The shifting distribution landscape - how technology and consumer behaviour is reshaping travel distribution

Transcript

Martin Warner:Good morning everybody, and thank you, thank you for coming back for this last combined session. So we're going to demonstrate some flexibility in the travel industry with some last-minute changes as usual, but here we've got 2 groups that were originally going to be individual panels, but we've merged the two together. So let me start just by, by welcoming everybody, and you'll notice from the names that I'm only going to be brave enough to do the first name version of introductions. So let me welcome Golan from Kiwi.com, Sonja from PLAY, Marcel from Transavia, Roland from Aegean, Laura representing SkyTeam as an alliance, and Mustafa from SunExpress. So welcome everybody and thank you for joining me. I'm covering 2 sessions. The original session was going to be— well, these in combination are about the shifting distribution landscape, how technology and consumer behaviour is reshaping travel distribution. So the original panel was focused on social media. The second group was focused on—

Golan Shakéd:Advertising.

Martin Warner:AI. We're going to run the original scripts, so we'll take the social media landscape relating to really the advances in booking technology and shifting consumer behaviours, that they're coming together fundamentally that alter the way travel marketing and distribution is working. So let me start by asking Golan, to kick off with an opinion, let me get to the first question. How at Kiwi are you attracting intermediary customers to use flight comparison and OTA? I know you're seeing growth, but is that at the expense of other travel agents, OTAs? Is it share shift rather than attracting customers who've previously booked directly with airlines?

Golan Shakéd:Hello, good afternoon everyone. Thanks for staying for the graveyard So Kiwi.com is a travel tech company in its DNA, and what we do with our technology is we are trying to provide incremental revenue for airlines. And the way we do this is by combining unique routes to create new itineraries that are usually outside of the traditional airline network. So that creates an increment. And the other element that we are focusing on is trying to touch on underserved segments or markets. That is because we are focused on Generation Next. So sometimes these are not exactly the same audience that the traditional airline would focus on as a brand, but Kiwi.com is a brand focused on that generation. And with the fact that we have creative itineraries, we can sometimes find destinations that an airline, a traditional airline, will not necessarily spend their marketing budgets on. So I think that overall you can say that while obviously we're all competing sometimes for the same customer, having the technological capabilities, the technology capabilities, we can extend the networks and also serve some underserved segments for the airline as a complementary product.

Martin Warner:When we, when we talked in preparation for this session, I happened to— a friend of mine who runs my local bar in Mallorca, an Argentinian whose parents were visiting Istanbul on an extended holiday, and Andrés wanted to go and join them for a short weekend.

Marcel de Nooijer:Mm-hmm.

Martin Warner:And he came to me and said, you know, Martin, I can only find something here, it's over €700 just to go for 2 nights, so unless I can find something cheaper, I'm not going to bother going. And so I recommended that he take a look at Kiwi.com and came up with something which was half the price.

Golan Shakéd:Yeah, I mean, it depends on the target customer, and obviously some customers would go for direct flights, comfort, and we all know that.

Martin Warner:So he self-connected over Vienna in one direction, and I think it was Amsterdam in the other. So it was an interesting incremental— I know that that's a sample size of one, But that's an example of creating incremental opportunity and business.

Golan Shakéd:It is, but just creating the itinerary is step one. I think the magic— and Tom Wu spoke about technology and customer service— it's about how do you actually service a connection between 2 airlines that don't necessarily cooperate on the same network. And the key here is not only to build the itinerary, but to make sure that if there are First of all, the communications are streamlined into one customer. So we put those 2 airlines under one booking and we communicate to the end consumer the information that we take from sometimes multiple airlines in an itinerary. But much more importantly, if there is a delay in one of the airlines and all of a sudden you have a connection that is tight or you missed a connection, the key here is to build a robust enough business model that you can actually put the customer on an alternative flight and still stay in business. And that's what virtual interlining is all about. It's about building the itinerary, but focusing on the customer satisfaction as well.

Martin Warner:Now, given that you're the only non-airline member of the panel, I should have explained that that was why I was going to start with you first of all. But let me move this to ask both Sonja and Marcel for their perspectives around that shifting distribution landscape, is it enabling a growth in the proportion of bookings that are made directly by travelers? And how do you see what those share shifts have been over time between direct and indirect, and why? Sonja, let me—

Sonja Arnórsdóttir:Ladies first.

Martin Warner:Ladies first.

Sonja Arnórsdóttir:Yeah, we have clearly been seeing quite growth in the direct sale our website. We have always set our goal quite high on the direct bookings and have been having it around 80% of our total bookings. It's very important for us both on the yield side, passenger data. So Colin knows that and he has fixed it on Kiwi site for us. So the passenger data is very important. We are operating from Iceland. There are some weeks of the year that we have some disruption, so we need to be able to reach the passengers. What is interesting about the direct booking through the website is the traffic sources. It's clearly social that is growing quite a lot. I would say around the last 12 months, even a little bit further out, we are seeing quite good social traffic, and I think it's an indicator of customer changes in customer behaviour.

Mustafa Egilmezbilek:Interesting.

Martin Warner:Marcel?

Marcel de Nooijer:Yeah, like you're explaining, also we have seen over the years a growth in direct sales towards the website or the mobile app. What you can truly see is that our passengers are looking for a kind of frictionless mobile option, so that is really going pretty fast, I would say. We don't have the same percentage yet As you were explaining. So we grew, I think pre-COVID we were about roughly 45% direct sales. We're currently at 60-62% in this.

Martin Warner:That's quite a significant lift.

Marcel de Nooijer:It is, it is in a short period of time. But we worked hard on that, both in investing in the website or in the app, making it much more seamless and easy to use, and we see the results. Of course, it also depends a lot on the brand awareness to go to and have confidence in the brand for direct sales. And as a Dutch carrier, we predominantly have also Dutch customers, which makes it a bit easy. Therefore, I think it's also a balancing act to work together with OTAs, such as you are explaining, to grab that additional business in a smart manner. And indeed look for solutions for the seams, because that one is one of the elements, especially I think you also apply an ancillary revenue strategy just like us. So apart from the ticket, it's also on other elements of the journey which we need to capitalize on, and that is not always an easy thing to do with OTAs. And you want to prevent misunderstandings, because if you virtually connect and carrier A has this sizing of hand-carried luggage and carrier B has that size, you get a disconnection, whereas you think it's a seamless flow and it's not.

Martin Warner:Yeah.

Marcel de Nooijer:So these are the elements I think we have to work on. I think it was also mentioned in the previous panel about uniformization, that we use each other's services or requirements in that.

Martin Warner:It's a good That's a really good observation. Let me drill a little bit though into the elements of social media and how social media has changed the way that you as airlines, first of all, go to market. I'm going to stay with you, Marcel, on that question.

Roland Jaggi:Okay.

Martin Warner:Go to Sonja and then ask our other airline panel members if they have a view on that. There's no obligation to answer that because that's completely unrehearsed, but Marcel, your thoughts?

Marcel de Nooijer:Happy to do so. Social media as an outreach is really changing our marketing concept, so it moves us away from, I think, a static brand message which we throw out into the world, into a truly more customer platform engagement way of working. With that, we create a new reality. We have been working with, for instance, since June last year on TikTok, and it has been truly explosive in terms of reactions, views, engagement. We see as a result of that that the brand awareness has moved up with 10%. That also accounts for brand preference. And what it is actually is that we get in touch with a younger generation. So this is mostly the 18 to 34. Years of age segment, which actually get their information by this channel, not by other means. And that is working out. I have to say, we develop a lot. Actually, we develop everything in-house on the content we bring to the platform. But you also sometimes have to be a bit of a dinosaur. I'll give you a small example. 2 weeks ago it was World Pilot Day, and I've got a very enthusiastic young communication team. They built, in their view, a great movie about a pilot going to the toilet, putting some strings to the steering of the pilot, and then handing that over to someone in the cabin to say, can you control while I'm—

Martin Warner:While I go to the bathroom.

Marcel de Nooijer:No, some people will think it's great humor. This is exactly not what I want to be as an airline. So they came and said, well, this this is it for tomorrow. I said, this is not going to fly.

Martin Warner:Scratch that one.

Marcel de Nooijer:I love the fun part, but this is not what we're going to do. So that sometimes is a learning curve, but it's really building up engagement, it's building up brand awareness, and it's really inspiring young people, young generations to choose Transavia as a brand to fly with.

Martin Warner:Yeah, and that's also bringing first-time travellers into the brand. So how have you focused on which social media channels to use by age demographic?

Marcel de Nooijer:Good question. So TikTok, it's really in the segment of 18 to 34 years. We use Instagram for, I would say, 18 to 54.

Martin Warner:Facebook for me.

Marcel de Nooijer:Facebook a bit, yeah, through 25 to 64. And there's different content. Attached to it. So as I said, TikTok is all about the engagement, the fun part, to relate to one another. I would say Insta, it's about visuals to inspire, also to share a bit of activity. Facebook, we really use more informative about deals or product changes. And yes, we also use LinkedIn, but that's not for a sales perspective. That's more stakeholder management, to give a view on the always entertaining Dutch political developments. And yeah, we found it a very effective way of communicating and getting in touch and sharing our message to our customers, but also to other stakeholders.

Martin Warner:I find it fascinating because, you know, there are certain brands and products where you wouldn't dream of going to TikTok with. But it is, you're stimulating first-time travelers, you're stimulating the ideation, for example, that TikTok brings in destination selection.

Marcel de Nooijer:Yeah, it's changing dreamers into bookers, I would say.

Martin Warner:That's a really good—

Marcel de Nooijer:I think that's the element on what it does. And that's, yeah, we're excited on it and we do it for that reason also in-house.

Martin Warner:As Willy said earlier, I think I'll remember that, try and remember that one.

Golan Shakéd:Okay.

Martin Warner:What about from a Play perspective?

Sonja Arnórsdóttir:Yeah, I was thinking about it yesterday. And I think 5 to 10 years ago, it was very easy to just launch a campaign on Facebook. And all your followers saw it, you saw a pickup in sales, and it was quite easy. Today, it's very complicated, very costly. So you need to do it very smartly. The algorithm is very personal, not on the business side. So you need to have the experts to optimize it. It's, like I said, costly, so you're not spending money on something that's not working. So you need to do this very correctly. But the opportunity is also quite a lot. And he called this the dream platform because every one of us has been scrolling at home and not thinking about traveling. But then there's pictures that's coming and it's That we are inspiring people.

Martin Warner:From my own perspective, I think the word personalization in our industry is often overused, but I think social media lends itself towards a personization rather than a personalization. Would you agree with that?

Sonja Arnórsdóttir:100%. This is just the platform where we can plant the seed for everyone. For TikTok, We need to have it very bold, fun in the right way. And yeah, maybe more like inspire people, plant the seed that this is the destination you maybe want to see. And then you need to tell them that PLAY is the airline of choice to go to the destination.

Martin Warner:I saw everybody in line nodding of heads. Anybody else outside?

Roland Jaggi:I just want to add personalization. I agree that's too far-fetched. It's contextualization. If I have a day trip from Athens to Thessaloniki and back, why ever offering me a paid bag? I'm not going to have a bag. So it's contextualization.

Martin Warner:That's it. That's a good example.

Golan Shakéd:Can I suggest, I mean, I know the approach and we as a company that is Gen Next focused. We obviously spent a lot of time on social media and we've contextualized to, to the points raised, but actually in the last year we've taken a bit of a new approach, which is to take a step back from being active and recruit a sort of an army of not professional influencers, because everybody goes and pays someone, but actually recruit an army of ambassadors. And, and, and that is Basically something that we are experimenting with for the last year. So while we let them make the decision on what content fits the best platform, because actually, to be honest, I think they know it better than us, we also allow them to choose the content that they want to push, but we give them a bank of product developments or updates regularly, and we developed an ecosystem for them so that we can reward them based on the work that they're doing, either by paying them in credits to book flights or for any affiliation. And what has happened was, in a sense, we lost some of the control. And Marcel touched on a really good point, especially when you're an airline, you have to be very careful about what goes out. And we did, of course, put some measures in place to ensure it doesn't go gaga. But we have now, I think, 600 people and growing around the world who are talking about Kiwi.com randomly. Sometimes it's product marketing because we launched a new product feature like locking your price, or sometimes it's an inspirational piece about Greece because one of our ambassadors might be here and it's a good season. And what that did for us was it shook a bit the traditional way of going social media, it expanded our reach, and what we are busy doing is control and feeding them with content. So, and it's been working quite well for us. I mean, we're engaging with 20-30 million people every month that we don't even know.

Martin Warner:I mean, that's a fascinating number. Yeah, you know, the return on investment of 600 ambassadors, that's a, that's a very significant—

Golan Shakéd:And we pay them in vouchers.

Martin Warner:Mustafa?

Mustafa Egilmezbilek:One thing I want to add here might be also a challenging point, which is using the technology or AI for direct sales or for the social media using for this channel. So we already entered into the agent era, right? And we expect that instead of users, those agents which use the users' preferences to make those searches. And when we talk about Gen Z, they are really great at adaptation. According to the statistics, the usage of ChatGPT in 3 months increased by 100 million users. Almost 400 million users are weekly ChatGPT. Whenever they need, whenever they are searching, looking for something, their first place is getting to ChatGPT and asking for it. The behaviour is changing here. Either it's social media or anything else, should be ready to get or collect this information from our offerings for airlines and so on. This brings us to the point where, how are we going to be ready for those kind of interactions with the technology?

Martin Warner:Those are all really good points, and that probably takes us as a natural leap into the discussion about AI. And, you know, AI seemed to me over the last 12 months to just simply be an explosion—

Mustafa Egilmezbilek:Exactly.

Martin Warner:Of opportunity as well as activity. So, Laura, let me bring you into this. What have you seen, or what have your members seen in practical applications of AI in aviation, whether that's in distribution or in other areas?

Laura Abraham:Yeah, so there are a few areas that AI brings opportunity for airlines. You can look at this as a cost savings, operational efficiencies, creating better experiences. So those are, in my mind, kind of the 3 categories. And we see all of our major airlines are kind of either dipping their toe in the water, or some of them are You know, up to their knees in the water.

Martin Warner:Already, yeah.

Laura Abraham:With AI. You know, for example, at Delta, they're using AI to optimize turn times at some of the hubs. I was at Delta before SkyTeam, and I personally helped develop a natural NLP way to turn qualitative voice of the customer data into quantifiable data points that we then could spread across the enterprise. And that's really bringing together different data points to have a complete, I think to an earlier point, complete persona of the customer to make decisions. And then at airlines like KLM, we see things like using this to understand operational efficiencies around baggage, for example. Having said all that, at SkyTeam, the alliance, we have our digital products and our digital strategy really firmly placed in the customer experience realm. And so we're not really using or even thinking about AI in some of those other areas. We're looking at ways to use it really in service recovery, and we believe that that will help improve the experience, but also open some doors to operational efficiencies. But at the Alliance anyway, we're not thinking about using it for cost reduction.

Martin Warner:Yeah, well, that's an interesting observation because I think a lot of people are thinking, well, this will help me reduce costs. Yeah, I would prefer to think that this actually helps move our industry forward faster.

Laura Abraham:Yeah, yeah. So at the Alliance, we really think about our role in the digital space as a multiplier, not a— we don't want to duplicate what airlines are doing, right? It's my role to multiply anything that's working well and really scale it. And so that's how we're viewing AI as well.

Martin Warner:Let me ask Roland, you want to add to that or with a different perspective?

Roland Jaggi:I'd like to start one step before. We often hear AI is a revolution. I don't think so. Revolution takes place overnight at the palace, at the castle, nowadays at the parliament, at the residence of the prime minister. This is not what AI is. AI is for me clearly an evolution. It's coming. It's coming from all sides. You can do hundreds, thousands of things. Customer journey, booking optimization, revenue optimization, cost optimization.

Martin Warner:Absolutely.

Roland Jaggi:But you're not going to have this one big boom. The one big boom, a revolution, would be I replace my— excuse my French— bloody PSS overnight and I have offer order in place from today to tomorrow. It's not going to happen. It's going to be years. So for me, AI is evolution. And the key is we need to understand that this is way too often used as buzzword. And AI is going to solve all my problems. I'm going to be at the beach here in Greece 24/7. I am not. So it's coming and we need to make sure we understand AI, we understand the capability. We need to have a certain understanding. But, you know, we're an airline. Golan has 600 tech people. We have 500 pilots. We're not going to have 600 tech people. We are going to work with vendors, and one is doing this, another one is doing that, and that's where we need to be on top of it, and we need to get this into the company.

Martin Warner:I think what's a subtle, or perhaps not so subtle, change is that we've seen the earlier forms of AI in machine learning in the industry for many years now. The general public, the passengers, have not really been aware of what's that— of what that's been contributing to the way the industry has developed. I think in the last 12 months, the awareness of AI and the launch of ChatGPT, that's put it in very firmly in the public domain. So I think that that's the change, that all of a sudden it's out there in the minds of the general public rather than being used in an ML sense.

Sonja Arnórsdóttir:Yeah.

Martin Warner:We have an interesting Slido question that I'm going to pick up on from anonymous. New technology in booking platforms, is this the beginning of the end for GDSs?

Roland Jaggi:I volunteer to answer.

Martin Warner:You may go first. Well, we could be on this question for a while.

Roland Jaggi:We talked about distribution at the beginning. And we're all working towards direct, we all do. But there is a market for indirect. My brand is not known, corporate travel and so on. There will always be indirect. And for me, every booking is a good booking that is not a GDS booking. It's all technology, they are bloodsuckers, it's expensive, they have never been cooperative. It's okay. We have reduced our GDS segments to less To single digit. And it is the end of the GDS edifice. They try to be part of NDC. They will manage NDC to be part of it as well. But it is clearly the end of GDS edifice, and that end has already started.

Sonja Arnórsdóttir:I agree, definitely. I mean, it's— they should have stepped into the game earlier and make it more in the future. Distribution is very important, direct is very important, but at the same time, just like you were saying, brand awareness. We as a new airline, maybe not new, not many years more.

Martin Warner:Of everybody on the panel, your proportion of OTAs is probably the lowest.

Sonja Arnórsdóttir:Yeah, and we need the OTAs, we need all the media to get the brand out there, but DTS is not very helpful.

Roland Jaggi:And it's also good, it keeps us on our toes. Our digital director is in the room. He needs to be chased by the OTAs. He needs to be motivated. He needs to get to be the one who wants the booking. And likewise, I want Kiwi to get the bookings. This is okay. We need that competition. It's good for us.

Marcel de Nooijer:And yeah, we're 60 years of age and we're non-GDS actually as a local carrier. So there is room to live without the GDSs, I can confirm.

Martin Warner:Well, your growth suggests that's true. Mustafa, I could see you agitating at the beginning of that question. Is there something you want to add there?

Mustafa Egilmezbilek:Yes. So it depends on how they are going to— I mean, how GDSs are going to utilize or leverage the technology as well. I'm sure they are not going to stay behind, right? I mean, our GDS share is also small. For us, direct sales is very important. But there is also one dependency, which is the tech giants, I would say. In other words, owners of those ChatGPTs, LLMs, and so on, because the search is starting from there. If any search gets to the GDSs, right, I mean, depending on their agreements and so on, then definitely not. But if not, it might change. So, this brings us to the importance of having a very good brand in terms of maybe we need to revisit our marketing strategies, right? How are we going to be on the top list on those new LLM-supported search results? So, we used to have the search engine optimization at the moment on Google, right?

Laura Abraham:Yeah.

Mustafa Egilmezbilek:it will be different, and the agents will make the search, let's say, or the users, they will make the search through those ChatGPTs and so on. So how are we going to make sure that we will be in the top 10 results of the organic search of those searches?

Martin Warner:Yeah, and how do you keep refining your approach to that?

Golan Shakéd:Exactly.

Roland Jaggi:Yeah.

Martin Warner:Colin?

Golan Shakéd:Well, I think any new technology that enables more customers to be inspired or book is good for the industry, and I think it provides an amazing opportunity. Now, to Mustafa's point, we need to adapt, and whether it's adopting the way we communicate and our itineraries in order to be relevant to the new platforms, and to begin with, to be visible on those new platforms. Today we are experimenting and we've We're getting a few bookings a day from LLM-based requests, and there's a lot to optimize. But to answer the question that was asked, is it the end of GDS? I mean, I don't want to be terminal about anything. I just think it just adds another layer which grows, and it was mentioned here, it inspires more people to travel. It makes it more accessible to some people to book on that platform, and we should embrace it.

Martin Warner:There's definitely a proportional shift that's accelerating, that's for sure. And that takes me to not just the question of adoption of new technologies, but adaptation in order to use new technologies. I mean, I almost dread to ask the question what the average age of the tech and the development teams are, but it's certainly not going to be in my age group. That's for sure. That's for certain. Let me look at the next question. Future use of AI. Let me ask that question, and perhaps Roland, you'd like to pick up on that in terms of what are you seeing next in your use of AI?

Roland Jaggi:Well, it will really go across We do some things on revenue optimisation, upsell, fare families. We do that together with the University of Glasgow. We do things with our local IT powerhouse that helps us. Lost baggage, for example, taking pictures, and our communication from the crew. to ground, voice of customer where we get the feedback, email sending optimization, at what time of the day do we send promotions, UI/UX testing. It's going to be everywhere, but it's going to be small from everywhere and it will grow big.

Martin Warner:Yeah, and it's an always-on 24/7 approach. Laura, what about the Alliance?

Golan Shakéd:What—

Martin Warner:how are you anticipating and working towards additional use of AI?

Laura Abraham:Yeah, so like I said earlier, we're really focused in the service recovery area. But in addition to that, to kind of take it back a little bit to the earlier question, we see this huge shift, not only, you know, with our airlines, but in the industry towards modern retailing. And what that's doing, of course, is flipping everything on its head. So it's not not an industry focused on the airline anymore. It's an industry focused on the customer and the consumer. And so any of our AI efforts at the Alliance will be with this in mind. We will promote and multiply anything happening at our airlines, and all of our airlines are becoming hyper focused on customer-centric digital strategies. And so that's, that's what we will focus on with the first start in the service recovery area, because we know that when things are going well, uh, on a travel day, everything's awesome. And when they aren't, we need some help. So yeah, that's what we're doing.

Martin Warner:Yeah, interrupted operations remains one of our one of our biggest nightmares in the industry.

Laura Abraham:Absolutely, right? And there's a lot of regulation around that. So we're just getting our house in order, honestly, at the Alliance. We want to make sure, you know, these tools just sit on datasets, right? It's really a very basic premise at the end of the day. And so as an industry, we loved to give our data away for decades. And so for me, this is an opportunity to gain control of our own data, build something pulling it all together, and then put these tools on top. And so that's really our focus.

Martin Warner:And can you facilitate best practice sharing within the Alliance to a practical level?

Laura Abraham:Yeah, yeah, absolutely. We could be better at that. I'm trying to push that in our alliance. I've been pretty successful so far, but I need to go even further.

Martin Warner:Yeah, very good. Colin, I'm going to come back to you, but on this, on, on this topic of, of AI, how are you guys at Kiwi using that to develop ideation opportunities? So across your Personas of travelers across the different social media areas. How are you using AI to build inspiration, build, you know, opportunity around destinations or around, you know, the virtual interlining?

Golan Shakéd:Yeah, so as a tech company, it is in our DNA to get AI everywhere in the company. Sometimes it takes a little time for us non-techies to catch up with our tech teams.

Martin Warner:I know the challenges.

Golan Shakéd:I send an email to my CEO and I'm not sure who's answering, whether it's him or his AI bot. But we have embraced AI across the organization— legal, tech, marketing. What we're trying to do is to now take this technology into several areas. So, Ronald, you mentioned voice of customer. We definitely listen to our customers across all platforms. Excuse me, in order to try and get a cohesive report that tells us not only what they're saying, but actually what do we need to fix. And AI, as you train it, is able to not only give you data, but actually give you instructions.

Marcel de Nooijer:Mm-hmm.

Golan Shakéd:And we actually use AI to give our developers and our marketeers and anyone who's facing customers instructions rather than just take the data and do something with it. We spoke about social media. I spoke about our 600 ambassadors. AI presents an opportunity for them. Today, any one of those ambassadors is a producer. You know, they can create videos, they can write scripts in any language, they can communicate with anyone all over the world instantly in real time. And so what we're trying to do is provide tools for people to do their jobs better. And, and we can see how the quality of work has, has changed dramatically once you do it. So we try to embrace it by giving our teams, first of all, not only data but instructions— supervised, of course, but coming from machines— giving our ambassadors and our outlets the tools to be more effective. And you spoke about inspiration. I think inspiration comes from the work that these ambassadors are doing by spreading the word, and then comes the network effects. Social media is a visual media, it's a shareable media, and I can give you an example. We saw that, I think everyone saw it last year, Albania all of a sudden became a massive tourist destination. And I think social media played a very, very big part in it because you've got beautiful beaches, very Instagrammable, you've got the right economics for our target audience, and then you've got a couple of airlines starting some routes into it.

Sonja Arnórsdóttir:Yeah.

Golan Shakéd:And all of a sudden you've got a destination I don't think anyone sat there, scratched their head and said, we're going to go Albania. I think social media played a very, very big part in it. And that's the way I think it works.

Martin Warner:Yeah, it is a combination. Again, from my own personal experience, I'm doing a fairly complicated, deliberately complex tour of Australia at the end of the year, starting in Sydney, finishing in Perth for the first cricket test. And that's a combination of planes, trains, and automobiles that takes me literally on a tour for a month across the country.

Roland Jaggi:Yep.

Martin Warner:I'm not sure I could have planned the itinerary without AI that gave me the kickoff, which is probably 80% of my itinerary. But I realised that that was responding to— it was as good as the questions I was asking, and it was creating a general persona response for me that I then had to invest my own time to then customize to my, to my individual needs. But I thought it was a fantastic—

Marcel de Nooijer:In a way, you are creating your own personalization in terms of how do you engage your art.

Martin Warner:And then by the time I've finished with it, it's there and available for somebody else to pick up and use or to adapt for their own personal use.

Marcel de Nooijer:So that also implies that you do it, others will do it, everyone will do it. So this is also for us, and we all share from each other's work, how to deal with that in a proper way to provide and to continue to provide value towards customer engagement and experience. Still, I also think, I think that has been discussed also yesterday, the challenges of the industry, the rising cost. How do you maintain your cost leadership in a certain way? AI also has a great opportunity to work on the productivity side within companies and to maintain that cost leadership so you can offer your services at a reasonable price.

Martin Warner:Absolutely. Roland, and then we'll come to you, Schaeffer.

Roland Jaggi:AI also has a cost. Just think about what you just were talking about. This is going to create massive polling. And if we go to dynamic Pricing and all that stuff, it's not cash anymore. It's going to be a poll every time, and it's going to be expensive. And some blue and other coloured vendors are going to make big money out of that.

Martin Warner:Yeah, yeah, there are definitely implications of it, some of which we can predict, but some that we— some we can't even predict because we don't know what those implications are going to be yet. So it'll be Fascinating to see how they emerge, how quickly they emerge, but then how we all react to what we see.

Mustafa Egilmezbilek:Yeah, it is evolving very fast, right? I mean, even we don't even know how to adapt to that one because tomorrow we might, you know, come up with another more advanced technology and adapting to a new one will take some more time. In terms of the usages at the airlines, I mean, I keep talking about the agents, sorry about that one, but after you engage with the agents and the more it is used, I guess hyper-personalization maybe will be achieved through the technology. Other than that, it's already stated that productivity increases, right? I mean, the tool is really good about creating content, creating summaries, sentiment analysis. They're out of the box, they are available.

Martin Warner:Mm-hmm.

Mustafa Egilmezbilek:So you don't need to spend any development or time on it. For example, for our customer services, we are getting the emails and classification is very, very easy, whether it is positive or negative.

Laura Abraham:It is—

Mustafa Egilmezbilek:it's very successful on that one. One more thing I want to add here, the tool is also very good, and I guess in that area development is also very fast, which is programming. It's called vibe programming, which means you program via prompting. All right, and this helps also to your innovation, you know, studies efforts, because now you can create the prototypes in 2 hours, in 3 hours, and so on. If you have any idea, implementing it, building the MVP, minimal viable product, It's not a hassle anymore. So you can start one, if it doesn't work, you can get another one. Yes, like at the same time that before you were developing the MVP, now you can try out, you know, maybe 10 or 12, 15 prototypes. It is accelerating the whole process.

Martin Warner:It supports the process of fail fast or learn fast and adapt.

Mustafa Egilmezbilek:Exactly.

Martin Warner:Yeah, yeah. Anybody else have a view on that before we move on to another topic? Now, killer apps. Um, has anybody seen a killer app using AI in aviation, and what would it be in terms of what you, what you might have, have seen?

Laura Abraham:If you, if you ask me that in 12 months, I'm going to say it's the one I built. For the Alliance. But you're asking it today. So no, no, but I would say I haven't seen one, but I would say that the 2 areas are the areas that I've mentioned a few times today, and I think those are the spaces we should keep our eyes on. One would be around service recovery, and I think the other one is using it for voice of the customer, because I think Both those areas are going to help transform what I think we need to do in aviation, which is to turn everything towards the customer, center everything on the customer. And the service recovery piece, of course, also assists the employees, which is the other side of the customer experience coin. So I think—

Martin Warner:Around the employee satisfaction and employee engagement and retention.

Sonja Arnórsdóttir:Exactly. Yeah.

Laura Abraham:So I think in the next year, those are the 2 spaces that we're going to see some great apps coming out of. But, you know, keep your eye on SkyTeam.

Martin Warner:Will do, Laura.

Mustafa Egilmezbilek:Thank you.

Martin Warner:Anybody else have a view on killer apps before we— we've got a few questions that have just appeared, so we'll try and cover those off. I'm not going to take these necessarily in top-down order. Let me take the Question of— I think there was a question which has just disappeared around at what point do you tell the customer that they're interacting on an AI basis rather than a real human basis. I'm sure I saw that question, but it seems to have disappeared. Mustafa, taking that one, do you want to—

Roland Jaggi:It's back.

Mustafa Egilmezbilek:It's back.

Martin Warner:There we go. Until when will airlines inform customers whether they're interacting with AI or a real human?

Mustafa Egilmezbilek:Maybe never.

Roland Jaggi:Because—

Martin Warner:Is it important?

Mustafa Egilmezbilek:It is important, right? Because that technology, based on a probabilistic model, which means every time you ask a question, you are getting a different answer. So that means And it has some issues like hallucinations and so on. I'm talking about the generative AI here. So in order to cope with the mistakes that the technology produces, I guess upfront, we should inform the user that we are using this technology to help them, right? It might work 95-98% of the time, but for the remaining 2%, we should inform them, I believe, as soon as possible.

Martin Warner:Yeah. And how do you get from a bot or from AI/ML into then interacting with a real agent. I still feel that we've got a big gap in that, in that process. Not everywhere, Golan. I know you're going to correct me, but I've had a couple of experiences this week where it didn't.

Golan Shakéd:I expect Kiwi customers a few, maybe a year from now, to actually demand an AI over a human. Based on the CSET, that's what I'm seeing.

Martin Warner:And on the, and on the quality of what the AI is doing.

Golan Shakéd:Yeah.

Sonja Arnórsdóttir:Yeah.

Laura Abraham:You know, I, I believe that we should be transparent first and foremost, always. This is about their experience. This is about customers. And oftentimes they feel out of control when they're traveling anyway. And so this would be a way to help you know, control emotions and keep people just feeling confident in who they choose. So I believe in transparency.

Martin Warner:Next question was around yield management. So inventory and yield management. Do you think that this job is on the list now of those to be replaced by AI? And I'd probably add to that or supplement.

Marcel de Nooijer:I think we touched base on that. Let's say it's already in use in revenue optimization. Let's say, with the developments on dynamic pricing and in a way also that helps to personalize offers to create value. So will it be on the short run without any revenue analysts? I don't think so. So it's about learning and adapting and transition. So perhaps 10 years from now, who knows? But yeah, it's already there to be used. And to be honest, we just moved a couple of months back to a new revenue management system, Kaysen, which is really into bringing this, this forward. And I had some initial worries on the transition. How will it interact with, with, with the team? Outstanding results. The speed of adaptation, mind-boggling, I would say. So I've become actually a little bit more enthusiastic as a dinosaur.

Sonja Arnórsdóttir:I would definitely agree. This will not replace the analyst. We already integrated some of the AI automation. So it's in my mind just better use of the manpower. They are using their strength instead of half of the day just gathering some manual data to analyse and then the day is over. So it's It's just better use of the manpower.

Marcel de Nooijer:But it improves the overall quality at the end of the combination.

Sonja Arnórsdóttir:And it's also so quick and violent. We are able to act more quickly.

Martin Warner:I think we have a couple of questions that overlap on the topics that we've just been talking about. Let me take the top one of what's left on the board. How do you believe AI will affect the frontliners of the company? And that's across All, really all of the different employee types. How well has that been thought through around the, around the implications of AI on the legacy employee base? Mustafa, any, any thoughts on that?

Mustafa Egilmezbilek:Yeah, again, the tools are very good at increasing the productivity. So one of the purposes of the technology is augmenting the people. So, their access to data will be very fast with the technology. I mean, the whole information, corporate information or anything, will be very accessible, which means in addition to developing some decision-making, let's say, tools, support tools with ML, with the generative AI, they will be able to access all the data quickly, which I think that increases their decision-making time. So it will be a positive impact on them.

Martin Warner:That's an interesting perspective.

Golan Shakéd:Thank you.

Martin Warner:Let me go to Laura. Any further thoughts on that?

Laura Abraham:No, I just think that it's going to open the door for our employees to have a better view into the other parts of the business that they may not have seen before. And I think it also opens the door to Helping them understand customers' frustrations, customers' journeys, all those things. So I really view it as, you know, I'm a really optimistic person.

Martin Warner:Clearly, you're really pro this.

Laura Abraham:I view it as a very transformative, like, window into areas that they may not have seen before.

Martin Warner:And the final Final question, and this is not trying to end on a negative, but we're just going to get close to time. How do you mitigate the potential reputational risk to the business of AI if it's perhaps not used in an optimal way? Marcel?

Marcel de Nooijer:I don't think it differs that much. I think you should work always on your reputation, and if this is a driver, to push up quality. We're all about transparency in our business.

Martin Warner:My personal view is, again, being optimistic, is that this is actually, if it's used correctly, it's reputational enhancement.

Laura Abraham:Yeah, absolutely.

Sonja Arnórsdóttir:And have a clear strategy in the company and everybody knows what it is, so it's not used in the wrong way. And data safety and everything in mind.

Marcel de Nooijer:Use it as a motivator.

Golan Shakéd:Yeah.

Martin Warner:Yeah, very good.

Mustafa Egilmezbilek:I agree.

Martin Warner:There we go, right on cue. We've managed that to zero. Folks in the room, I hope you found the combination approach to these sessions on social media and AI to be as fascinating as I have, but we couldn't have done this without a very flexible but a very knowledgeable panel. So my thanks to all of the panelists that agreed right at the last minute that we would, that we would take this approach. So please join me in thanking the panel.

Sonja Arnórsdóttir:Thank you.

Copyright policy: All transcripts on this site are the copyright of CAPA - Centre for Aviation. Our reproduction policy is as follows: you may quote up to 400 words of any transcript on the condition that you attribute the transcript to CAPA - Centre for Aviation and link to the original video page. All other use is prohibited. While we aim for 100% accuracy in the transcript, there may be some minor transcribing errors.