Florian Martin on Lufthansa Group’s Journey Toward Offers & Orders and Modern Retailing
In this video, Florian Martin, Head of Commercial Offer Methods & Automation at Lufthansa Group, talks with Justin Jander, Director of Product Management at PROS, about Lufthansa Group’s revenue management and pricing strategy, and the group’s next steps into modern retailing. He shares his view on how data, offer optimization, and revenue management play a critical role in the future of airline Offer & Order management.
Video Highlights
- 00:53 - Florian Martin on how Lufthansa Group has approached their revenue management and pricing strategy, and how it is evolving.
- 01:56 – How Lufthansa Group went about stripping down the internal complexity around traditional revenue management.
- 03:34 – Florian Martin on what “an offer” really means in the context of revenue management and offer optimization.
- 04:54 – Learn what Lufthansa Group's next steps are in their evolution toward modern airline retailing.
- 06:38 - Florian Martin on how data plays a role in the future of airline Offer and Order Management.
- 10:04 - How Florian Martin sees the role of revenue management in the future.
- 14:33 – The value Lufthansa Group sees with PROS RM and PROS RTDP solutions for revenue optimization.
Full Transcript
Justin Jander: So, thanks Florian for joining me today. Really excited to get a chance to talk to you about all things offer and order management. And so, to kick us off, why don't you tell the audience who you are and a little bit about yourself.
Florian Martin: Thanks, Justin. Always a pleasure and thanks for having us. Yeah, I'm Florian, Florian Martin working for Lufthansa Group for, so for all hub-based carriers in the group, Austrians, Swiss, Lufthansa, Brussels, you name them. And basically, we are doing all the R&D, the science, but also the whole business intelligence and everything system related, whereas you know, we got plenty of them....
Justin Jander: Indeed, I guess with that many airlines, it's a lot of hubs, a lot of complexity in the network. So, that probably introduces some unique challenges to how you approach the business. Can you describe a little bit about how you've approached your revenue management pricing strategy and how that's evolving as well?
Florian Martin: I think whenever you talk about an innovative subject, like Offer & Order, but it could be any other in the realm of revenue management. For us it was always key to first do our homework and get all the carriers on the same platforms so that we are ready for that innovation step. We learned, also the hard way to be honest, that doing all these integration efforts, having very different basis to start from, different backgrounds, different process, different systems, that really hinders you in your innovative approach, and in the end you're not faster, you're actually slow and it's a lot more expensive as well. So, we have really deliberately worked on this journey to also together with PROS on let's say bring everything onto the same platforms, have all the airlines of Lufthansa Group on single instances, cloud-based, such that it's scalable and we have this kind of level playing field to start from such that we can then make the innovation step jointly.
Justin Jander: So, I think towards that innovation step, the next thing that's sort of on the horizon is this idea of disentangling. And so obviously I'm going to say give me a one-minute summary of it. It's a lot bigger than one minute of course, but can you give us a little bit about what is that whole thing and what does it mean?
Florian Martin: It's really a topic that's very dear to us. It's an idea that has been floating around for a couple of years and we've been working on it ever since. First in-house and now in cooperation with PROS. And essentially, it's about disentangling or separating all the complexity that an airline faces internally in terms of revenue management, like all the network complexity, the fact that you have thousands of flights kind of interacting with each other, and you sell those over extended period of time, up to a year. All that kind of complexity is an airline perspective actually. And if you want to be very granular on the willingness to pay side, you need to take a customer perspective, a market perspective. Well, as a customer you don't see all this complexity and your willingness to pay doesn't depend on it. You don't care about that the airline has such a huge network and everything kind of holds together. So, we really said let's divide and conquer and have specialized models for let's say the part of the capacity optimization, which involves bid price on controls and everything, and have another machine learning based model that is really targeted to trying to estimate willingness to pay as granular and as accurately as possible and kind of bring it together and do modern revenue management with it, including continuous pricing and also then in the end being ready for offer and order.
Justin Jander: Alright, so let's talk about offer and order. So, how do you see that structure that you have today? What does offer, let's focus on offer for now. What do you see that offer really means in the context of revenue management into offer optimization?
Florian Martin: I think offer is a lot about flexibilizing the product dimension of it. If you think historically what others have done over the last decades is always kind of getting better with the pricing of things, but it's like the stuff that you actually sell. So, the product itself is kind of predetermined and fixed and it's pretty limited also in what you can sell actually. And I think we have made great advances on the dynamic pricing side of things. Now it's just time to also get into the dynamic product dimension in some sense. And it's certainly going to be a stepwise approach. It's going to start with just make the shelf a bit wider, maybe, a bit richer in its contents, but in the end, it's also going to involve more flexibly determining what the product actually is and maybe targeting it a bit more to the customer segment that actually approaches you and you're not always suggesting the same things irrespective of customer needs.
Justin Jander: So really creating, the term offer is being used all the time now, but offer really thinking of it as all the things that the passenger can buy. So, in that sense, what's the perspective on the offer being? How much of the stuff is stuff that you sell versus third party ancillaries? Where do you see that evolution going, starting and going?
Florian Martin: I think we are all pretty aligned on the vision that, of course you say airlines want to be more also as a retailing partner for travel and anything that kind of relates to it. I think what still needs to be determined is what steps do we make in what order until when, and I'm sure there will also be some different paths being taken by different players in the field. I personally believe that it's a lot of got to be flexibilizing what you can already sell maybe at one or the other external ingredient. For example, we are big on the multimodal part of it. We try to bring in other means of transport, like trains, into our offer, which for us is also in terms of megatrends, like sustainability, a very important thing to do, whether we going to add the Uber ride to the airport as well. Let's see. And when we're going to do that, but I think first we just want to make sure that whatever request comes in, we can target our reply in a better way to the specific customer segment and then do all the magic of AI-based continuous pricing on top of that.
Justin Jander: And so, I guess then when you can add the order piece of the puzzle to this, the value of order, rather than it's not just a record keeper, right? But the value that it adds is one, providing some of the data in the catalog, for example, so that you know what you can sell, and the conditions associated with that. And two would be the actual data, right? Feeding in the wins of the win-loss data, the bookings being the wins, feeding that back in. So, can you talk a little bit about how data plays a role in the future of offer and order management?
Florian Martin: I think it's going to be huge and for me, on the one hand I think you need to make sure, and that's one of the key stories of offer and order in general, that you as an airline are actually in the position to get that data. If we wouldn't do offer and order, other players will take the field essentially more and more. They will fill in all that needs that airlines cannot cater for. And that means that there's going to be someone in between the customer and the airline, which means you don't have this direct customer contact, which also means you're not going to get the data or just very little information that the partner might forward to you. So, I think it's kind of imminent and that's one of the major motivations also behind of an order that you keep that customer connection and once you have that secured, then you will also get all that rich data that in the end you need for all the fancy AI based things, whether it's on the product or the pricing side that you want to do.
Justin Jander: Yeah, it makes a lot of sense, and I think obviously airlines were kind of the first big data company, right? Back when big data was the buzzword, big data was the thing, and airlines were like, oh look, we have all this data, it's only grown, so there's only more and more data that you have at your disposal, but ensuring that you're getting the right amount of the right data in those high volumes is really critical.
Florian Martin: And what is really worth collecting is also right? The volumes that airlines are dealing with are almost unheard of, and we know very well what we sold, but what did we offer? That's a different question. So, tracking everything that we kind of offer to the customer, which might not always have led to a successful sale, I think that's the big challenge because that's a lot more than what you just sell.
Justin Jander: Yeah, and that's I think one of the most interesting parts because on ancillary pricing it's a little bit easier to have the wins and losses because you don't present the offer of ancillaries until after you've selected an itinerary. But if you think about the offer on the itinerary side of things, on average you're probably getting 20, 30 itineraries to choose from with potentially 20 or 30 different prices, and that's just one direction. Your other direction is another set of itineraries to choose from. Just that amount of volume of loss data is huge. So, then you have to decide how you can sift through the noise. Especially I know something you guys deal with a lot is bots hitting your, your webpage, your sites, and taking data from your site costing you more money, of course in the process.
Florian Martin: I think you're immediately in that realm of where you need very smart routines that already govern the data collection part of it, and then you need very sophisticated architectures that are able to cater for the data that you actually decided you want to keep, which going to still be huge. So, I think that's also where you as an airline, you really be up for that game and you need to partner with the right players that will enable you to keep up with these requirements because we know one thing for sure, some of the very traditional data storing techniques will not cut it. So, I think that's an often overlooked challenge, but a very important one.
Justin Jander: So maybe a slightly controversial topic, I don't think it really is, but it could be, what do you see the role of revenue management in the future of offers, especially given that a lot of people want their hand in the offer space, people, the e-commerce team says that offers are theirs because they are actually presenting the offer to the passenger. So how do you balance the fact that there's a lot of touch points from the passenger all the way to the revenue management system? How do you balance the ownership of offer given that there probably will be multiple owners in the end?
Florian Martin: Yeah, it's certainly a topic that is hotly debated either explicitly or implicitly at the different players. I think what we see in the e-commerce domain of course is that if you have control of the customer touchpoint, the sales channel, then you have all that kind of contextual information. You can do a lot of optimization of what you actually display when to home already with that, that usually results in some form of conversion or take rate maximization, which is not bad. It's just I think worth the question why do you do it? It's probably because you cannot do actual revenue maximization. So, that's of course much better. That's what you want to do. And I think this is where RM comes in because revenue maximization is all we do. It's just that we need to keep up with that. It's actually a challenge, I think for revenue management to involve itself in adapting its processes to be able to use all this rich data and make sure we include it in the revenue maximization where it's supposed to be and not start a kind of shadow optimization that actually is not in sync with revenue maximization in the end.
Justin Jander: Yeah, I mean the classic anecdote is you want to maximize conversion, set the price to zero. Exactly, and you'll do great, but clearly that's not going to achieve any kind of revenue goals or profitability goals for the airline. And I think that extreme example is obviously facetious, but there is an element of showing the point of why conversion can't be the only objective.
Florian Martin: Exactly. I think the one thing that really comes out of this whole offer and order discussion on the highest of levels is just that all of these things are connected, right? Like what you display, which is the selection of the product for this customer and what it costs to the customer, or the price of it. These are decisions that are intrinsically connected, and if you rip them apart, you are going to lose a lot and you're going to create a lot of problems. So, of course, I think the real approach has to be the first thing we try is let's do it together because that's where the real optimum is.
Justin Jander: That makes sense. And I think that really puts the full picture on offer optimization is it really is about every step of the journey, whether it's the itinerary, the extra products that you have, the price of those products, how you bundle them, how you show them to the passenger, all matters. And I mean, even where we are today, the discussion of what order do you show itineraries and has been around for a long time, like ordering itineraries, do you do it by shortest? Do you do it by cheapest? Do you do by where there's less demand because you want to promote those, tons of different things that you can do? And so now we're just adding to that complexity with other things that you're offering the passenger.
Florian Martin: Absolutely. I mean, internally we often refer to it as business transformation because I think it even better describes the scale of the challenge. There's almost no functional airline that is not affected in some way by essentially going away from traditional PNRs, tickets, EMDs, you name them. Everything is kind of based on that somewhat. And if you replace these things, you've got to rethink almost everything. The good thing I think is, I mean, it can be daunting, right? You're like, oh my God, I'm going to transform my whole business. Can I pull it off? I think the important part is you can do it in steps. You need a very well thought out approach. And again, that's I think where it's important to partner up with the right ones that have some experience that, have already thought about all these challenges, and then just make one step at a time and you're not alone in it. The whole industry is facing the same challenge. It's just you maybe have different starting points, but the target I think is pretty the same.
Justin Jander: So, I think we can then, with that in mind, we can go back to the specific products that you're using with RM Advantage and RTDP. When you think about those, one of the things you've said to me this week is talking about how they can be, those systems can be RBD free or class free. Can you talk a little bit about that? How that's those incremental steps towards the full thing. Can you describe a little bit about how that works in your mind?
Florian Martin: Yeah, I think it's also a good example of how such a transition can look like, in this case for the use case of Core RM, let's say. So, basically bid price generation and buy-on adjustments and how you get to an availability and a continuous price in the end. Of course, also where we are coming from is a very legacy, class-based, fair-based world, right? So, our data is structured like this. The systems that have been there before are working like this, and the output it generates also looks like this. So, what we just did is we started together with PROS essentially on making sure the core RM systems are kind of agnostic to this class-based information. So, they're kind of internally continuous, but when they get data feeded in, it might come in a RBD fashion, but it kind of gets then mapped into a kind of price demand relationship and what comes out of it equally might be continuous, and then you map it back into an RBD world, and this is kind of how it looks today.
Florian Martin: It's kind of beautiful actually. You can actually deal with discreet and continuous information. You can also output both. So, you're kind of in this transition state where you can do both. And of course, the outlook is then to at least first maybe the RM itself makes itself completely cost free, which means we're not the bottleneck anymore. If we in the offer and order journey decide, hey, we're going to replace this and that engine or this and that process or revenue accounting system, whatever, we actually will not have to post a requirement, but hey, we need RBDs, and classes and everything and we're going to mess up everything again. So, I think that's important that everyone does first, the things he can do in his shop and has a bit of hindsight of what's coming, and everything is so much easier in a continuous world. It's beautiful, it's elegant, it's neat and tidy.
Justin Jander: And it starts people thinking like your analyst especially, it makes them start thinking like economists or something a little more mathematical in the process rather than like Q, Z, M, Y, these letters of the alphabet. Breaking that into being an actual price demand.
Florian Martin: I think we command, I think, have gotten so used to these things that we don't question them anymore. Right? But if you think about how long it takes to explain to a newcomer how you actually, as an airline, get to a price, you said with all these letters, need to know which ones are part of which compartment, then all the fares and you have bid prices, buy down adjustments or fare adjustments. It's hugely complicated, but if you would explain it just like you have an opportunity cost, it's a bid price, that's some value, and you add a margin on top, which reflects the specific way you have to pay up the customer. You add some variable costs, some taxes, and you're there.
Justin Jander: Exactly. Imagine being a college graduate, coming to work for an airline who's an economist, economics major, the first thing you described, they would've no idea, none of that would even come close to relating to economic principles. But literally what you just described is probably first-year economics and allowing somebody to think in that mindset changes the whole game because now you're thinking about what happens when I change the price, demand goes down. How fast does it go down? Instead of thinking, when I go from M class to Z class, what happens to the demand? That's not a natural thing to think about. So, I think it's a really great approach and yeah. Well, I want to thank you very much for taking the time to talk to me today. It's been a pleasure and looking forward to continuing the discussion and the partnership between PROS and Lufthansa.
Florian Martin: Great. Likewise. It's always a pleasure to exchange and also to collaborate so, much appreciated.
Justin Jander: Thank you very much.