A view on cloud technology for Modern Airline Retailing
21m 58s
The discussion between Boris Padawan and Cindy Falchlena focuses on the infrastructure and strategic shifts required for modern airline retailing, emphasizing cloud technology and AI. Cindy, from Google Cloud Consulting, highlights two major trends: the move from structured to unstructured data and the transition from applications to agentic platforms. She argues that airlines can adopt these changes without overhauling legacy back-end systems by first enabling front-end capabilities to understand natural-language queries (e.g., "I need a hotel near Times Square") and then setting up agentic distribution protocols. Key challenges include mindset barriers, risk aversion, and the risk of "reverse direct distribution" if OTAs better serve agentic queries, potentially eroding airlines' direct sales. Cindy notes that change cycles are accelerating—AI mode reached 2 billion users in one year—and urges airlines to focus on "moving fast in the in-between" rather than waiting for infrastructure overhauls. Examples include Alaska Airlines' natural language shopping and Google's agentic retailing for hotels via Universal Commerce Protocol. The conversation underscores that airlines must act now to remain competitive in a rapidly evolving digital landscape.
Introducing Cloud Technology for Modern Airline Retailing
Hello, and welcome to the June Tim cast and the next episode in our 2026 series.
I'm Boris Padawan and Chief Commercial and Marketing Officer at Travel in Motion.
Today, I'm joined by Cindy Falchlena.
Cindy is director at Google Cloud Consulting.
Cindy, thank you so much for joining us.
Thank you, Boris.
It's great to be here.
Cindy, why don't you just introduce yourself and share a bit of your background and what you're currently doing?
Absolutely.
So, Boris.
Speaker 2
We've known each other for what, 20 plus years now?
Speaker 1
Absolutely.
Speaker 2
Going back to.
Speaker 1
Our EDS.
Speaker 2
Days if we're getting in the Wayback Machine I've been at.
Speaker 1
Google about four.
Speaker 2
Years now and I joined.
Speaker 1
Google as.
Speaker 2
A travel and transportation expert and during my time.
Speaker 1
Here I have.
Speaker 2
Kind of moved in advanced and now I run the North America regions for go to market as well as our incubating industry.
Speaker 1
Practise of which?
Speaker 2
Travel is 1 and some of the others are logistics, power and energy, manufacturing, construction.
Speaker 1
Biotech Auto.
Speaker 2
Those sorts of things.
So I'm definitely expanding my industry footprint here at Google.
Speaker 1
It's great to have you here as one of the representatives of one of the three big cloud providers, if not even the biggest cloud provider.
I don't know.
But of course, if you go through the aviation industry, if you go through the airline industry, cloud is everywhere and Google is everywhere.
So you have quite a footprint here in the area.
Framing the discussion we will have now for the next about 15 minutes are the infrastructure needs airlines have when transitioning to modern airline retailing.
When Cindy and I met for the first time and it's, it's unbelievable but true, it's about 20 years ago, we were working for a company that was providing reservation systems in a very legacy classical way based on TPF infrastructures, highly transactional based, very centric, very often outsourced from an operations perspective to one of the big outsourcing providers.
And it's interesting to say that even 20 plus years later, we still see kind of similar architectures within our industry, but the world around it has changed.
And this is also one of the motivations by airlines are moving to modern airline retailing.
And we don't need to go further into modern airline retailing here right now because our earlier blocks, our earlier Tim cars have touched on this, of course, quite extensively.
Nevertheless, whenever we talk about modern airline retailing, let's say from a team perspective, we talk about the commercial processes.
We talk about the ability of airlines to move, to offer order, to create offers to passengers at every point in time.
We talk about customer segmentation, personalization, We talk about willingness to pay and so on and so forth.
But it is a commercial view and the projects we do are very often driven by commercial functions.
We work with commercial roles, but we shall not underestimate the implications this has for IT and here the implication this has for the infrastructure.
My colleague, also Cindy's ex colleague if I'm not mistaken, next Dot, wrote a block about the infrastructure challenges for modern airline retailing exactly coming from this view.
And I can only recommend all of you to take a look at the block because the move to modern airline retailing also implies a very big shift in thinking about how infrastructure is handled.
We are talking about modularity, about API's, about providing customer facing retailing experiences, we talk about scalability and so on and so forth.
Google's Vision: The Shift to Unstructured Data and Agents
Cindy, from your perspective, and I don't want to corner Google in the infrastructure area only because I know you do much more at Google.
But starting maybe with the infrastructure view, what do you see as the biggest challenges our industry has when moving to modern airline retailing?
So.
Speaker 2
I'll combine the Google.
Speaker 1
Aspects of it.
Speaker 2
Both from an infrastructure perspective and being Google.
One of the benefits of being at Google is.
Speaker 1
The data.
Speaker 2
That we have and understanding consumer and human behaviour because.
Speaker 1
They're all Google.
Speaker 2
Plays and market and so we sit down with the search teams on a monthly basis and we.
Speaker 1
Look at what are the.
Speaker 2
Patterns of consumer and retail.
Speaker 1
Behaviour.
Speaker 2
And human behaviour and and.
Speaker 1
Retail behaviour is.
Speaker 2
Everything from I'm buying a shirt to I'm taking a trip urging insurance to.
Speaker 1
Whatever it happens to be.
Speaker 2
And we've seen this fundamental shift.
Speaker 1
Over the last 12 I'll say.
Speaker 2
14 months since April of last.
Speaker 1
Year.
Speaker 2
Where we've moved away as an industry, and I'm talking technology as a whole, so stepping back into infrastructure, we've moved away from the construct of I'll say.
Speaker 1
Structured data.
Speaker 2
Point and click Filling out fields in a.
Speaker 1
Database to this.
Speaker 2
Construct of unstructured data.
Speaker 1
The idea that.
Speaker 2
In in Google's words for that when we think about search would be moving from keyword.
Speaker 1
Search, which is where we were.
Speaker 2
18 months ago to moving to fully unstructured.
Speaker 1
Search so the idea that as a traveller I could.
Speaker 2
Go out on Google.
This is very unique to Google right now and I could say, hey, I have a meeting in Times Square.
I need to be.
Speaker 1
There by two.
Speaker 2
O'clock in the afternoon and I want.
Speaker 1
To stay.
Speaker 2
No more than a 10 minute walk from my meeting.
And I want it to be.
Speaker 1
A hotel that has a.
Speaker 2
Good.
Speaker 1
Gym has a.
Speaker 2
Restaurant and is.
Speaker 1
Quiet that.
Speaker 2
Being translated into JFK, LaGuardia, Newark.
That being translated into who?
Speaker 1
Flies where for.
Speaker 2
When at what?
Speaker 1
Price.
Speaker 2
That's a very different world than going out to an airline website.
And I live in Houston, so that would be and I fly a certain carrier because.
Speaker 1
Of where I live that would be IA GWR.
Speaker 2
For me and this kind of, you know, point and click type process, that's not where we're at from a technology perspective.
And so when I think about what?
Speaker 1
Travel suppliers should.
Speaker 2
Be thinking about right now.
Speaker 1
It is the concept.
Speaker 2
Of unstructured.
Speaker 1
Data so the idea that.
Speaker 2
In my traditional world.
Speaker 1
Whether it's TPF?
Speaker 2
Or mid range or even traditional cloud.
Speaker 1
That I have this structure database.
Speaker 2
Construct that is not where decision making is happening right now and travel suppliers thinking about OK, how am I going to create this intersection of?
Speaker 1
Structured.
Speaker 2
Data.
My traditional world.
Speaker 1
Whether it's the.
Speaker 2
Traditional world of PNR or offer and order?
Because offer and order structured data as well.
How am I going to move into the unstructured?
Speaker 1
Data space, where the context of travel lives.
Speaker 2
Understanding the context and being able to use that unstructured context to drive conversion yield.
Speaker 1
Share of wallet attach rate.
Speaker 2
Like all of those things so unstructured.
Speaker 1
Data is the.
Speaker 2
The big #1, the big #2 or actually maybe the number one is.
Speaker 1
We are.
Speaker 2
Exiting the era of applications and entering the.
Speaker 1
Era of agents.
Speaker 2
And.
Speaker 1
Regardless of what?
Speaker 2
Your back end infrastructure.
Speaker 1
Stack looks like.
Speaker 2
Understanding how agentic is going to impact.
Speaker 1
Your ability to distribute is is very.
Speaker 2
Front and centre right now And are you prepared for Agentic?
Speaker 1
Do you have all of the?
Speaker 2
Platforms in place, the security in place, the protocols in place, the marketing in place, all of that stuff that goes along with an agentic.
Preparing Airlines for Agentic Distribution Without Backend Overhaul
Platform, I'll take a deep breath and let you react to that.
I think as Google you're of course one of the front runners on all the subjects we are talking about.
But talking about the aviation industry, which is always since the times of TPF when they were actually very far ahead of all the other all of lots of other industries talking about airlines, I think this all sounds good to airlines, but they are not there yet.
They are still struggling in cutting E tickets, in doing code share in a traditional way.
They are still sending teletext messages in form of type B messages.
How would Google support the first step to a more modern platform of retailing?
And I'm very happy that you started your initial answer with returning with the move of airlines to become retailers before airlines will even be equipped to follow the paradigm or to execute on what you described and what we all believe in actually because this will be the future.
But what is the intermediate step there for airlines to get ready, but still to support the things I mentioned earlier as and as well also to support a legacy integration because over all the airline industry will not move in form of a Big Bang.
There will be front runners.
There will be airlines which will still stick to their traditional architectures and setups.
Speaker 2
So that is a really interesting question and I will be provocative and I will.
Speaker 1
Say that.
Speaker 2
We shouldn't assume that we have to move our back end infrastructure.
Speaker 1
To be able to adopt.
Speaker 2
Front end practises.
What I mean by that is generally when we coach in this space and I'll go to a couple of announcements and a risk.
Speaker 1
If you are a travel supplier right now, so Google has.
Speaker 2
2 main conferences.
Speaker 1
We have Google.
Speaker 2
Cloud Next, which generally happens in the April time frame where we announced our technology advancements like Gemini Enterprise Agent platform, things like that.
And then we have Google IO, which is very focused on our consumer announcements and at.
Speaker 1
Google IO.
Speaker 2
A couple of weeks ago in May, we announced Agentic.
Speaker 1
Retailing for hotels.
Speaker 2
And so the.
Speaker 1
Idea.
Speaker 2
That right now, what's happening, people?
Speaker 1
Go to Google to.
Speaker 2
Discover and decide about travel and we have.
Speaker 1
The statistics to back.
Speaker 2
That up in terms of where?
Speaker 1
People are.
Speaker 2
Discovering what they're going to do and the decisions that they're making.
What happened with that announcement is now through Agentic and.
Speaker 1
Through UCP.
Speaker 2
Or universal commerce.
Speaker 1
Protocol.
Speaker 2
We've opened the door.
Speaker 1
For travel suppliers, whether it's the OTA or the hotel to.
Speaker 2
Conduct bookings and the Google consumer upper funnel.
It's starting with hotel.
Speaker 1
But the OTAs?
Speaker 2
Are playing and.
Speaker 1
Guess who the.
Speaker 2
OTAs also conduct traffic.
Speaker 1
For that would.
Speaker 2
Be the airlines.
So if you're an airline and you're thinking.
Speaker 1
That the website.
Speaker 2
Or the mobile app is the centre of the universe.
Speaker 1
I would argue.
Speaker 2
To pay attention to what's happening with hotels right now, so starting there.
Speaker 1
There is pressure that is rapidly, rapidly, rapidly.
Speaker 2
Evolving here and if we think that we have time.
Speaker 1
I would.
Speaker 2
Argue when you look at the change cycles.
Speaker 1
A little stat is that it took.
Speaker 2
16 years for mobile phones to release to reach 100 million users.
Speaker 1
It took.
Speaker 2
AI mode with Google which was launched.
Speaker 1
Last April at IO.
Speaker 2
One year.
Speaker 1
To reach 2.
Speaker 2
Billion monthly users, so the change cycle is phenomenally accelerating.
So going back to like thinking about as an airline, how would you?
Speaker 1
Prepare for.
Speaker 2
What is going on or how would you react or get in front of what is going on?
I think the first question really centres around are you?
Speaker 1
Able to speak.
Speaker 2
The customer's language, and what I mean by that is can you understand unstructured data?
Can somebody come to your?
Speaker 1
Website somebody.
Speaker 2
Come to your mobile app and instead of pointing and clicking, can they type in?
Hey, my 30th anniversary is in November.
I want to go someplace warm where you can fish.
Very simple example, but it's true in business travel as well.
I've got that meeting in Times Square at 2:00.
Speaker 1
O'clock.
Speaker 2
And that doesn't say that you have to change your back end systems.
Speaker 1
It just says.
Speaker 2
That you have to understand to be able to translate that into what your back end systems would require.
Speaker 1
So the idea that I say.
Speaker 2
Hey, I've got a meeting in Times Square and that can be translated into IA.
Speaker 1
HEWRIHIAHLGA that.
Speaker 2
Sort of thing that generally, you know, when we coach in this space.
Speaker 1
We have two.
Speaker 2
Steps one that's.
That is one step one, which is can you understand unstructured data?
No, by the way.
Speaker 1
That might be.
Speaker 2
Video it might.
Speaker 1
Be a photo.
Hey, I saw this beautiful.
Speaker 2
Beach.
I want to go there.
Speaker 1
Or.
Speaker 2
I saw this monument.
I want to go there.
Speaker 1
Or this is the office I have to?
Speaker 2
Visit Tell me how to get there.
Speaker 1
You know we.
Speaker 2
Look at that from a step one perspective and starting to understand when we look at step.
Speaker 1
Two, we say capture that data like actually.
Speaker 2
Do something with it, learn from it, and step #3 is.
Speaker 1
Stitch it together with your structured.
Speaker 2
Data from the history perspective and be able to respond in context, but going to the other side of the fence and looking at what's going on from an agentic.
Speaker 1
Perspective.
Speaker 2
Are you taking the steps now to set up your agentic platform again?
That has nothing to do.
Speaker 1
With what your?
Speaker 2
Back end systems look like and everything to do.
Speaker 1
With, do you have the ability to distribute or to?
Speaker 2
To handle agentic distribution, What is your crawl, walk, run strategy for that?
The last thing that I'll.
Speaker 1
Say is that?
Speaker 2
When you think.
Speaker 1
About AI Geospatial.
Speaker 2
Grounding.
Speaker 1
Is a really really.
Speaker 2
Important thing in travel.
Speaker 1
I think this sounds fantastic.
I really see the vision you have, the vision Google has here.
I'm still a bit conservative, stuck in the airline corner where change takes tremendously long as we have experienced while fully sharing the paradigm you are very well describing here.
Addressing Mindset and Distribution Risks for Airlines
From the experiences you have had with Google, you have had with working with airlines over the past 20 years, where do you think are the biggest hurdles?
Is it a technical hurdle?
Is it a hurdle in in mindset?
Is it being risk or worse as an airline per se or how would you summarise it?
Speaker 2
Any and all of.
Speaker 1
Those I'd say the risk.
Speaker 2
For the airlines, you know, I think about the last 10 years and in particular in the airline industry and it's been about.
Speaker 1
Direct distribution, right?
Speaker 2
What are the things that I have to do to take control of my distribution and be in charge of making a good decision with the information that I have at that point in time in this scenario?
Speaker 1
Where the OTAs?
Speaker 2
Are playing in the agentic distribution world what we're seeing in human and consumer behaviours?
Speaker 1
You go where it's easy, you go where you're understood, you go where you get the.
Speaker 2
Exact answer.
Speaker 1
To the exact.
Speaker 2
Question that you're asking is and for the airlines, I think the risk is, is.
Speaker 1
That if the.
Speaker 2
OTA can be asked and can the answer.
Speaker 1
The exact.
Speaker 2
Question and you as the airline can't are we going to see call it?
Speaker 1
Reverse direct.
Speaker 2
Distribution, are we going to go back to an indirect distribution model That to me is the risk.
The second thing that we see overall in what's going on is we one of the things that we track very.
Speaker 1
Closely is.
Speaker 2
Generational behaviour and what we're seeing in the numbers now and we've.
Speaker 1
Validated.
Speaker 2
This and in some of our direct relationships is it's no longer Gen X and Baby Boomers driving travel, Gen X Baby Boomers point and click, it's the generation.
Speaker 1
Of I own a file, I share a file.
I take back that file.
Speaker 2
And it's mine that is not the digital native generation in particular.
It's a completely different way of working with technology and if we continue to revert into that point and click and not recognise.
Speaker 1
Who the actual customer is.
Speaker 2
I think.
Speaker 1
That is also.
Speaker 2
One of the.
Speaker 1
Big risks or big hurdles that we have?
Speaker 2
To work through in terms of the mindset.
Speaker 1
Associated with this.
Speaker 2
We were talking about earlier on the podcast the.
Speaker 1
Idea is is.
Speaker 2
It about the core infrastructure or is it about something else?
Speaker 1
I would argue.
Speaker 2
If we make it about the core infrastructure, we're still going to be having this conversation five years from now if we make it about how we move fast in the in between.
Speaker 1
That is.
Speaker 2
Where the magic?
Speaker 1
Happens.
Speaker 2
And there are certain airlines.
Speaker 1
Globally that have.
Speaker 2
Embraced that methodology and.
Speaker 1
We have a.
Speaker 2
Texas saying which is make hay while the sun shines There are airlines out there that are making hay while the sun shines.
I mentioned before you know this change cycle the.
Speaker 1
Idea that.
Speaker 2
You know, we think we we tend to look at the.
Speaker 1
AI.
Speaker 2
Application.
So the Gemini's the.
Speaker 1
Crocs The.
Speaker 2
Crocs, the clods, you know that sort of thing and watch market share but.
Speaker 1
If you look.
Speaker 2
At What's going on?
Speaker 1
Actually the largest.
Speaker 2
Kind of.
Speaker 1
AI use globally.
Speaker 2
Is AI mode, it's not the applications now that will those levels change over time.
Speaker 1
Absolutely they will.
Speaker 2
Will it change month over month?
Speaker 1
Absolutely it will.
Speaker 2
But as an airline?
Speaker 1
As a hotel as.
Speaker 2
A car rental company as an OTA do.
Speaker 1
You have a strategy that allows you to play where?
Speaker 2
Decisions are being made and now where commerce or distribution is emerging.
Speaker 1
And I think it's interesting that's not as a surprise to our conversation about kind of a very conservative infrastructure view, infrastructure discussion, classical cloud is turning into an AI discussion.
But the change will happen and the change has started.
We see this in our daily lives already.
Do you have some examples for airlines how Google may be proof of concepts or things in production where airlines make what you would say very good use of AI?
Real-World AI Examples and Critical Next Steps for Airlines
Yeah, absolutely.
So I I would say that the.
Speaker 2
Front runner I'll I'll name 1 carrier and then I'll use a couple of examples where I can't name carriers.
Speaker 1
So Alaska.
Speaker 2
Airlines came out about 18 months ago with natural language shopping.
Speaker 1
So the idea that you could go to the.
Speaker 2
Alaska Airlines site and type in I want to go.
Speaker 1
To a.
Speaker 2
Beach with bioluminescence or I want to go?
Speaker 1
To a city with great food or.
Speaker 2
Or whatever it happens to.
Speaker 1
Be to Charlie.
Speaker 2
Jane and the Shop.
Speaker 1
Up at Alaska.
Speaker 2
Are doing some really, really interesting things.
Speaker 1
I'll give you a couple use cases.
Speaker 2
Without naming carriers because it's considered.
Speaker 1
Secret sauce.
Speaker 2
Another use.
Speaker 1
Case.
Speaker 2
With one of our partner carriers involved using.
Speaker 1
AI with distressed.
Speaker 2
Inventory, so the.
Speaker 1
Idea that.
Speaker 2
You know when you go and you do a standard shopping transaction on a website and you return, you can return a lot of things, but if you look at, say your top 10 logical returns and you've got.
Speaker 1
Two of those flights that are not.
Speaker 2
Healthy from a load.
Speaker 1
Factor perspective.
Speaker 2
We run call it experimentation, where you take those flights and instead of returning them in the list, you return them using Gemini.
Speaker 1
Very delightfully with.
Speaker 2
Photo within with.
Speaker 1
Text that's.
Speaker 2
Appropriate for the input.
Speaker 1
Query with.
Speaker 2
Ads, placement, those sorts of things and what we found.
Speaker 1
Is by.
Speaker 2
Responding in a delightful.
Speaker 1
Way you get.
Speaker 2
Fear of missing out and we have inventory.
Speaker 1
Recovery for those.
Speaker 2
Flights and very successful inventory recovery.
We have another carrier in Europe that we're working with where?
Speaker 1
All we did.
Speaker 2
We changed the interface on their website mobile app to natural language and it doubled the conversion in their direct channels.
Speaker 1
So there are lots of cases out there.
Speaker 2
And we're seeing acceleration like we have never seen before.
Speaker 1
And I think these examples and the vision you are sharing is actually a good wrap up of the discussion we are having today in our Tim caste.
Is it a true statement if I summarise you that the infrastructure requirements, they're they're actually not the hurdles because the infrastructure is around.
It is more about thinking what you do with it and going the next step with all the examples you brought earlier from an AI perspective.
And the challenge for the airlines is after they hopefully all move to successful retailers in a modern airline retailing environment, how they go to the next step in order to follow all those items.
Speaker 2
I'd summarise it in three things.
I'd say that if you're waiting on your infrastructure catching up.
Speaker 1
That's a very, very high risk.
Speaker 2
Because the OTA's are not waiting, the second thing that I would.
Speaker 1
Say is that?
Speaker 2
Agentic is here and having an agentic strategy in using AI.
Speaker 1
As a decoder.
Speaker 2
Ring between the agentic world and.
Speaker 1
The.
Speaker 2
Existing traditional travel.
Speaker 1
World is really critical.
Speaker 2
And not to get too googly on this, geospatial crowding is really really really important.
Speaker 1
So there is.
Speaker 2
Kind of singularity in in answers in that space.
The last thing that I'll.
Speaker 1
Say is that?
Speaker 2
We are entering an era of unstructured data and it doesn't mean you have to blow up in your back end systems, but it does mean that you have to bring together your.
Speaker 1
Structured World.
Speaker 2
With the unstructured.
Speaker 1
World.
Speaker 2
And you have to consider what that means in a commercial ecosystem.
Lots of hops are not a good thing.
Having a back end platform that seamlessly integrates structured and unstructured together is a now imperative, not a future.
Speaker 1
Imperative and I think that's a very good summaries of your view.
Cindy, thank you so much.
It brings us again to the end of this month's Tim cast.
Cindy, thank you for the time and the perspectives and thanks for everyone listening.
If you're thinking that this sounds all simply and also slightly terrifying, welcome to the world of airline retailing.
Welcome to the world of airline IT.
You can find out more about us about Travel in Motion on our website travelinmotion dot chapter and connect with us on LinkedIn.
The links to our profiles are in the show notes.
If you haven't already subscribed to our Timcast, please do so on Spotify.
Thank you so much, Cindy.
Again, a big thank you to you for your visionary talk here.
It's always great catching up with you, talking with you about the developments in our industry.
Thank you so much, Cindy.
Absolutely.
It's been.
Speaker 2
A lot of fun.
Speaker 1
Thank you and bye bye.
Podcast Summary
Key Points:
Airlines must shift from structured data (e.g., PNRs, databases) to unstructured data to understand customer context and drive conversion.
The industry is moving from application-based systems to agentic platforms, where AI agents handle complex, natural-language queries (e.g., "I need a hotel near Times Square").
Airlines can adopt modern retailing without overhauling legacy back-end infrastructure by focusing on front-end capabilities like understanding unstructured data and setting up agentic distribution.
Key hurdles include mindset (resistance to change), risk aversion, and the threat of "reverse direct distribution" if OTAs better serve agentic queries.
Google's AI mode reached 2 billion users in one year, highlighting the accelerating pace of change that airlines must address.
Real-world examples
Summary:
The discussion between Boris Padawan and Cindy Falchlena focuses on the infrastructure and strategic shifts required for modern airline retailing, emphasizing cloud technology and AI. Cindy, from Google Cloud Consulting, highlights two major trends: the move from structured to unstructured data and the transition from applications to agentic platforms. , "I need a hotel near Times Square") and then setting up agentic distribution protocols.
Key challenges include mindset barriers, risk aversion, and the risk of "reverse direct distribution" if OTAs better serve agentic queries, potentially eroding airlines' direct sales. Cindy notes that change cycles are accelerating—AI mode reached 2 billion users in one year—and urges airlines to focus on "moving fast in the in-between" rather than waiting for infrastructure overhauls. Examples include Alaska Airlines' natural language shopping and Google's agentic retailing for hotels via Universal Commerce Protocol.
The conversation underscores that airlines must act now to remain competitive in a rapidly evolving digital landscape.
FAQs
Agentic distribution uses AI agents to interpret unstructured customer requests (like natural language queries) and complete bookings, shifting from traditional point-and-click interfaces. It moves control from airline websites to wherever the customer asks the question, such as search engines or OTAs.
Airlines can focus on front-end capabilities like understanding unstructured data (e.g., natural language shopping) and setting up an agentic platform for distribution, security, and marketing. This allows them to translate customer inputs into structured data their legacy systems can process, enabling gradual adoption.
If airlines fail to offer seamless, context-aware booking experiences, OTAs may dominate by providing those experiences, leading to a loss of direct customer relationships and revenue. This could reverse the trend toward direct distribution.
Unstructured data comes from natural language queries (e.g., 'I need a hotel near Times Square with a gym') rather than structured fields in a database. Airlines must process this context-rich input to understand customer intent and drive conversions, even if their backend uses structured data.
Start by enabling natural language shopping on your website or app, as Alaska Airlines did 18 months ago. Then capture and learn from that unstructured data, and later stitch it with structured data for personalized responses. This avoids waiting for backend changes.
Younger generations (digital natives) prefer unstructured, agent-driven interactions over point-and-click interfaces. Airlines that cater only to older demographics risk losing relevance as these younger travelers become the majority.
Chat with AI
Loading...
Pro features
Go deeper with this episode
Unlock creator-grade tools that turn any transcript into show notes and subtitle files.