also quite involved and very important for the quality of care we provide to customers end to end.

Yes. Lots of complexity, and I'm sure we'll get to the data point later. But with that complexity, I imagine you've spent hundreds of millions of pounds getting to where you are today on that. And that will obviously be a barrier to entry for anyone looking to come in, but perhaps you could expand more around the moats. And there are large players out there that have bigger pockets, but is there something specific that -- about your space that you have that moat? How do your tech stack say against other rail carriers and also other retailers in different verticals? How should we think about that?

Right. Yes. The way I think about it, I'm actually quite keying for both Mike and Pete to kind of pitch on this side. I think there are 2 -- actually 3 parts or all the way I think at it. The first one is sheer scale. Pete will tell you that rail is a thin margin business. So it requires quite a lot of scale, quite a lot of transactions and kind of that's your volume in order for the economics to stack up for us to be able to invest in the kind of powerful and modern platform and consumer experiences. For us to be able to -- for Trainline to be able to hire people from Google and Facebook and kind of the best marketers and technologists that can be had in the role. So I think that sheer scale that enables economics to sort of work. The second part of the moat, and I slightly alluded to this, I think in the previous question, is the technical integrations with carriers. And there is both the breadth and depth point to it. The breadth will be the number of carriers [indiscernible]. We like to say that we were in it all the carriers and whereas you might be able to pick up some very niche one that we might not, but we have almost all the carriers, all ticket types and kind of all the fares and kind of all that. The reason both breadth and depth are important is in order to see in a local market like -- European markets like, like France and Spain and Italy, you need to be able to serve customers with everything they would expect for rail. And usually, that's not just ticket, that will be all the fares on the tickets, kind of all the railcars, kind of all different things. The complexity it was not obvious to me before I joined was that child fare in France could be different one in Italy, the age range and then the level of discount you get and we get child for free. The same applies for senior railcar, the flotation tickets can get within seat maps, obviously, for different carriers and models of trains. And now on top of that, the fact that each one of those kind of hundreds of carriers has a completely different API that we have to integrate one by one, again, going back to that point of no GDS, no other aggregators apart from what really we are doing now in [indiscernible].

My favorite complexity example. So when I joined Trainline, same, very similar story, really. Rail is hard, is what I very quickly learned, the next hard in ways it's not immediately obvious. So one of the things I discovered to my surprise early on is that if you retail rail tickets, you have to worry about something called a look-to-book ratio, which means you can't -- there's a limit to how much you can even check the price of the inventory that you're trying to sell. And that's mostly because of the historical technical systems on the side of the carriers, right? So it means as a retailer, you're fundamentally serving more requests for searches and prices than you are allowed to check with your suppliers, right? So that represents a fundamentally hard technical product. How do you do that accurately fast, real time and sort of get that modern experience on top of quite an old technical architecture that you're backing off to invest in the industry. And so I mean, there's examples like that all over the place, right? So it's just 1 thing where that middle layer of Trainline just kind of solves that kind of problem makes it go away in sort of slightly magical ways. But it's hard than I've seen it, right?

Yes. And it's just probably comparing. It took Trainline hundreds of engineering years to build out all of these, so sort of expecting new companies that come in and rebuild all of this. There is obviously the investment point, but there is the expertise point as well of like understanding it all and figuring it out and that's where...

And the margin point, it's more right . You do it at all and the margins are fixed, you need to go with it for a long time to actually get the return. And that's a key part of our rails.

Yes. And just I'll add one more. So just kind of third I'd really like to add in. I kind of touched on this a little bit as well, but because we cover the entire customer life cycle in terms of traveling by rail. And this means that we're not a search aggregate or like journey search type technology provider, we are there with the customer irregardless of whether they are on that big special trip or on the daily commute, we're there. Even if they need to change their journey, we're there, is -- kind of as they're traveling or kind of dealing with the chaos of train stations that kind of all that. So again, it's a very broad technical and product for more than anything offering that's kind of, yes, regardless of whether -- whoever trying to get, there's quite a lot of service to cover and build out. And I think this comprehensive customer care is both a backup in terms of loyalty, but equally more just be a pure build-out sense kind of the cost of replicating something like that.

So one of escape most people have noticed is that Uber has been upping its game within the transport and rail industry. And obviously, they are a tech platform, they do invest in things, and they're actually going hard quite on price promotions at the moment. But obviously, as Pete pointed out, that's probably not a sustainable business model given how thin the margins are. They do have big pockets. They can fund loss -- being loss making for longer than others. How do you -- does this -- is this -- how do you think about this? Is it something to be worried about?

Yes. I think we certainly pay a lot of attention to the competition, and we keep an eye on what competitors are doing and we certainly noticed Uber starting to play in the rail space and that they have a large enough customer base and big pockets of business we're paying attention to. All that said, based on what we've seen, it looks like Uber for a rail in an experimental test effectively. We currently do not supply, technology, does not provide technology to Uber. And we say that's appropriate given that the platform -- the size of the platform player that we are. We're not really interested in doing a lot of customized, high change work for someone who's working on testing out an experimenting the hypothesis in a certain space. I think small partners that they have gone for are actually very appropriate for kind of that they're running. However, if their tests were to be positive, if they get a good signal that this is kind of a business they can make, we think we're actually well positioned to be a technology partner with them and other entire platform leverage point and so on. And maybe just to be super transparent based on the market data we are seeing in the U.K., we do not believe they're getting any traction right now. This kind of industry level data, which is public shows that there are no non-Trainline, online players that are getting share, whereas Trainline continues to grow ahead of the fund.

Yes. Just to pick up on the kind of investment point, you're picking up James that the 10% Uber credit is cash outflow that you redeemed it right and that compares with a commission in the U.K. of about 5%. So 10%, 5%, this is definitely a loss leader -- and yes, as Milena says we haven't got the evidence that they're making material progress. Obviously, we continue to watch. They are well funded. We have seen other big players come and go in the past. The way Skyscanner and Expedia have experiments with rail in the U.K. And at certain points, they decided that it was not worth the pixels on the page that could be otherwise dedicated to driving other revenue streams. So -- yes, we're still watching, but we've not seen a huge evidence of great traction yet.

Excellent. Just before we talk about strategy, Mike, perhaps you could add a bit more sort of numerical context the complexity of this industry that Milena has flagged a number of times. The conclusion might be that there's billions of pieces of data going through your platform on a daily basis. Can you just give us an idea of scale and the growth of that data asset and how -- and the uniqueness of it in this market?

Yes, yes, certainly. I mean just to put some rough numbers around. We process about 6 terabytes of data every day, so it's constantly flowing into our platform. And the way to think about it, the uniqueness of our platform is that we sit kind of at the intersection of supply and demand. So on the one side, we ingest a lot of supply data. So we said we ingest a whole lot of detailed information from 200 different rail operators all across Europe. So that's real-time timetables, ticket availability, pricing, ticket rules, real-time information about when trains are on time or delayed and so on. So there's one half of our data light comes from there. The other half of our data light is more on the demand side, which is then everything about the users. So on that side, we're ingesting not just when people buy tickets, but all of the other user behaviors that enrich that. So we're getting search data. So we're really starting to be able to understand demand at what people are looking for. What their desires and needs are, whether or not they are actually able to build that by buying a rail ticket, it's quite interesting. And then we're starting to be able to put that together and look at people's behaviors at the time. We can -- we get scan data from barriers so we can see people are actually using those tickets when they're taking journeys. And then, of course, there's a whole lot of data then around what happens, as Milena pointed out, as we are getting further down the journey. So helping people to manage those journeys, make changes, make cancellations, make refunds, all these sorts of things. So the uniqueness, I think that we bring to this is this combination of these 2 things, because a lot of players in the rail industry have a lot of information on the first half, but not very much on the second part. And when you start to think about how users are interacting with the rail system, if I back to me is a really unique valuable part that we're finding -- returning a lot of value after that.

I'm sure. Yes. I think there's an interesting sort of commercial angle that we'll touch on later because there definitely seems to be value in there that's building. So Milena, in an industry that's as complex as yours, there must be lots of different priorities, wish less ideas running through the business. How do you think about the group's overarching tech strategy so you don't lose sight of your end goal?

Right. So the Trainline differentiates itself with technology. I think that's important to say and it's pretty clear and I reviewed example, 4.9 star rated apps and the trustpilot [indiscernible] things we're very proud of and the things that are essential to our business and technology that kind of effectively drives that. We continue to innovate and invest in our kind of product and technology surface that's absolutely key to kind of how we move. And then also tech team is kind of driving priorities of the business as a whole. So I'm kind of supporting purchase of the business whole. So as a business, as a leadership team, if we decide going to pay a goal, for example, the tech team will go and build that. So there is both innovation points and obviously kind of wherever we point our ship as a business, the tech team goes and builds technology that we need to do that. The other thing I would say kind of on top of that is sort of -- we have a large platform that serves quite a lot of users, and it kind of makes money and all that. We have to invest in constant ever greening of that technology service and the evolution and all that -- we do that in a way that's aligned with kind of key KPIs like availability. We go after anything that could potentially affect scaling of our tech platform or reduce robustness or reliability or tech productivity of our engineers. So -- and just to give a couple of examples, we might decide to move from -- or we are actually in the process of moving formulation of database, validated storage to kind of further improve our kind of robustness and then we're moving from [indiscernible] to the containers to a better -- improve productivity of our engineers and so on and -- as well as working on the cost efficiencies within what is a large technical platform and sizable cloud bill and all that. So the debate that often plays out the exact, and I think it's sort of really healthy and robust discussion is how much of our engineering time should go to business goals and kind of driving growth versus sustaining and scaling and growing and modernizing and keeping our platform as up to date and powerful as possible. And I think we are really happy with where we end up with those sessions. I think it's kind of very healthy. So kind of ultimately that strategy ends up being about that as a being about how we innovate, how we drive business goals and how we keep our platform as modern as possible.

You made it very clear that technology is the heart of the business. But in terms of what's the weight within the business in terms of technology people in terms of the size of that infrastructure. Are you a tech business? Or actually, do you just -- how should we think about it from that perspective?

So in terms of -- if you're asking organizationally in terms of size, the tech team is half of the business, very sure that, that would put us in kind of a tech business and kind of the field that Pete talks kind of regularly and as it's healthy, kind of I want to make sure we have the right team and kind of obviously, the appropriate sizing and kind of all those things. So we have about 500 engineers, just over half the company, 80% of those are software engineers, so people developing our product offering and our platform offering, and then the rest are security, privacy, reliability operations and so on. Among the 400 that are software engineers that are actively advancing our product and platform offering, about 100 work on this spec strategy part that I spoke just earlier. And then the rest are advancing the chasing business outcomes across our U.K. business, our European business and Trainline Partner Solutions.

So again, there's lots of priorities, lots of different moving pieces within the business. There is the consumer, there is B2B, there is white labels, Global API, and that is international. How do you ensure every team across the business is aligned and fighting the same fight.

It's a very good question, and so important. I take it really personally to run an efficient organization, but also to run an organization that's focused on the top priorities for the business and kind of making sure we get the most out of the deal that way, and I think it's also most respectful of the people we have on the team and their hard work to put them on the things that are most important. As a comment to summarize it in 1 sentence, we make sure we put people on very clear strategic opportunities, and then we just make sure they're kind of set up for success. So there are a lot of good standards, things we would do like strategy planning and [indiscernible] objective secure results and goaling our team in that way. And that applies both teams to take care of technology and technology strategy. They have clear metrics related to reliability and cost efficiency and productivity and so on as well as our either market-specific targets or product proposition capability-related target. So it's kind of -- I deeply believe that this clear alignment of people to strategic goals, very often to individual metrics that they're kind of expect to drive makes team more motivated, more empowered and the kind of that really helps. And then the last thing I'll sort say on that topic is, we look at quite a lot of our initiatives and generally kind of our investment in the people within tech through return on investment plans, kind of what we expect to get if we have a [indiscernible] team in this particular area. And based on my experience, it's not something most tech businesses do. It's kind of like so many of them play around with technology and sometimes something comes out of it and sometimes things don't. But again, I think it's just really respectful of the time and effort people put into it and we're hearing such a good feedback from the team that they enjoy this transparency. And naturally, because of that feels gravitate towards areas that are highest impact and again, that's how we end up with the teams that are more motivated, more resilient and more ready to be agile and change when something like that is needed.