[Music] [Michael Colombo] All right. Let's get started. Good to see everybody at Summit. Hope you guys are having a good first day in Vegas. Good to be back.

Excited to talk to everyone today with Kevin here about the transformation of Brightline trains. We'll do our best not to let the session run off the rails.

Introductions first. So I am Mike Colombo. I'm the Chief Marketing Services Officer for WillowTree. That's a title that no one's ever heard, but basically, what it means is I run the marketing services business for WillowTree. I've been in the industry for 25 years. I started an agency, a digital agency, about 20-ish years ago. We joined the WillowTree family in 2021, which is when I met Kevin at a previous company and now at Brightline. But my team, what my team does is essentially what you'll see in the presentation today. So it really runs end-to-end marketing transformations from strategy and design to development and data and activation. So you'll see all of that in our story today, but that's what my team does and what my background is. I will hand it over to Mr. McAuliffe. [Kevin T. McAuliffe] Yeah. I'm Kevin McAuliffe. I am the Chief Technology and Digital Innovation Officer at Brightline Trains in Florida. I want to bet to have the longest possible title with the least amount of responsibilities possible. Anyway, I've been doing this for almost 30 years. I've worked for Disney, JPMorgan Chase, and a whole slew of other startup companies that you might remember as AOL, CompuServe, ICQ, and a few others. So I've been around the block, and we like to run a team that gets stuff done, and I use other terms to indicate the stuff. And I believe that this will tell the total story of what we were able to accomplish in a very short period of time at Brightline Trains. Thanks, Kevin. Super excited. So how many folks in the room are in the travel industry? Just a show of hands.

Not many. Okay. So a couple of things. I think just setting the stage for how we entered Brightline, what the situation was, I think, a lot of this will be-- Let's see if my clicker works here.

So where we entered post-COVID so right now two big trends are travel returning post-COVID, and then what that means for marketing? So what we see as far as demand in the space goes. So obviously, COVID affected every industry, none more than travel and hospitality. I think, as far as, industries impacted, if you look at all of the subgenres of travel and hospitality, they were impacted most of the entire business world. I think what we see happening, though, since 2022, 2023 is a huge return to travel. They were calling it revenge travel at one point getting back at COVID, getting out of your house, and getting back on the road. This year, we're predicted to see more international tourism than pre-COVID for the first time. So since 2019, 2024 will be the first year that there's more travel since pre-COVID. So certainly back in a big way. Also, spend share of wallet. In the US, the average person took over three trips last summer. That's in a time when inflation is high, a lot of economic uncertainty, share of spend in other categories was going down. At the height of inflation, share of spend for travel was going up. So that's what we saw happening over the last couple of years, the last 18, 24 months in particular. What that meant for marketers in the marketing space is really chasing that demand. So in the travel and hospitality industry is growing the fastest of any industry in terms of digital ad spend. That's for the last three years. That will happen again this year. So more marketing dollars more quickly into this industry than anywhere else. The other really cool thing, really interesting thing about travel and hospitality that we'll get into is the ability to personalize, the ability to think really deeply about your guests is unlike a lot of other industries. Travel is personal. It's memorable. There's a lot that goes into it from a preference standpoint, from what are the parameters, how many children do you have, what your budget is, all of that stuff. It's not like selling dishwashers, right? It's not-- There's only so much personalization we can do in that space. In this space, though, we can really think a lot about what do we know about our guests and how can we help them really choose their own adventure. So I think that's what's really neat about the space. That's where at a macro level, we enter into Brightline Trains. Kevin, if you want to talk about Brightline's position in those within those trends. Yeah, I think what we saw pre-COVID was one thing, and then we shut our business down. So COVID really turned us off completely. We threw everything out, and we had to start over again. But what we're seeing is, especially in our short haul from Miami to West Palm, we're seeing a very large group of people who both went back to business. So our commuter business is very healthy and people who enjoy events. So every single event up and down the corridor. And we treat our guests as guests. We treat them in a hotel like way. When you go to our stations, they're gorgeous. When you ride our train, it's a different experience. And we're seeing people come back to travel for a lot of different reasons, and they're taking the train, which is super exciting to us. Yeah. So working to reengage after COVID. But I think there are some challenges that we stepped into in terms of where the infrastructure was for serving that guest experience? So in 2021, they made the decision to come back, and they made the decision to build a whole new infrastructure, whole new tech stack. By November of 2021, they launched that tech stack, Drupal Radix, which is a plain back-end system owned by Sabre, a whole other group of disparate solutions that didn't connect. And the way that we went after the guest experience was, every station was an individual station. Everything was a siloed technology installation, and all of those things equaled a broken guest experience. Step by step by step, it was not connected, and you couldn't talk to the guest equally in any of those stops. So the guest felt it, the guest told us, and we started to think about how do we reinvent that opportunity into a seamless end-to-end guest experience. And what are the things we needed to take out? What are the things we needed to put in? That was all part of the conversation.

So how are you thinking about it? How are you guys approaching transformation? Yep. So we took a very big step back. I came in December of 2021 to take over the digital part of the business, and I almost didn't take the job five times 'cause it was such a hot mess. It was that much of a mess. I was called the fourth or fifth chef, by the CEO because I was another guy who was going to come and put a menu in front of them to say, here's what it costs, and here's what you're going to go put out there. And is it going to work? Is it not going to work? Are we going to go through the same steps of learning and failing and changing? What we did do was put a big focus on people, process, data, and technology. And all of those things were meant to build trust with the leadership team, to build trust with our guests, to build trust across the board. And we picked opportunities to go after that were going to be meaningful MVP moments, moments that really did matter to the guest. As part of that, we added people along the way, we added technology along the way, and we made some big decisions on the way that we just laid out our tech stack in general. It took a lot of time. So in those moments, we found other moments to win. And I'll talk a little bit about the journey mapping that we underwent. Sure.

So first and foremost, I got to Brightline Trains, and our digital data was a mess, hot mess.

Everything was everywhere, different systems under desks in like stations, and we worked really hard to hire a new team because the old team left, and we started to reconstruct our data. Part of the data decisions were around what systems to use and what to keep, what to throw out, what to try new. And as we were doing that, we decided to build some trust across the organization, and we spent a lot of time on a customer journey map. And that's where WillowTree came in. At my previous job TNL, we did a mapping of everything to gain trust with our leadership teams, to gain trust with our peers. We did the same thing here. They did a fantastic job of helping us really understand what we were trying to do as a business, as a train. We're not a big train. We're not Amtrak. We're not somebody in Europe, right? We are a small train that goes from at the time Miami to West Palm, right? Three stops, Fort Lauderdale in between. We now have six stops up to Central Florida, and all of that was into consideration of how do we treat our guests and all of their differences along the journey. How do we give them the right content? How do we give them right marketing? How do we hit them at the right moment at the right time for the right reasons for the right cost, right? And all of that stuff was to build favoritism across the organization about we know what we're doing. We can achieve what we're saying we're going to do, and we're all going to work together to roll this thing out as a group, as a company, as an organization. So we're spending a ton of time looking at the journey, thinking about experiences across the journey. Of course, the first thing we have to look at is the architecture. Can the architecture deliver the experience? - Right. - Right. That we want in the end. So this was a full rebuild of the whole architecture. Total rebuild. So remember that menu? I had a new menu. The old menu was a costly endeavor that was millions of dollars loss to the business that didn't achieve what it was supposed to do, that didn't achieve what it was promised to do, and had probably somewhere in the range of 40 or 50 different types of vendors, SaaS solutions, and other things in the way of it making it work. So we peeled it back, we skimmed it down, and we talked about things in terms of simplicity. And the architecture really has a couple of bases to it. Siemens, they build our trains. They're Germans. They're very good. They're passionate about technology. They have three different systems that we use, onboard, which is the rail agent system. It tells us where our trains are. HAFAS, which is their operation system. It's a fleet management tool. And then Sqills, which is their backend system, which is all of the things that help us sell tickets. They're passionate about trains, and they're really passionate about technology. They came from technology to solve train systems. They didn't come from train systems to build technology. They care about technology. They care about us as a partnership. The second part is, I think, why we're all here, Adobe. I would be remiss if I didn't bring up Adobe in this conversation. Adobe plays a big role, right? So we started with really thinking about the experience side. How do we build our websites? I'm a big fan of retiring Drupal websites. Don't like them. I don't know if anybody here likes it. Sorry. I apologize. But I do like an Adobe site. I have a team of people who really embrace the tool, know how to use it, and we've known how to roll it out. But then we also needed a tool to consolidate our data and our personalization opportunities and our activation opportunities through AEP, and we did this all to really consolidate our guests themselves so that we can then represent opportunities in all of the experiences that we developed down the road. And then our backend systems and our frontend systems eventually talk to each other. We're using serverless backend systems so that we can serve things up faster, API structures. We can do everything on the fly as opposed to before it took us a month to deliver code. And when we delivered it, it was broken. So now we have all the checks and balances. We're doing two-week sprints. We can do hotfixes on the fly, and we move like we're actually we know what we're doing, which is pretty exciting. So we moved to this much more logical backend where we have the backend for frontend middle layer. We have these new backend layers. And then we have Adobe serving data and content on the frontend. So that's how we're thinking about the infrastructure. The next step in the process is delivering those marketing experiences, those pre-login experiences.

Yeah, we spent a lot of time thinking about how this project was going to run. And by the time we got to the starting point, we were struggling with our deployment of content, deployment of assets, and supporting the marketing teams. So we spent the first four months developing components, developing those libraries, building out our experiences, and getting those pages ready for the marketing team so that we can launch an Adobe Sites experience in front of everything else all the way up to the login. The login we had to do separately 'cause we're going to build a new backend. Hadn't even decided on the backend yet. We did decide on Okta as our login, our customer profile so we could tie it all together. But that website enabled us to start have to go all the way down the road to launch something 12 months from now. We started. We built our chaps. We learned. We grew from that, and our teams learned the tools as we were going.

All of that stuff enabled us to prepare for deploying content on the fly and being responsive to our marketing partners so that we can hit everything that they needed to continue to drive that business on the short haul in terms of filling trains and bringing the return back to the business. And I think in all of that, it really gave us the confidence to run websites differently than we had before and to drive not only the content seamlessly out the door, but it drove our SEO up. It gave us all sorts of benefits for our PR teams to be responsive. It gave us opportunities to hit all of the marks as we march towards launching a new station in Orlando in '23. Right. And that's how we want to think about things in the Adobe world as well as the world at large, but start to think of everything as a service, right? Our personalization as a service, commerce as a service, content as a service, so that then we can serve that out to any experience, whether that's in-station kiosk, or digital signage, or the mobile app, or the website, etcetera. We're doing all of that authoring in one place. And we're serving that out to every single digital channel. Can you talk a little bit about how you approach bringing the data together? Yeah. So the first part of data was where is it? Who's storing it? Who owns it, right? Where's their governance over it? How do you want to use it? Are we protecting it? All of those things needed to be ironed out. What we realized was we had pretty good systems. So we brought a lot of data through Fivetran into Databricks. We visualize it for our business through Power BI. But we were missing was how do we build that 360 view of the customer? I've never been anywhere that I've worked where there wasn't a million systems that were disparate all over the company that you couldn't combine your full guest experience. We had the benefit of being a very small train that could in Florida that gave us the opportunity to really rebuild everything from the start, put everything into a place that was consumable. We have a customer mart. We've got a web and all the application data, and all of it fills into the CDP. And so now we have an actionable profile on every single guest that comes to us from email, from website, from social, from other so that we can say we're going to do something with that person when we're ready. And we're going to set the systems up to enable it. Before, we had all of these different systems, including Bloomreach. We had Antavo. There was a whole separate dataset that sat to the right that we couldn't do anything with that we had some guy offshore running point on delivering our emails, while we were also running a website and looking at traffic and seeing how people utilized our assets. And none of that came together like peanut butter and chocolate does when it really tastes delicious. So we decided to really strip it all out and put those systems together and use AEP as the opportunity to pull it all together. Yep. And there's peanut butter and chocolate. - Yeah. - Yep. I'm allergic to peanut butter but it's still delicious.

So bringing that all together and talk a little bit about the timeline, standing all of these up, what use cases you're looking at? So unfortunate for WillowTree, I had timelines to hit and I'm aggressive, right? So I said to them, I have to get out of a contract by the end of the first quarter. And they smirked and laughed a bit, and I said, "No. No. I'm not kidding. I need this live." So again, for all of you guys thinking about doing something, don't do everything. Do what you need to do to take the first step that enables the rest of the steps forward. I think we said we're going to go do this. We're going to set up AEP, CJO, all of the analytics package. We're going to do all of this by the end of the contract that I have with Bloomreach, and we did it. Did it in 12 weeks. And we didn't do it perfectly. Made some messes, deployed some bad emails, learned from it, right? Took that all in and continue to get better and better. And the best part was the relationship with WillowTree was to set us up for success to learn how to use the tools, how to set up the tools. And then it was on us to go hire and teach our people how to use the tools. And now we're activating data. We're literally activating data. We're taking opportunities in our different segments, different types of customers, different types of rides, different types of offers, and we're representing that stuff back to people automagically because we're using these tools correctly.

Yep. So we've rebuilt the infrastructure. We've launched content in marketing. We brought the data together. So now we have the opportunity to activate. I also talked about commerce. That was that logged in experience, the purchase experience, etcetera. Want to talk a little bit about that? I do. This is my favorite part. Okay. This is where the peanut butter and the chocolate come together, and it gets all over your face. The commerce part was tricky, right? So we had a system called Radix. It was part of Sabre. Radix is really good at taking off and landing, right? We're really good at leaving a train station and stopping intermittently throughout the day at different train stations. So we have origin and destination multiple times across the journey. When you can't sell those pieces, you lose your business, right? So we set out to look for a backend system that enabled us to do that, and we chose skills. And skills enables us to enable train, revenue management, train bookings, and everything else that's related to train systems because that gives us the ability to now sell every station and every stop. When we did that, we connected Okta. We built all of the pages for web and app. We launched native applications, and we connected that ecommerce system, and 100% of our revenue goes through that system. We don't take cash. We don't do anything else but digital and technology, right? So everybody's buying it on the app, buying it on the web, buying it on the kiosk, or they go to the GSA desk, and they connect that way. We replaced our backend payment models. We did all of the code itself. We moved to AWS. We've got a lot of opportunities in there to connect the dots end-to-end for our guests. So now every single thing that happens across the system goes into our data lake, tells us what's going on, and we're able to react to it. We have active monitoring. We have active L1 teams who are looking at the data all day long, at the issues all day long. They're tracking. They're monitoring. And more importantly, we have 99.9% uptime. We used to have 93 to 95% uptime. And I don't know if you know, but that's a lot of money that you throw out the door because you don't have a great system to connect people to.

So talking a little bit about the journey to that 12-week journey, right? So I think I've talked to a lot of folks, like yourselves over the last couple of years of how do we get started? How do we get out of the gates? How do we get value out of this new stuff, out of the CDP, out of, CJA, AJO? So for us, that journey is always about focusing on where the value is, right? So focusing on, first, a use case-driven approach to making sure that we're after the value first thing. And then like Kevin said, getting out of the gates with something that's not perfect that we can build on and that we can expand. So our approach is looking at what are those use cases, then how do we go and build it out, starting with segmentation, segmentation in the CDP. So understanding who's critical to the business, where is the value. If we could do a couple of things better to drive revenue, what would those be and how do we focus on those? So setting up what are some key strategic segments for us new riders, zero rides, one to two rides, pass holders, premium versus smart folks. The cool thing about this, and I think for those of you that are working on these type of tools every day is Kevin's whole point of data being trapped in all of these different places wouldn't even allow us to get this far, right? Wouldn't allow us to get, let's set up a segment where we know who's buying what with what frequency. That's just that we can market to. Those type of things with no data exports, imports, calling IT, doing this thing. It's all being done in the tool. So starting with these segments, understanding how to build them, understanding what the offers are going to be for them, and how we can activate. And then-- And by the way, one quick thing. So what we didn't know when we launched our Orlando business, our business down south is pretty predictable commuters, event-driven, heat games, all that stuff. When we launch Orlando, there's this idea of people coming in from multiple airports, international airports. We're in the Orlando airport, so our train station's right there. 80% of our guests now are new to system, so if we don't have that data coming into the system to reactivate them, you lose them completely, right? And part of our charge is to take that 80% of that Orlando audience, that new to system audience, and really build a relationship with them so that they understand that the train is a viable solution to driving up and down the turnpike, and it is a viable solution to taking a flight from Orlando down to Miami. You could be on a train for three hours. You could be on a plane for seven hours. Could go the other way too, but we don't tell them that part, right? But we have the data in the system to activate. Sorry. Yep. Nope. No problem. And so simple use case of taking an audience, right? We want to target all repeat riders who have not made a purchase in 30 days, right? And then we can look at the data and say, look at all of the purchase events, calculate who's buying a premium ticket versus a smart ticket, add an attribute then to that profile that says these people prefer smart versus premium, and then start to build campaigns out in CDP and AJO that then move people from this offer to that offer. And then I think the critical point then is having the journey track that we can analyze, right? So then once we launch these campaigns, and this is a planned campaign approach. And once we launch that, then we can start to look at, well, where did they drop off, what adjustments might we be able to make, etcetera. So this is all table stakes. Once we get everything in one place, once we have the data in one place, we can start to do these type of planned campaigns. But then also, we can start to do a lot more on the real-time side as well, right? So instead of having a plan, it's something that's more campaign based, having things that are event based, right? So a train is delayed or canceled, we can immediately push that event into the user's profile, trigger a journey in AJO, send a notification in email or in app, and then push them into a rebook. If they don't rebook, put them in a different journey, which is a reacquisition journey that then pushes $20 off or a free premium ticket or something like that for an event that just happened. So these are things now that moving to a more automated real-time approach from just the planned campaign approach, all enabled by the fact that we have all of this data now at our fingertips. We've centralized it, and we've got the ability to work on it. The funny thing about data is that it can be paralyzing our C-suite sits in and they cannot get their heads wrapped around how to tell a story or figure out what to do or how to activate. We have all of this data. Part of it is take that out of their hands, put it in the system, and let the system activate opportunities, right? Build little segments, little models, start to do stuff because if you do stuff, at least you're in motion towards the right opportunities of growing and learning and changing. If you wait for the C-suite to figure it out, it won't happen, right? You'll just be paralyzed. So I think sometimes you have to really feed the beast and just put some data in there, some ideas in there, and see what happens 'cause you're going to learn a lot from it. And as you activate those things, they start to tell you more about who you are as a business and who your customers are more importantly, which then is amazing unlock for the activation ahead, right? If you do nothing, you get nothing.

Talk a little bit about what else you're using the infrastructure for where you're going in the future? Yeah. We set out. I set out to not be locked into waiting for shit to get done. Excuse the S word.

You sometimes have systems that are unmovable, teams that are paralyzed because they can't deploy systems that don't connect that make it really hard to adjust. What we built was we built an API structure that enables us to do hundreds of hotfixes and releases and configurations a week if we wanted to. We do two-week sprint cycles that enable us to deploy frontend connections to systems on our app and our websites. But most importantly, when our teammates in the field need something, we can tap into the systems and build tools through Power Apps in Microsoft that enable that data to be consumed. We built a couple of different tools that enable us to look up guest experience stuff so that when somebody comes to the front desk and they have a problem, they can find the answer of what's going on in front of them immediately without having to open up 12 different things. We build tools for our people on board to look at the manifest and say, are you in the right seat or you're not in the right seat. And then we put information on the boards that's sitting in our systems to tell you that a train is on time or delayed. Seems simple, right? Like, seems simple to be able to do that, but when you don't have the right systems connected and you can't connect those systems to other things, they don't work. And so we've really streamlined everything so that we make it available to consume anywhere. The stuff that we're doing for our onboard, on-time, delay stuff, we feed to the organization that runs the Orlando airport. They consume that data, and they publish it to their TVs across the airport to tell people when our trains are on-time or not. So we can give this information to anybody, anyhow, anywhere. It's because we've made it available, and we made it secure, and we made it useful is why it's out there. So we're using it everywhere. Nice. Talk a little bit about outcomes. Yeah. I don't know if you-- We're in Vegas. There's a train that they're going to build here. I'm not supposed to talk about it, so I won't talk about it.

But this is going to be the first bullet train that we're going to build from Vegas to Rancho Cucamonga. And in order for us to support that team, what we did was we built a website for them using the same AEM stack, right? Team put together through a lot of work, hard work. They're sitting in the audience, so I want to say the right things. But they were able to spin up a website that was for a completely different business entity under our enterprise view and deploy that site so that they had PR and content and context for what was going on in the news about the government loans and all of the different things that come with Brightline West. So that is one example of a scaling. There's so many of those examples of us just being able to spin up things that enable our marketing partners and everybody else. And some of the impacts are huge.

We have a goal of hitting 2.5 million people in our database this year. It's a tough one because we're at 800,000 in email, but it makes it a lot easier because we now have reactivation programs and all sorts of things through AEP. We're also-- We launched, I'm going to brag a little bit. We launched the first satellite on a train across the globe. We used Starlink. We ripped out the whole network. We put a new thing in with satellites to drive free Wi-Fi for the whole train. We're doing amazing Wi-Fi. As part of that, we're going to use an AEP or Adobe website to capture people's information on a splash page just like you'd probably login at your hotel room to get in the internet, and all of that information is dropped right into the CDP. So it automatically attaches to people who are on profile. It creates new profiles for the people who aren't on profile, and we do somewhere in the range of 10 terabytes per train per month. And we do 30,000 to 40,000 unique devices per month per train. So we have 10 trains, so 40,000 unique devices. That means every single one of them will have to login to get the internet on our train, which means all that data goes right into the CDP. So app downloads, we're killing it. We're doubled what we were before we launched our new web and app. All of those things add up. And some of these numbers are quite frankly, super exciting and impressive to me because a lot of times you launch something new, you change an infrastructure, you change the whole stack, and teams have to learn a lot. They have to change a lot. Guests have to adjust. We didn't see a blip. We saw 4x our sales. We see 50% more downloads on the application. We did all of it in 18 months. Another task that I put on my team members to go fast and furious to meet our deadline of launching a new station in Orlando, which is a big entity for us. And in all of the opportunities, our page loads were 13 seconds per page load. They're now down to 900 milliseconds average per page load, right? So the changeover to serverless to AEM to all of the things that we've done, the power of all of those data points equals, and the 99.9% uptime on our backend, all of it equates to more people through the funnel, more sales into the coffers, and more repeat ridership because now we have you in our customer database, and we can represent offers to you whenever, wherever we want based on your needs, based on where you go, how you go, why you go. So it's pretty exciting. Yeah. It's awesome. Huge shift to a digital first infrastructure and experience across the board.

So with that, I think we wanted to open it up for any questions you have for either Kevin or I.

[Music]