Dave: We're super excited about this next panel. We're going to bring some energy, keep it lively, maybe be a bit provocative and we absolutely want to hear from you. So please log into the app.

And for this panel, we want to hear questions. We want to keep it very, very dynamic. So, Dave Harvey, Harvey Global spent twenty-five plus years airlines, mainly commercial tech roles. Now I'm giving advice, I'm consulting, advising.

I've got some dear friends, longtime friends in the room, some new friends. So, this is a neat opportunity. I did want to give a quick shout out to Brett, the entire ATPCO team. This program has been phenomenal. It's been so smooth getting to this point, you know, working with Megan and Kathy and the rest of the team. So huge thank you.

Let's, I want to introduce and let the gentleman, Brent and Matthias give us a few moments on their background before we kick into the questions.

Brent: Thanks David and Aloha everyone.

Dave: Love the shirt, by the way.

Brent: Thank you. Try and bring a little Aloha wherever we go. 30 years in the industry, I spent time with American, Etihad, and I finished up as the Chief Revenue Officer at Hawaiian Airlines. I was fortunate during that last tenure, so this event is near and dear to my heart. I was a board member of ATPCO for 10 years. So, thank you for participating and attending. It's a great event and then kudos to the ATPCO team for the event today.

Post airline, I started with a travel tech company and I'm now the Chief Commercial Officer for TravelX, a company that focuses on post booking revenue management. So great to be here. Thanks, David.

Dave: Awesome. Matthias.

Matthias: Yeah. Good morning, everybody. I'm Matthias Freeman. I'm a Senior Advisor for Travel in Motion. For those of you who don't know us, we are helping airlines in the transition to modern airline retail and consultancy. I'm particularly focused on offer optimization and dynamic pricing. That's also where my background is. I used to work for Lufthansa for several years and there we started off the transition of the revenue management systems also implementing continuous pricing back in the day.

Dave: Wonderful. Let's jump right in the first one is everybody's throwing around AI first.

I think we've, over the decades, it was digital first, then it was app first, we started writing these questions two months ago and now it's not even AI first, it's agentic first, but there's, that's been thrown around. There's a lot of confusion. What does it actually mean?

Kudos to a couple of panels yesterday that put the definitions up there for us all. And maybe I'll take a first stab and then let you guys go. I, just from an airline perspective, it's all about efficiency, complex systems, a lot of cost, a lot of moving parts moving together.

And I think the, I think a lot of the airlines right now we're thinking about how do we use AI to drive another level of efficiency and it's less about kind of AI native new business solutions. So well let me kind of throw that out and let's see what you guys want to hit with that.

Matthias: Yeah, I would agree to that. I think the confusion starts where you say an AI and it's all over the place and it could mean anything.

Dave: Larger than life.

Matthias: And as you say, the efficiency, but there's so many fold applications that you could go into. It's really a question of where and how to get started.

I think the great difference is if you're talking about the interface to the customer, otherwise internally you can decide for yourself when to get in and where to start. They're going to be driven by the customers.

So that's the greatest confusion I think that's still in there. And also, how the entire value chain is going to develop, right? Which plays are going to win, which plays are going to disappear potentially? How is that commercially? I think from the talks that we've heard before, I had the impression that most are talking GenAI agents that are internally owned. But I think the more important is to think about the external ones, the ChatGPTs and so on. Who’s going to own the customer? We're back to that discussion now.

Brent: I think it's kind of twofold. I think there is kind of that operational efficiency, and you think of the operational side of the business, what it can do to drive things there. I think there's a lot of opportunity there. It's a complex business with, and it can clearly declutter a lot of that. I think that the commercial side it's interesting when I go talk to customers about kind of AI and their journey, everyone's really interested in it, but they're not really sure what steps they should be taking.

And part of it is really they're taking their cue from their customers and customers are interested in using AI to start their journey and look into it. But still only two, you know, 2% of people are using AI to kind of buy.

So, I find commercially people are interested, but they're not ready to fully commit. And so how much it really it comes down to kind of trust and transparency. And so, when we're talking to folks, it's data trust and transparency. And that's going to sort itself out here, I think. But we're still in those early stages where people are kind of figuring out what that means to them, and it means different things to different partners.

Dave: You know it is interesting. There's kind of this level of stress we all have like every day if we're not testing, learning, doing something, we feel like we're falling behind. We have seen and Matthias, I'm going to tee this one up to you. This is a little bit more on the provocative side. We are starting to see more and more carriers get out there and start making some larger bets in this space.

So, over the next two to three years, what kind of bets are we seeing that may be missing the mark from your perspective?

Matthias: Well, I think maybe we should better be talking about what we should do and not what we should not do, but well, if you ask me like that, I think it's dangerous to actually think it's just a change in the customer interface and we put in a layer on top of our legacy systems with all that fragmented systems and data.

I think that's not going to work, but it's going to be amplified with AI. So, you need to get that ready. Customer expectation is going to increase, right? We will have more information, more insights about the intent. So, it's also the expectation to use that and you need to have the back-end systems for that.

So yeah, probably shiny nice demos put on to fragmented data and systems. That's probably what I would not do.

Brent: Yeah, I think absolutely getting your data sorted out and cleaned up is going to be imperative. If you haven't done that, do that. It's a journey, but it is going to be vitally, it's always been important. It's going to be more important now than ever as Matthias mentioned. I think the other challenge for airlines is and particularly folks in revenue management is we're always chasing the next thing that drives more and more and more complexity, and AI certainly has the ability to do that.

So, if we think of the end consumer and that complexity, how do we add complexity and value to the airline, but also kind of make it simpler for our customers. And I think that if we don't balance that right, that could be a real challenge and a real problem.

Dave: You know, and Mathias, I'd love to hear your take and, and thank you, Brent teeing it up and bringing it into the RM space. When you think about some of the modern retailing that we're seeing, is this evolution? Is it changing more of the core RM practices? We'd love to get your take.

Matthias: Well, I think for sure it's a great change, but the underlying economics don't change, right? We still got scarce capacities, we've got seats, we've got ancillaries potentially in the future and we are price differentiating. So, it's the willingness to pay of the customer on the other side. At the same time, it's going to, the scope is going to increase dramatically, right?

We're going to have offer optimization. So, it starts with compiling the offers in a contribution optimal way. Also of course having a customer in mind, so it's propensity to buy and contribution for the airline.

Then it goes to dynamic pricing. We're not, classes are going to disappear, fares are going to disappear.

We have ancillaries to price without the data to really have the price variation in it to determine the willingness to buy. Of course, we need to be continuous. If you have continuous or dynamic offers, we need every single price point we can take.

So, it also helps if you're already in an up-to-date revenue management system potentially with continuous pricing. Also, there it's a real time component that's new to revenue management, right? I'm not used to that. As said ancillary pricing, contextual pricing as well, right.

That's a real time component there. And then the final curation of the offer, again, the ranking we heard yesterday from Skyscanner, how successful they were with ranking the offers. So that's going to be the last step in that optimization.

But also, the decision space increases, right? If you're thinking about ancillaries, if you think it, flexibility, for example, you not only have the price, but you also have the fee you can play around with. You've got the timing, right, Brent, you're going to be talking about that, I suppose. So, the decision space is also going to increase. So yes, it's going to change dramatically.

Dave: Yeah. And I'm jumping around here a little bit, but I'd like to then layer on a little bit of AI disruption on top of this. If I think about just kind of base RM theory, it's effectively scarcity and control. And you control the offer, the timing in an agentic world when agents are now shopping, comparing potentially, holding inventory in a dynamic world.

Does RM, those things don't tie up right now? Like what, where do we go with all this? And does the kind of the locus of control potentially shift on us a bit? Any thoughts on that?

Matthias: Well, I think for example if you're holding inventory, you can price that right and you need to do that in the future that. So that's going to be OK.

I think volume is going to be a problem, right? We do see increasing look to book ratios already and airlines are worried about that already. But we have way away from the agentic world when the volumes are going to explode.

So, I think there we need to be thinking about the systems, the performance of the systems, but also about cost, right?

I think about that cost when all these agents hit your offer engines.

Dave: Yeah.

Brent: Yeah. No, I think airlines are going to need to, you know, figure out how to be effective communicators with customers, agents, but how they do that in a way that focuses on cost, delivering, being responsive to the agents in the right way, but doing it in a cost effective way.

And that cracking that model is going to take a little time. And certainly, the economics of how that's done is going to have to change I think.

Dave: Yeah, and Brett did a great job with his opener yesterday talking about the over 99% of searches are thrown away today. That's just going to accelerate how do we balance this equation of having the right offer the right place, the right time, agents, you know, doing all of this comparison and then the cost part of the equation.

Brent: Yeah. I think you're going to have to get to a point where the airline that can respond the most effectively to agents and the most efficiently. And that's going to be difficult. You're not going to be able to respond with the most robust detailed answer because certainly you're going to drive that ninety-nine down to 99.9999.

And the airlines are going to have to figure out a way to strike that right balance of how is not only being responsive to agents, getting them in the answer that they want, but just not overloading systems and managing the cost component. And to me the cost component, we're going to have to figure out a way as an industry, both suppliers and airlines, to handle that.

Matthias: Yeah, I would agree to you. I think also the vendors are not having solutions at the moment for that, right. Most of them are saying, OK, trust us, we can handle volume. But I think that's underestimating the problem.

In a dynamic world with agents, both of them driving the volume there, AI itself might be the answer, right? Smart, very, very smart caching. I don't like the word caching in the dynamic world, but very smart caching went to the offer engine and when to respond maybe with a less accuracy or other vendors we've been discussing with are trying to make it very fast and very cost efficient offer engine in competition to the legacy ones.

Dave: Well, you guys both kind of went with data foundation and kind of getting your house in order with that comment. You know, we heard Bridget this morning talking about the importance of data and trust with the end consumer. Nicholas, JetBlue yesterday was just talking about the fidelity of data. So, everyone's out there making decisions. As we think about the next two to three years, there's going to be leaders and laggards when we think about the data architecture and the foundation, what things should this group be thinking about as they're making some of those architecture decisions?

Matthias: Well, I think yeah, data and content is obviously key, and you need to be accessible and understandable by the agents, by GenAI models. The goal at first, I think, is staying visible, right? And then you can take it from there, take the whole transactions into that new world.

What you need for that, I said earlier on, you also need the back-end systems for that. So, you need the offer logic to construct all these relevant offers, but also you need to fulfill and service them, right? So, I think the investments, that is also often a question is the investments into modern airline retailing some cost now? No, it's not. I think it's a necessary foundation.

Speed is going to be important. So, I said become more, stay visible, monitor that, measure that and then you can take it increasingly from there. And the brand is important, right? Building a strong brand also in this new world will pay off eventually and you need to make sure you don't disappear.

Brent: Yeah, I think as Matthias said, I think, you know, having a clean data layer to start with is super important. But as you experiment, develop and add to that, I think making sure you're maintaining that level and then stay curious because you're going to, I think the explosion in different kinds of data in our customer base now.

You know, we give them a bunch of performance reporting, but the vast majority of them want the raw data because they want to be able to incorporate that into their decision making, understanding that and being able to take that. So, I think it's really important that folks kind of stay vigilant and stay curious, right? And keep those standards high but stay curious and figure out how it all fits together.

Dave: And Brent, I'd love to get your take on this, you know, operating at a high level with a carrier now more entrepreneurial startup Travel X having some success. What were some things on the operating side of the airline that were near and dear, and you thought about how to operate the business versus what you're seeing now on the other side? I'd love to get your take.

Brent: Yeah, it's been a big adjustment. Yeah, 30 years of being kind of in a structured corporate environment to being a startup travel tech, it is been a real difference for me.

And being 7 hours away from most of our team in Buenos Aires. I think part of it is the biggest thing for me, a couple things, I guess is pace. And so, in a startup environment, we're moving fast, things are dynamic, we're changing and that's great. We also don't have some of the same kind of legacy structure that is at times good but also times bad to where we're able to move dynamically.

And so, I, we were in my early months with the company, we were talking about an enhancement that I thought made sense to do. And I was like, OK, this is going to take five sprints in three months to get done. And they're like, no, it's going to be two weeks and we're going to get this done. And so, I think kind of resetting my expectations.

I'm the grandpa of the company. I'm the oldest guy at the company now. And I never thought travel tech would be next for me, but it is adjusting my mindset and also it just for me personally, it's from being, you know, an airline to hearing the voice of our customers and, and prospective customers. And so, it's been an interesting adjustment, but one I've really enjoyed. And we've got a great product and a great team. So, it's been a real fun transition.

Dave: Yeah. And Matthias, I know you have a similar journey. Anything you'd like to add to that as well?

Matthias: Yeah, sure. I think we started the transition at Lufthansa back in the days with continuous pricing. And it's that I think the transition we started there really helps us also in the new world because of continuous pricing, because of dynamic offers and the optimization actually is yes, very different, but very similar as well. So, it's not going to be all throw away. So, it really helps us there. We reduce a lot of that legacy complexity already. So, that's going to help us as well. And yes, AI might help us in the future there as well, like, you know, handling all these rules, handling fares, but also all these influences you have in the different systems, for example.

Dave: Yeah. Well, let's pivot here just a little bit. We've talked some about the customer but like to would like to spend a little bit more time here. Personalization, another big focus the last few years, could be a little bit of a buzzword. But what I'm trying to get to is anybody truly delivering personalized offer or is that more of a longer-term vision that you're trying to develop delivery mechanisms toward Matthias?

Matthias: Yeah, I think we are still very much segment based, right? First of all, we need to differentiate between pricing and content there. Yeah, I think personalized pricing is very problematic. I don't know anybody who wants to do it right for legal reasons, you don't want to irritate your customer. How would you even do that?

So, I think there we are on the content domain, we do have some airlines who use entitlements. You get them for free then. But that's about the personalization where it is. Otherwise, it's still segment based, but the good news is the segments are becoming finer and finer and finer. So, if that progress is going to continue, you're going to end up at some point in personalization if that's even your goal. I'm not sure. We've had a lot of discussions with airlines on this. Some are for it, some are for micro segmentation, but I think yeah, the journey is the right way wherever you want to go.

Brent: Yeah, I think we've progressed a lot as an industry. We still have a ways to go in our company. We do a degree of personalization, but it is attribute based, it's not individual base. So, we know things about the PNR that David is in, but we don't know anything about David.

And so, I think you can continue to do that. I think the other interesting thing is we think about personalization and is the journey is not just the initial offer, the initial booking is that as you kind of move further down the value chain, airlines have the ability to monetize that customer not just at time of sale, but offer the right thing at the right time post sale as well.

Dave: And Brent, I'd love to hear, you know, TravelX, you guys are doing a lot of things post booking to monetize kind of every step of the way. Just how important is the initial booking versus that point to the time of departure?

Brent: Well, the vast amount of value is kind of getting as much as you can right the first time and so, you know, everyone in this room is trying to do that and the value of that is immense and the disproportionate amount of value for carriers is going to be in that. But as technology continues to evolve, I think ability to communicate with customers continues to evolve, having a deeper relationship with customers is going to enable carriers to get that extra post sale activity.

I mean we focus on post booking RM. So, where airlines may have not got it right the first time, we give them the opportunity to get it right again. And there's a real opportunity. But it's a high value transaction. It's not a low value. I want to sell you something for $10. If we go do a transaction with an airline, it's multiple hundreds of dollars of value for them and it's a win-win for the guests.

So, I think there is more in that space, and I think as we think not just of, we're focused on post booking our end today, but I think there's more and more post booking optimization and offer optimization that airlines can do. We're early in that phase, I think in terms of becoming better retailers post sale.

Dave: You know, would, love to get your gentleman's take on, and this has come up in a lot of other panels where the language models are going to be more, more and more the front door to this activity. So, the apps and websites that we're interacting with today are going to are going to have a lesser role.

I think about the tens of millions or hundreds of millions all the brands have spent AB testing, trying to optimize human behavior on those interfaces and now it's going to be agents doing this, like that is a huge disruption. What, you know, I know we don't, we don't have all this sorted out, but I would love to get your take on that shift, what this audience should be thinking about when the agents are actually driving that decision making and serving it up to the humans.

Matthias: Yeah, I said, I think there is an agent in in the middle, right? It's a gatekeeper. You need to kind of convince the agent to get to the customer that, so you need the data to be machine readable, understandable, and accessible, but it's still going to be the customer in the middle.

And also, the transition is going to be long, right? So, I think yeah, it's still worthwhile optimizing websites. It's not going to disappear tomorrow while you have this hybrid world right where you need to also feed the agents.

Brent: Yeah, I think if you look at kind of customer behavior today and I think how agents are evolved, they're going to be an information gathering source. But until folks get comfortable enough to actually use them for a booking component, which today is small and I think that will continue to grow.

But the pace of that, I certainly don't have a crystal ball as to what the, how that changes. And so, living in both worlds there's still value and, you know, website optimization and things like that. Maybe that diminishes over time, but there's still value of that while the agentic world is catching up.

Dave: There's a bridge. I'm going to jump into a few of the audience questions here and then we're going to wrap with really looking into the crystal ball 10 years out. What one of the questions that I don't see it here on the monitor, but it was effectively around the PSS. We're talking about AI first, Agentic first. You know, there's APIs and a lot of the architecture. Well, what, how do we see changes in the role, if any, with the PSS in this new world that we're entering into?

Matthias: You want to go?

Brent: Yeah, sure. I think PSSs are going to continue to have a role, again, I think we're going to have a have an evolution, not a revolution. I think what airlines and vendors are going to need to do is continue to work with PSS providers to minimize friction, because I think people are going to want to have more vendors working on more things and having more access to both data and APIs.

And the efficiency of that is going to be really, really important. And the effectiveness of that I think is going to be paramount to seeing this transition because airlines aren't going to do this all on their own. You're going to have more third parties in the mix. And so that relationship between the airline, PSS, and other partners, I think it's going to be really, really important. And hopefully the parties in the middle will handle it the right way.

Dave: Yeah. Matthias.

Matthias: Yeah, you could break it up, right? You can break it up into the offer and order piece of that. We do see that many of the especially large airlines are committed in the order space now and with the old PSS vendors for good reason. I think it's just a big risk if you have too small a company and you can be unsure if they can handle it.

At Lufthansa, we had a really bad experience couple years back with some vendor on our  dot com, right? Then Amadeus took it over eventually. That was a big write off. So, I think that fears is good. Of course, there's new players in in the game, which is good as well and then they doing a good job. So, it's worthwhile looking at them.

In the offer space, it's more free, right? So, there are a lot of new players also are worthwhile to look at. Yeah. So, the PSS vendors will still play a role and they do, they have most of the portfolio they can offer. So definitely they are players to look at.

Dave: Another interesting one here. I'd love to get your perspectives, maybe some hot takes here, but it's like we talked about cost and look to book, but at some point, these data centers, tokenization, all this computing power, somebody's got to pay for this. You got to pay the piper.

Any hot takes? I read, actually literally read something this week that there's real estate firms that are submitting one hundred claims for to build new data centers, knowing that only five are going to get through there. But they're just casting a wide net. So, it's like the gold rush. But how does this stuff all net at the end of the day from a cost equation?

Matthias: Well, I think the cost is going to come down eventually as well, right, as we evolve. But yes, it's a definite question. And that's also, for example, the question is NDC obsolete with MCP servers? Well, people tend to forget about this token overhead that you have in MCP servers, right?

So yes, you need to pay the bill eventually. The customer of course needs to pay the bill. And the question is, again, if you're thinking about the external agents, then that's not your cost, but then you're giving up the customer in some way or at least that has to show, you have to prevent that.

Brent: Yeah, I agree. I think you're going to see kind of a bubble in cost and that will continue to come down rapidly and thereafter. And you know, the airlines have traditionally been such high margin businesses, right? Just kidding.

Dave: Everybody except the airlines.

Brent: Yeah, fair enough Fair enough. No, I, yeah, there will be revenue and there will be value creation for the airlines on this. And so they're just going to have to manage that cost revenue balance and take advantage of the opportunities, test, experiment, leverage all that they can out of this to have a better relationship with their customers, to better monetize areas where they have value leakage today, and then manage that cost piece because inevitably it's going to be a challenge there.

I think over the, you know, not maybe today, but a year from now, two years from now, that's going to be a little bit more of a challenge.

Dave: We're going to do rapid fire down to our last two to three minutes. I one topic that I'm actually a little surprised we haven't heard more about is the role of voice in in all of this. Any quick thoughts about how that's going to play out? I'm old school. I like, I still like to type. I'm doing more and more voice interaction, but any perspective there?

Matthias: Well, I think it's definitely going to take off. Actually, at Lufthansa, we've been thinking about it years ago as a channel, but back then the quality was just so bad that it just didn't work out. But I can well imagine that, especially the new younger generations.

But also, it's going to be very efficient if you're just walking away and saying like, especially those that know exactly what they want. Book me the same flight on Monday again to London, right? That's going to be a voice for the more inspirational Shopping I don't think that's going to be too quick. But yes, I do believe in that. Definitely.

Yeah, it's not it's not going to play a role if you're typing it in or if you're talking to the agent. That's just the interface. You might do typing; I might do voice.

Brent: Yeah. Ultimately, it's just providing more avenues for customers to communicate.

Dave: New channel.

Brent: And so ultimately you need to meet your customers where they want to be met and how they want to communicate with you. And the technologies improves and continues to improve that I think it'll be kind of interchangeable at some point.

Dave: OK, last minute hot take. We're sitting here 10 years from now. Bridget talked about five years from now and just easing friction and just making it easier to do business. What, any perspective on what this thing looks like 10 years from now?

Matthias: Well, I think hopefully offers and orders are very native and we can laugh about this whole transition, 15 years of NDC transition and filed fares and so on.

And I think, yeah, agents are going to be around. So, some folks think, OK, and then five years is going to take off. No, it's not. It's going to be very quick. Look at the development you've had with GenAI. I expect like 2 years and then we are full of agents.

Brent: Yeah, I think we'll end up in a world. I mean it, it's easy to kind of take a take a pot shot at NDC and how long it's taken and what's new anymore. But I do think we're kind of in a window where things can really accelerate.

And I don't know exactly what it's going to be like in 10 years, but I do think, really the next, I'll call it, you know, maybe isn't one and two, but I got three to seven years in there. There is going to be a big change. And I do think people who adapt, experiment, test, are going to have a real advantage and continue to kind of outpace the industry in terms of success and innovation.

Dave: Well, excellent. Well, this was so much fun. Brent, Matthias, thank you so much.

‍