S1E7 How AI will change travel booking with Trainline's Gary Stevens - podcast episode cover

S1E7 How AI will change travel booking with Trainline's Gary Stevens

Oct 14, 202430 minSeason 1Ep. 7
--:--
--:--
Listen in podcast apps:
Metacast
Spotify
Youtube
RSS

Episode description

Director of Engineering at Trainline, Gary Stevens, joins the show to talk about the role of machine learning in the travel booking experience, the importance of engagement as an engineering leader, and mentoring vs coaching.

00:00 Gary's journey to director of engineering and learning from great managers

09:00 Coaching vs. mentoring

13:00 The importance of engagement in engineering and the importance of celebrating good work

21:00 Machine learning and the future of travel user experiences

Gary's recommendation: The Hard Fork podcast

Transcript

Hello everyone, welcome to PriorityZero Elite Dev Engineering Leadership Podcast. I'm your host Scott Carey and today I'm joined by Gary Stevens, Gary. Thanks so much for coming on. Before we get started, tell us a little bit about who you are and who you work for. So I'm Director of Engineering at Trainline, which is Europe's most downloaded travel app. And what do you do there as Director of Engineering?

So we'll look after two of what we call our strategic themes, customer lifetime value, which is about getting customers onto our app and onto our business and then engaging them as well as our global customer journey, which is about building out that transformative product experience. Right, I'm really excited to kind of dig into how you do that. Before we do, I want to hear a little bit more about how you've got to that point in your career. So when did you become interested

in software engineering as a career? So during university, I did a very broad technology degree. There was bits of, you could be science, some bits of animation, there were bits of networking. And I wasn't really sure what it was wanted to do, hence the sort of the broader limit. As I kind of went through and learned a little bit more, initially it was drawn to some of the more creative elements, but I found the creative elements in programming and

software engineering, the bit that really kind of hooked me. After that, I surely started, I volunteered a charity helping them embrace coming online and following that, I had a few jobs at different agencies, again, sort of following that more creative element. And from there, my career took off. So when you were more of an individual contributor, were you focused more on the front-end side? It sounds like you were on that side. Yeah, I was very much on the front-end side.

Yeah, and at a time when I think the front-end was going through a real great period, when I started HTML and CSS were the main things that I did. And Flash was still very much on the landscape. And through my career, obviously the decline in Flash Store gave rise to that, to a lot of foreign frameworks, and had a lot of good opportunity to be in a place where we were trying to replicate a lot of what we'd historically done in a really interactive medium, natively on the

web. And as someone that isn't so much hands-on anymore, I feel like front-enders just change so dramatically. How would you feel about being a front-end developer today? Do you think it would be so different to what it was like when you got into it? Yeah, I think so. I think there's a lot more of things that people looked at as traditional software engineering. There's a full application lifecycle look at. There's different browsers, there's different devices. There's a plethora of

frameworks. There's a plethora of paradigms and way of doing things, whether it's assuming all of the content from the back-end or whether it's doing something truly interactive. I feel like the area is just kind of alone enormously. Yeah, and it feels like even for people who are hands-on with all of these new frameworks, all of these new paradigms, it's a lot to consume. As someone maybe who's at a higher level of remove now, how do you stay up to date with what's

relevant? Because I hear that from a lot of engineering leaders. It can be really challenging to stay up to date, especially in an area that you kind of know, but that's changed a lot since you were there. Yeah, I try to give myself a bit of a break and not try to know everything. There's obviously areas where I'm more drawn to that I'm more curious about how this often particularly problems today. We do a lot of showcases internally where we present a lot of the work

we've done. That's a really good opportunity for the curiosity part to take over. Ask a couple of questions of how they solve particular problems. A really just kind of like skip level meetings, listening to the team, maybe you're attending a few of their talks and stand-ups. Can I give you that insight of the things that they're trying to do? Yeah, so yeah, for you fairly it's much more valuable to kind of hear what people

are doing inside your four walls. You're less focused on what's going on elsewhere. Yeah, I want to know where they've got problems and where I've got opportunity to either provide support resources, lean in with advice or connect them to the people that can help solve their problems.

You know, there's still a lot of the timeout side of work where I'll maybe tinker on the number of unfinished side projects that litter my desktop, but usually kind of the areas I tend for in the office are to try and understand where can I help these people be more effective. Yeah, I was doing some leadership training recently and I became,

someone became made me become aware of the idea of the advice monster. I don't know if you've come across this concept, but I think for someone at your level it's such an interesting thing to understand whether it's something you're kind of consciously aware of and if you have any ways of kind of taming that advice monster and for anyone that doesn't know it's very much what

it says on the tinnitus. Just that voice in the back of your head that isn't helping someone solve problems themselves, it's that temptation to be like this is how I would solve it. Yeah, I'm very consciously trying to think about my impact. My impact in leadership is a big word, but I'm aware that to a junior engineer or a middle-level engineer advice I give may carry weight,

positively or negatively in how they receive it. So if I'm approached for the problem, I kind of go into that active listening mode, try to understand what the challenge is, try to understand it in a little bit more detail, and then I maybe try to play back to them. If I've understood you, I think the problem you're suffering from is this. One of the greatest advice I was given by a coach a few years ago was to then immediately start with and what do you need from me?

And those words I think have been really powerful in in immediately making the conversation one of like are you looking for advice or are you looking for some support and to offload? And if people are kind of then open the door and say need some advice and what to do next, that's my gateway

into kind of offering that without without style with what he's what you should do. It's actually really good to take that into your personal life as well, I found, because like it's even more tempting because the stakes sometimes feel a bit lower to kind of skip that step in personal interactions. So I become much more aware of that. So that's been your path so far. So you're now director at Train 9. At what point in your career did you side you wanted to move off of that IC

track and onto the management track? I started leading a team probably 2013 small small team of a few engineers and at that point I sort of shifted in how I operated. It was I saw it much more about enabling them to do good work and helping them solve problems that were coming rather than

solving problems they had and giving them space to do that. Through a number of projects I had to do more of that and I found myself slowly kind of getting away from the work but some of that consciously I got as much enjoyment out of supporting engineers and helping them develop to solve problems as I did doing the work myself. Now I enjoy the outcome. I don't need to be right there

in doing the hands-on work. I think that's a massive realization. I think the moment you realize that is the moment you're hooked and the moment you realize that it's probably all is going to be that track for you now and I feel like people who kind of go the other way that's the moment where they realize they find that too difficult or they don't get that same level of gratification. I don't think it's so much about the challenge of the job of management. I think it's that

gratification piece. I think you can learn the other staffers. I think there's almost you've got a really interesting developing or leading people. That's what as you say that's what hooks you in. That's the bit that kind of gets you into the role and that's where you can get some real satisfaction on a daily basis. The rest of it is all about the experiences. It all about the challenges you face and the scenarios you're dropped into and over time building the toolkit of

things of how you would approach a problem. Yeah. Once you realize that that was the kind of path you wanted to go down. It seems like most people get shunted into that and then you figure out as you go along, you don't get time to step away. A football manager does and they take a year between jobs. Where did you turn to when you realised that that was where you were going next in terms of figuring out some of those more like people challenges or some of those management

things that you hadn't learned in university or on the job? I had a really great manager. Previously to Price Trainliner, I was at Compair the Market, the Price Website. I had a really great manager there, Gorkwool Nick, who was himself, had been a coach and so he brought a lot of his coaching experience into how he managed. I had transparency and openness with him and I could

say to him, look, I'm not really sure how to do this bit. I feel a bit out of my depth. He was great at telling me, playing back real life examples of how you'd done something and say, what you're forgetting is the time you did this or the safety net you've got here or what isn't you're trying to go for. Through those interactions, it just over time helped build

that confidence. He was there to kind of catch you. Even though he would consciously, he wasn't right, you felt that that strong line manager was there to support you. I think that and a few techniques and strategies that he gave me for managing uncertainty were really, really powerful. Yeah, I had a great manager once who I'm really bad at numbers and Excel. As I moved up the management track, that became something that I had to figure out

because there's just so many spreadsheets in your life. I'm still not great at it, but I remember I was sat there looking at PNL sheet and I was like, I don't know what to do here. When I told him, he was like, okay, do you want to get good at Excel or do you want to get good at doing a PNL? I was like, doing a PNL and he was like, great, I'm going to give you the basics of this and you

can try and figure out. Then if you get stuck, you can come to me. That was so useful. I've actually said that to loads of people when they come to me with a slightly more technical problem. It's like, do you want to get really good at this technical thing or do you just want to solve the problem? Yeah. And then we can have that conversation. I think that's a really great bit of advice because we can so often be blinded by the challenge of knowing what will the outcome

to be an unfamiliar situation. And it can be tempting to try and solve both at the same time. And sometimes you do need to think, this will give me a use case of how I want to solve that technical problem in the future. But now I'm focused on the outcome. Yeah. And so you had that great manager. Is there anything you took away from them now that you manage teams and now that you're trying to coach people? Because I think we bandied the word coaching around a lot. But having

recently kind of like tried to get better at it, it's really, really challenging. Yeah. Is there anything you learned from that good manager or any other good managers that you now are really conscious about when you run your teams? I'm much more conscious. I completely agree. We say coaching and mentoring interchangeably, right? We use it in the same context. And one thing train lines have been really good is kind of investing in people in some of those coaching sessions.

And patience is the huge virtue in coaching. You're not trying to give the answer. You're trying to lead someone to the answer. And I try to be very deliberate. So I do a lot of mentoring internally and I try to move out of the coaching in the mentoring sessions where I ask, what's the problem you're having and what have you tried and how do I help point you in the right direction?

So it's more just being really deliberate about when those two skills are used and recognizing I think the situation I'm in and knowing, am I using the right tool for the job ahead of me? And how long have you been a director now? So is the director at the fair market for two years and a bit of train line two years? And so what was the biggest kind of shift for you going from whatever level you were before directed to director? Definitely being comfortable getting out of the

detail. A piece of advice I was giving early on was put down the pickaxe, right? You don't, the depth of information you have in an area about what's happening in the team doesn't mean that you're leading it in the right way. It just means you know what's happening. Learning to lead through others and learning to give them the space to try things and fail themselves. Those were very uncomfortable things to do straight away because you feel if you're allowing

the team to fail, that somehow you're letting them down. But in reality, you're allowing them to learn the value of experience and giving them the autonomy to do what they need to do rather than depower them and step in and take over. And it can be difficult, right? If you don't immediately know the answer to a question that you're asked because you're not in that detail, it's okay to sort of say to somebody, I'll check in with the team and I'll follow up on that. I think that's

been something I tried to do more and more. And so four years at the director level, or I feel like that's typically the time when people start to get a bit itchy, like by finding the next step after directors are really interesting one. Where are you thinking going next? I try to remain really open-minded about what I want my career to be. I've found that in the past when I've been

closed to focusing on a particular area, it may blind the direction I take. If I only focus on what I identify as the next role, I don't want to take a route that doesn't actually give me the fulfillment of what I like to do. So right now what I'm trying to do is think about what I know my obvious development areas are, where I need to guess some outside experience, an outside perspective, some feedback, some look back at what I think I'm doing, well, what I think I need

to do better and try and build the plan around that. And where am I the minute? I think I've got ample opportunity to grow in that and take on new challenges and develop a part of things I've not really been exposed to before. Perfect. Right. We're going to say quick break there, and then when we come back, we're going to talk about Gary's Priority Zero. So Gary, this podcast school Priority Zero, we talk about the biggest priority on your list right now. What's your biggest priority at

train line right now? My biggest priority is probably the word engagement. And I'm using engagement as a very broad kind of catcher for a lot of things that I'm trying to do. For me, it's about making sure that the engineers that work in me, the part of the business I'm responsible for, feel that they know what their work is contributing to. And with that, it kind of opens up a whole

new set of challenges. So providing visibility and clarity of goals, ensuring that they have attachment and see the value in those goals, that they feel they're able to contribute to them effectively and their work matters. And for me, that's the little bit of the recipe for success. The more we kind of do around giving people that sense of connection and engagement to their work, the more we can do celebration moments about things to go well, the more we can provide feedback

or development areas and things that they're struggling with. So really trying to build that strong connection and problem solving around what good goals they're like and what good focus looks like in those goals. That's where I'm spending a lot of my time. Yeah, and you're a consumer facing

app with a huge amount of users. And I feel like you could use engagement in this sense with a dual meaning because I feel like you're talking about the engagement that your engineers have with their work, but I feel like that's probably driven by the engagement levels that you're seeing from your users as well. Is that the kind of feedback that you're trying to build? Absolutely. So, you know, we're a consumer facing business. The engagement and feedback of our customers

in form, we're really customer-centric. And ultimately, we're here to build a transformative experience for customers, giving them the tools in our app or in our website to plan journeys that matter, to make the right savings, to choose the right thing that matters to them. Yeah, the engagement that we get from them is just as important. And we want that to be the catalyst for people kind of finding the exciting things to go after. Yeah, and we've talked a little bit about

impact. And I'm always interested by how people start to measure impact. Like with a train line particular, you said your biggest priority is engagement. How do you start to measure that impact and then create those celebration moments for your team? I think if everybody is clear on what the team goal is, and if they can answer themselves, do we feel that this goal has impact either for our customers, for our consumers, for the partners that we work with? Then that helps

them answer that question. The next layer down, I think, is for the engineers themselves, is the thing I'm contributing to this substantial and meaningful? Is my individual role in this adding and creating impact? Is something we ask through the promotion process that we've rolled out? We ask people to focus on their individual contribution. We don't want to build a team

of solar heroes, where we want teams to be harmonious and contribute together. But when we're looking at promoting individuals, we want to know where was your direct impact in this work? Where did you solve a problem that the team were facing? Where did you uncover a particular gnarly challenge that you solved? I hear from a lot of leaders, it's very difficult, especially in teams that may be a little bit more remote to actually create those celebration

moments. Have you got any ways that you found that kind of deliver on that? Yeah, we talked about it back to earlier when when your words matter. And I think sometimes I can be thinking if I give a team a shout out, does it feel genuine and authentic? Is it just me? Set it over the team. But it does matter. People do know that if somebody external to the team is noticing and appreciating their work, they feel valued and they feel their

work is being noticed. So when I see particular goodness of work, or when I hear about team members or direct reports of my teams that are doing really well, trying to give those call out moments, the big one that we've done this year is a bi-weekly showcase. There's no format. There's there's a drop-in, everybody kind of comes in and then there's just a couple of minutes on either the AB test they're running, the technical challenge they solved, the great bit of research that

are research team have done, or a great bit of insight that our data team has spotted. And that's really organically grown into something that has been a few people find like the highlight of that fortnight where they get to see exposure to things other teams are doing, they get to see cool things are coming and they just get a sense of celebrating each other's work. So it's led by the engineers, it's the whole content of this facilitated individual contributors are presenting

the work, non-managers, and it's become a really organic celebration point. And do you find with that format, like I love that format, but do you find that you kind of get the same contributors over and over again, and the are you conscious that you're trying to kind of use it as a showcase for people who maybe don't feel like they're delivering that impact and don't feel as confident I'm putting their hand up, but you know that they are, they just don't have that confidence.

It's the bit I've been most pleasantly surprised about is that the breadth and variety of presenters has been really broad. I've got to give that the line managers a huge amount of credit in creating that support to allow their engineers to do that. We have teams in London and in Barcelona, and we have presenters from everywhere, right? Like we have some presenters who are around that week, well that day and decide to present their idea remotely. And the engagement has just been

consistent whether someone's in person or not in person or whatever role they've been at. It's been really great rewarding to see. Nice and I'm going to go back to what you were talking about when it's talking about kind of engagement and measuring it. And you mentioned like having clear goals, you've said that a couple of times, really important. Like with how you kind of start to measure that, do you believe in having kind of one goal that you can like shoot towards or is that just an

impossible thing to actually do? So we run on OKRs really heavily and we try to have a nice big succincts here, a metric. Yeah. You know the one thing that the team conceptually can just focus on within that. They can have different chaos that maybe have different types of metrics. As long as they're confident that they can prove those things and they ladder up to their

hero metric in the past of our teams, we've kind of picked one or two different metrics. And while that's great, because they maybe are trying to prove different things, creating that central focus and commitment around moving that one big thing in a meaningful way tends to add a bit of cognitive overload. So it may not always be the right thing, we'll iterate that and we'll learn, it may not always be getting the chaos right, but trying to simplify at the end of this quarter,

we're going to say we move the needle on this. That's the mindset that we try to embed. And are those things like spend a user time on app, things like that? Eating conversion rate, so simplifying the speed that customers get through, the number of searches they have to do. Yeah, we tend to focus more on kind of a haver or things that make the users experience of the app better.

And one thing I always am interested in with talking to people at your level about metrics is this idea of kind of safety latches with that, because I think if you have one measure, it can be very tempting for teams to kind of gun towards it and ignore any peripheral damage that that metric can have. Layers, that's something that you think about. Yeah, we'll talk a lot about

boundary metrics or metrics are leading indicators. We have a really talented data science team that partner really effectively with engineering and product counterparts and they can have that conversation around, is this the right indicator for this behaviour? Do we have the data that led us to form the hypothesis that said that that metric leads to this behaviour? And bringing that confidence in those things and facilitating those conversations and creating that clarity

has been really great at making sure we're picking the things that do matter. And if we don't, then we can kind of do a bit of a retro on that, figure out why, what was the hypothesis that we got wrong or the data point that we didn't notice? You mentioned data science, obviously machine learning, big topic right now. I think with an application like yours where you're trying to make really smart recommendations, you're really trying to help people through that kind of journey.

How much work are you doing? How much time are you spending thinking about the elements of machine learning and the application and then obviously the step change we see with Gen AI and trying to bring that in as well? Yeah, I think it's, we obviously pay a huge my attention to talk about where there's a lot of work we're putting into building out more of our ML capability, looking at where does Gen AI start to cross over with travel planning or information that supports or gives

context around these things. And does it? Because it's recommendation engines, obvious. Gen AI, slightly less obvious. I think if you take someone like Google who are, who are already displaying train times directly in search, hotels, flights directly in search, I think if you imagine a world where multimodal journey planning is where people want to get to, where they can say, I've got a hotel here, I've got a hotel in Edinburgh, what's the quickest way

for me to get there? That's when I think things will start to kind of come in and disrupt and dismdmd and, and provide that overview and how, how we play a valuable role in that by using our content and our experiences where we have a wealth of knowledge in our experts in this area, how we can really contribute to that or create experience that, that matters is, I think, a direction that,

that we may see a merge. And at your level, how challenging is it to kind of bring in those machine landing elements while still thinking about safety, thinking about how you bring those teams together? Like, what are your biggest challenges with that side of things? I think we, customer trust has to remain vital, right? And I think we have to show customers where we got some, how we got to

an answer, how we landed on an insight. And I think customers want to know that, right? They, I think, users across digital landscapes are now, we're not naive anymore, we know a little bit more about how the sausage gets made, we know a little bit more about how our data is used and collected.

And I think we have to reach more transparent with, with providing answers on that. So the customers trust the we're making a, the right decision or the right recommendation, especially in an area where they may be spending 60, 80, 100 pounds on a train ticket and it's a big outlay for someone, knowing that it wasn't just a machine that recommended something arbitrarily that it was the right

call, I think is vital. Yeah, and I think it's interesting when I ask you what your priority zero is, you talked about something that's very team oriented, you know, and I'm interested like, is that how you think about your role now? Director level, is it much more about kind of bringing that culture of engagement up through the layers?

I think so, I want engineers in, in the areas of the business I work in to, to feel pride and attachment to the work that they do and know that what they're adding is, is contributing positively, is developing their career in the right way and that they've got support to try to run around them. Now, I tend to do that through the people that I lead rather than

directly. And so I have to provide them with that context and so it's consciously thinking about what are the, what's the tools and the information I need to give to my team so that they can effectively lead their teams and allow them to manage and work with their teams in different way. But providing that North Star, providing that feedback and providing that direction and context, also about to the business as well as internally to the team is where I see my role.

And what's the biggest block of that priority at the moment? I think time, I think it's really tempting to want, you can't expect overnight change when you're driving cultural change. I would love to do that and there's this power where you, where you see a problem or you, we simply do a lot of engagement surveys, you see a particular negative piece of feedback and you know that solving that problem only comes with proving what it

is you do and letting somebody else feel that that change has been made. I can't just, it's not like code, I can't release something, it's fixed and it's live, it takes time. And I think the breadth of things that you kind of want to move in and engaging in or create time is always the biggest blocker.

And I want to look into the future a little bit here in terms of like when I think that that priority is super interesting because it feels like a very long term priority, like at what point would you be happy that that is no longer priority zero and you can move on to the next one? I would say I'm never going to be done because it's a continual continual investment and continual process. I recently read a quote that

Utopia is described as the process of improvement, not the indestination. And that's what I will see. There'll be how much time is needed and how self-sustaining is will be the true success of whether it's working effectively when it becomes something where I feel that everybody is positively contributing to the things that they want to drive. Then I'll feel that we've created an environment that's self-sustaining but we'll always pay attention to it. And what are you really

excited about in your domain at the moment? I think you know we touched upon the opportunities around data science, around machine learning but in terms of kind of book and travel experiences, what kind of keeps you going there in terms of the next big challenge or the next big opportunity? I think rail is a really interesting point where people are starting to understand is a more

sustainable and more effective transport measure mechanism versus plane. I've had the opportunity in the privilege to travel, we recently traveled from London to Milan and then on to Rome. Yeah, you're trying to do a lot of rail travel, don't you?

Yeah, we try to really dog food and live the brand value. And doing that 17 hour train journey through Europe whilst at a huge investment of time was so great of having breakfast in London and lunch in Paris and then going into snow, and then shorts and t-shirt in Rome was really fun and really great and it has such a massive benefit to the environment to travel that way.

And I think as more people start to realise that that shift is a worthwhile one and time spend on trains is an enjoyable time versus try to an honour plane, they're going to look for ways to take better trips, to find cheaper prices, to take more comfortable journeys. And so so big thing is like how do we create those, how do we capture them momentum and how do we create those experiences that our app works just as well for the people booking, they're on the day

commute as it does people planning that big holiday. Yeah, in theory you can get a lot more word done as well on the train, but yeah, a ton of more word done is great. I remember when I was travelling around Europe as a 20 year old and that was before a lot of these options existed and yeah, trying to navigate the different operators across Europe was a full time job. So being able to take that out of the equation is it really opens it up, it makes Europe more of a

continent less than a bunch of connected countries. Yeah, absolutely and that's a huge, that's a huge concern point for customers and a huge anxiety moment of in Spain and Italy there's complete aggregation across routes, you know there's a lot of lot of companies operating on multiple routes and with different class of trains and different

carriages and different types of journeys. So knowing what's best for either your price point, your time or your comfort is yeah, can be a little bit daunting and that's what we try to make easy. Perfect and I like to finish every show with a recommendation, this can be anything, it can be work related, not work related, something you read, something you've seen, what would you

recommend? I've, I've been listening to the HardFort podcast that there's a weekly snippet and dive into the industry at large and what I really love about that is the nuanced way that they

kind of dive into a particular topic and talk about it and it's real world impact. That's been a really great one at just getting me to think a little bit differently about the problem zone facing and how challenges, how in the industry are large and Gen AI and other things are becoming more prominent, privacy concerns, safety concerns, information consumption has been really great for just kind of keeping me I think abreast and informed. Yeah, those guys are great, they do a really good

job. Gary, thank you so much for coming on today especially on one leg, we didn't even got onto your injury but I really appreciate the time and hopefully see you again soon. Thank you again for listening to Priority Zero Elite Dev podcast. Remember you can get us wherever you get your podcast Apple Spotify but when you do, please remember to like and subscribe so you don't miss an episode and we'll see you at the next one.

This transcript was generated by Metacast using AI and may contain inaccuracies. Learn more about transcripts.
For the best experience, listen in Metacast app for iOS or Android
Open in Metacast