• Ryan Haynes

AI in hotel and hospitality retail

Looking at AI in hospitality, a hot topic as tech leaders look to optimise performance of systems and get an understanding of the value that AI provides to the tech stack. Joined by Mike Webster, CEO and founder of Arvoia, hospitality's leading AI platform, he talks us through where AI is being deployed and how far we are yet to push it to gain real value.

We’ve seen AI deployed in pricing and forecasting, AI has been applied to identify recognition systems and within guest communication, but what about in retail of hotels and hospitality? With Mike we cover:

  • What AI is and what it is not

  • How AI elevates marketing, sales and operations

  • Where AI is currently being deployed well in hospitality

  • How AI can take hotel ecommerce to the next level

  • Mitigating the lack of staff, talent and skills to utilise AI

Discover more of Arvoia: https://www.arvoia.com/

Check out our other episodes: https://www.haynesmarcoms.agency/travel-market-life-podcast



Program Notes


Ryan Haynes:

Hello and welcome to Travel Market Life. I'm your host, Ryan Haynes. And today we're Looking at AI in hospitality is a hot topic. As tech leaders look to optimise performance of their systems and get an understanding of the value that AI provides to the tech stack. We've seen AI deployed in pricing and forecasting. AI has been applied to identify recognition systems and within guest communication, but let's have a deeper look at the real value of AI to hotels


Ryan Haynes:

Joining me is Mike Webster CEO and founder of Arvoia the hospitalities leading AI platform. Thanks ever so much for joining me today, Mike. So AI is such a hot topic. Can you actually define though, what is AI and what is not AI?


Mike Webster:

Good question. Good to be here. Thanks Ryan. So simply post AI is the ability for a machine to, to think and to make decisions. There's lots of different areas within AI. I suppose everything from, you know, speech, it's almost like the easiest way to consider it is I think of a human and the human features and functions and how machines replicate that. So speech, for example, would be speech recognition, read and rifle would be natural. Language processing movements would be robotics. Seeing is like, is the area of computer vision. And then pattern recognition is, is, you know, kind of the area of machine learning, but fundamentally it's the ability for a machine to think and to make a decision.


Mike Webster:

Ultimately, you know, at a broad level,


Ryan Haynes:

I mean, there's been so much utilization or the word AI, artificial intelligence machine learning, and the perhaps the way that we've got to understand the way it works when we know the Ammons Amazons, Alexis of this world and how they recognize what you're saying. We also see how chatbots are working in a way to really facilitate better engagement online and, and through the, through the website without actually using human being as customer service. But is that enough for us to really understand what the power of AI is?


Mike Webster:

No. And, and I think, you know, if you look at, particularly, if you look at chatbots, for example, you know, they, they, they provide a certain function and there's an efficiency to us because you don't have a human necessarily on it. But I think your chatbots, you know, for example, if you look at where AI is, well deployed it look at Spotify, for example, right? When's the last time you went on to Spotify, I want to talk to a chat bot to try and figure out what, the minute you go on to Spotify, Spotify, you're being served, you're being understood, your preferences, your needs, state, your mindset, even when you're driving. Sometimes it'll pick that up and it'll serve you what you want. So chatbots, yeah, they are use of AI and they work pretty well. And, but is the ultimate destination, how brands can effectively communicate with their digital clients?


Mike Webster:

I don't think so.


Ryan Haynes:

I mean, it's fascinating as you say, you go on to Spotify, you see all the artists that you've been listening to, you've seen related tune the, in a similar, you go on Netflix or these recommended programs, new films and features that have just come out and it's like, oh, Hey, ho, ho, how are you really understanding me? Why have I got no horror on my list? Netflix understands me really well, but how can it really be applied to travel and hospitality? Because I mean, you know, travel hospitality, it's something you do, you know, two or three times a year, maybe. So as a consumer, what are the ways that we could recognize that we're getting value from a business and actually underneath it, AI is doing the work.


Mike Webster:

Yeah, I think that, I think there's a couple of things there it's on page. I think when, if you consider a guest arriving at a hotel's website, for example, you know, one of the reasons that they tend to find the go, there is an expectation that they're going to get served better. You know, you know, in a way that they're not necessarily getting safe from an OTA or another party. And the reality is, or not usually is that if they're understood, when they arrive at their needs, state their mindset, the product, they want to see the image. They want to see that the price brackets that they're engaged in, and maybe all the additional services that they want us to, if they, if they get that, when they arrive at a hotel's website, they feel understood and served better. And that's what that tend to book when they don't get that, the tend to bounce, you know, they tend to bounce off and go somewhere else.


Mike Webster:

So if you look at how AI can drive hospitality, it is, and it's specifically around the idea. If you've never, if the brand has never seen this person before, the reality is that human behavior is unpredictable, but it's entirely measurable. So if you're a first time entrance onto a digital estate for a hotel brand, for example, there's a lot of other first time entrance, but the indicators that dash, that, that, that machine will spot, but a human won't necessarily spot there's this long trail of, you know, source origin, destination, lead time, device size, you know, in a fully anonymized way. And, you know, your lead time, particularly in the time of year and other social events and other macro events gives a very, very decent trail for the AI to ingest and make a prediction as to what that customer wants to see and how they want to engage.


Mike Webster:

And not this idea of a static website for everybody. So AI can transform that entire digital journey into something that the Spotify experience. I that's dumbing it down a little bit and that the consumer can come on and immediately go, this brand gets me and they've never seen me before. And that's the, that's the promise of really?


Ryan Haynes:

Yeah, I guess, yes. Right. I mean, if you do look at, you know, where the source of traffic is coming from, you can identify whether it's a Huawei foam or the latest Samsung, you can identify whether they're coming from France and what time of day, and, and, and what they're particularly looking at. You're a real idea already of the type of person that, that, that is. I mean, we're already seeing as well in booking.com. It was certainly a conversation I had of a couple of hoteliers@theftwentytwentytwoaroundsortofhowbooking.com is already suggesting and recommending alternative properties. And it's, you know, really is based on their, their search criteria is based on what they've been looking at previously, the types of rooms and packages that they've been exploring.


Ryan Haynes:

So if we look at AI, then elevating marketing sales and operations, you've made mentioned one example there of, you know, landing on the website and it being really customized to each visitor. What other areas can AI be effectively deployed to engage with the customer on a more nuance level?


Mike Webster:

Good question. I think it just to be absolute, just to paint a picture of what the opportunity is for hotel brands, they can take booking.com experience way off the table and present something incredibly better. So all the AI has pretty sophisticated, but the AI has a revenue trap is ultimately in the margin trap that booking.com is pointed towards opportunities for hotels, by the way, is that they can and have the opportunity now, which wasn't available before retail to customers. So what image they see, what, what room they see, what ruin package they see all delivered individually, each customer would have been hotel necessarily having to invest in, in that AI technology, for example. So this, and there's a way to leapfrog what the OTs are doing so that that's, the opportunity is much broader and deeper now than it, than it could have been before.


Mike Webster:

But somebody coming from a tech talk has an entirely different behavior than somebody coming from, you know, brand aware, direct channel or somebody coming from a measure, for example. So, so if, if you're serving up the same thing to each of those, you're wasting your marketing dollars because tick-tock will have a different behavior. So what, What AI does and proper AI platform as one of the journey, is that they will understand that. Yeah, look at T talk to user at this particular time with that particular device size or this particular lead time, et cetera, big, long string of data. That's the type of customer is, needs to get served this particular journey. So just taking static website delivery into customizable cosmetology products and delivering that to that customer seamlessly.


Mike Webster:

So you can, what would you do in terms of marketing, you know, knock yourself out in your marketing tactics, a hundred percent. You know, it tells a really good at that, but if you're landing in into your own real estate and it's all the same largely, or it's mod or it's or worse, it's, there's some bit of a rules based thing where this is for families. And this is, you know, that kind of thing. You're wasting marketing dollars. So AI will literally put your marketing on steroids, cause it's just bending relevance. You're building to an audience of one, but what's happening now is in the absence of having AI technology, somebody walking up to the, to the front desk and the front desk person who's leading the for first human space is also kind of a raft of all the things. If you've got the time, for example, you got to serve up, you know, try and see if you can upset and maybe sell some stays, or maybe sell maybe some other things during the stay or enhancing the, the, the, the opportunity for the, for more for the revenue.


Mike Webster:

Whereas it, if AI has done all of that and is known exactly what, how to upgrade, somebody knows exactly, you know, whether somebody wants champagne and, and also the comfort of knowing when they're in destination and they're in the property, what they want to see and what the heart is going to be served, whether that, and it doesn't matter if it's a champagne and in the room, or whether it's a theater tickets, all can be served. So from an operational perspective, and even that's just example of the front of house, AI would do a lot of the heavy lifting. So some people think of AI and they think, oh, it's going to roll, modify in hospitality, and it's going to dinner. No, it's putting the human back into hospitality, freeing them up to separate the differenciate brand, the hotel from, you know, digital first companies like booking, for example, in the delivery side or on for Airbnb.


Mike Webster:

So it works right across the journey. So it makes marketing more efficient. It gave the sales piracy and building the brand connectivity. I mean, customers to book easy without having to, to, you know, search and find them bounce back to 50% of customers who book@booking.com have checked the hotel's website and come back. So to get, to, to make, make it a more dynamic channel and an operationally, even at the front of house level, just at the most basic level, it builds efficiencies.


Ryan Haynes:

I must say this, there's nothing worse than working in hospitality and you're constantly having to upsell, or you always have this sort of like sinking feeling inside when you're having to sort of try to suggest additional products and services. When all you want to do is give someone a great time and you can see that value there because, you know, do you know if that person instance instinctively in front of you wants to go and see a musical or wants to go for a romantic meal? You don't, but as you say, the AI is starting to identify those patterns to really understand the nature of that person, the stage. So they can take from the collateral of marketing office offers the most relevant one. And that's just taken us to that unique experience that each of us like getting now online, like we've talked about with Amazon and Spotify and Netflix, and you know, that that's, that's certainly a leap frog in a new direction for, for hospitality that on demand nature in some respects of retailing.


Ryan Haynes:

And we've seen that over the last couple of years, that move towards contactless, being able to provide customers with portals, to be able to access a whole range of products. And they were talking about AI now feeding the products that are most relevant to every, every guest or every web visitor that you've got coming. Why do you think there's such hesitation in hospitality around AI?


Mike Webster:

Part of the problem with AI is people can call it things, AI, you know, AI driven pizza delivery, make it up, right? And then people don't quite in a lot of hotels, equipment get misunderstand what the impact is. So there hasn't, there's no hesitancy, we've had conversations with major US-based, Brian's at a very senior level and they at the chief data level, and they are already on the journey and things that like third parties like us would provide us increases through that. So, but there just hasn't been solutions. Brands are falling in a little bit between two stools where they think contactless is a good thing. I don't think contact was not meeting humans. For example, isn't necessarily a good thing for a brand would have been if the transaction piece can be totally automated and totally understood, and the price and the region, all that delivered to a customer.


Mike Webster:

So when they arrive, you can start hosting. So it's putting the human back to the hosting. So I would say it's contactful and all the transactions are done online and what you desire and require while you're in the destination could be served you perfectly, but the human that puts the human back in front of the human, I think the human in the destination has been asked to do too much in the transaction. Does that make sense?


Ryan Haynes:

Oh God. Yeah. I mean, certainly as a marketeer myself, there's, there's nothing more painful than making sure that you've got all your tactics in place and trying to identify the next tactic. You're going to deploy and segmenting all your lists and making sure that they're all planned and in place. And that they're now having to go through and trying to figure out what's working. And what's not the fact that you've got AI that essentially does all of that for you. Now, you've just got to figure out what it is you want to be offering. And our AI to select from that would just sort of alleviate so much of the pain, the manual pain that goes into marketing right now. But I guess the other areas really for, for hotels that I'm trying to understand is how they can mitigate the lack of talent and skills to utilise AI, because data science, eh, and, and it's, it's a tricky area.


Ryan Haynes:

It's hard to find talent. And especially for hotel brands, because they're not necessarily seen as ineffective or as exciting, sometimes there's maybe a pure tech provider. So what can hotels do then to, to not have to necessarily worry about the intricacies involved in AI?


Mike Webster:

I suppose there is a general shortage of, of a data science and AI, the necessary at scale, like to build an industrial level AI platform capable of serving thousands of models to tens of thousands of hotels simultaneously. So, but I mean, individual level individual at the big hotel brand group, yes, they have investment investments at a, at a, at a smaller hotel group. You're just not going to get to. And even if you did have a couple of data scientists, you're not going to make the impact because you have to be able to do what Instagram does. And as, as a person who's transacting, real-time understand that you want to see another Kathleen piano video.


Mike Webster:

That's an individual models running specifically for you in real time. So there's, they don't need to invest in their own mass team, you know, or they, what they can do is they can collaborate. And at the large brands can collaborate with companies like us. Who've already built its full tech stack at scale with the mass cloud compute and the smaller hotels don't have to do anything. They were already booked in with a lot of the booking engine. So in our example, so when they got to do is activate a, turn it on, and if it start working for you all, it was on always learning and always serving customers better over and over a particular period of time. So there they, I, and there's going to be a continuous drought of both the, if you want to call it the, the, the theoretical data scientists and the, you know, the, the experimental data scientists you want to call it, that there's a drought and what's happened is that all the AI has been controlled.


Mike Webster:

You know, with the, with the surveillance capital efforts of Google, they own the customer. They understand exactly what's going on. They sell that to booking. And then the hotels have to pay booking where the shift back to independent AI within hospitality is that hotels are going to own their customer. They're going to be able to serve their customer. Their own AI platform is going to be able to retail them and understand them better separate the Google, separate the booking. So the leapfrog you mentioned earlier is, is twofold. One is that this idea that guests will come to a destination website and try and figure it all out themselves, that's going to go. So Dave, what AI really does is it elevates all existing kind of book flows and experiences in websites.


Mike Webster:

Instead of having the static journey, you go through, it'll understand exactly what you need to see what image you need to see and distill it into an individual journey per person, per individual humanizes the AI humanizes the journey. So it would say, yeah, here's the package you want to see? Here's the goal. Here's the, here's the theater ticket you want to see? Here's what, and here's the upset and the reason it gets so accurate as adults doing that at scale for everybody in real time. So there's an opportunity for the industry to make, not only just a leapfrog, but to go beyond what Google and go beyond booking.com and go beyond, go beyond to what air and other people who've been doing machine learning for a while, and AI AI, and present customers, individualized journeys. And that's right in front of that's right, as a real and present opportunity for the tar hospitality industry to just serve that up.


Mike Webster:

And then they can add whatever product or service they want. So that puts them right at the top of the pilot, and that is available now. And there's, you know, it's amazing opportunity for hospitality. The first time, the biggest unsolved problem in travel can be addressed by serving direct hotel customers in a way like Spotify does.


Ryan Haynes:

From what I read from a couple of the customers that you're working, large customers you're working with right now, actually this is performing improved performance improves every single month that you're constantly seeing incremental growth in revenue. So as you say, this is, this is more than just implementing some AI. This is implementing learning that constantly understands who's coming to your site and what their needs are.


Mike Webster:

Yeah, I think at the, at the core principle, if you serve guests better, the revenue goes up. So when you're not doing some gimmick, right, what you're not doing is fielding. You know, you're not changing. We don't change the price. You know, what we're doing is we're we understand the customer and say, this is the price bracket or room and rate structure, and image and basket that you need to see. And if somebody scrolls and looks and views, so that's something slightly different than we learned to serve the next customer, have a similar profile, similar, slightly better. So you get this little of incrementalism, anybody that says talk in our platform, you get 30% more is, is smoke and mirrors. So what it does is have you just deployed a on one touch point. So let's, for example, which room and rate combination you need to see, it'll start off a, of percent, four or 5%.


Mike Webster:

And it goes each month, it increases because it's always on and all was learning. But the level of technology, which you won't go into needed to serve that is that if you have one hotel group say 50, 60 hotels, and they're just automating with AI, what room and rates somebody needs to see, you'll have a couple of hundred models specifically built and trained just to do that. It isn't an algorithm. It is hundreds of models, servicing a Tik TOK ad user. You know, you'd have a model for Ramadan. You'd have a ramen, you know, you'll have one for black Friday, you'll have one for, you know, for certain device sizes. You know, and each of those always on all was learning depending on how somebody interacts, the model learns from that itself and then get slightly covered to serve the customer. So it's a low of incrementalism always on, and then nudging it up.


Mike Webster:

But the reality is that the reason you're getting more revenues because you're serving the customer better. The reason everybody's on Spotify is that they serve customers better. That's where the revenue. So it isn't that we've got a little thing that's going to maximize because you're tricking somebody, serving customers, giving them a better experience, building a better digital handshake between the brand and the user or the brand and the guest early, and then allowing you to focus on hospitality. But if you serve them better online, the revenue goes up. And the speed at which they book goes up, their likelihood for loyalty goes up. The repeat goes up an explosion of positive metrics. So it is, it is allowing the hotel to focus on the hospitality, which Airbnb doesn't do and allows them to focus on hosting, which Airbnb says it does, but doesn't to a large degree, but it is putting the, the ability for AI to serve customers better.


Mike Webster:

That's why revenue goes up and you can do it everywhere. What image somebody sees, you know, what upsell they see, you know, what a room upgrade to get, you know, right through the whole journey, 40 different touch points. Each one, serving customers while you're asleep, it's becoming more intelligent and it's linked specifically to your brand. So every single brand gets a set depth and set of models, all was on. That was learning. It's amazing to watch.


Ryan Haynes:

I was going to say, I feel like I want to keep count over data models, but I think, well, we'll definitely leave that for another time. Mike, I've


Mike Webster:

Written down here.


Ryan Haynes:

Well, I do. I think, I think that's enough for us to sort of like cover off, you know, certainly how it's being used today, but definitely get you back in and we'll have a look at those sorts of like data modeling and what that looks like of hotels, because I'm sure that you're seeing some fascinating models within, within the hospitality.


Mike Webster:

Amazing. And what it's showing is book, but the, the, the behavior of a traveler in terms of a deep understanding of the behavior and their needs set and mindset that needs state and mindset, you will not get from a human. You will not get from a marketing or marketecture and understanding of what your audiences are like is that AI really tells at a very detailed level, what your customers are like and what they want in real time, historical data is useful, but it's really where they are right now with the second on how you can serve them like Instagram, cat videos, it's, it lives live elevating, you know, static like a wall of a website and, and booking engine and creating customized hospitality products specifically for that individual.


Mike Webster:

And it's all there and we're working and it's very exciting. I won't go into the data.


Ryan Haynes:

No, no, no, no problem. I think at one thing we have learned from today's conversation is that you like cat videos. So we'll all remember that. That's for sure. Mike, thank you ever so much for joining us and I I'm, I'm sorry. We have to bring this to an end, but I'm sure we could chat forever, but it was really insightful. Thank you. That was my Webster, the CEO and founder of Arvoia check them out as they start to look deeper at how you can better engage with your customers on, through your online portals. I'm Ryan Haynes. You've been listening to Travel Market Life. Don't forget to check out our other episodes. So particularly the outtakes from ICF 2022 and our hoteliers voice series have a lovely day.


Ryan Haynes:

Thank you for listening.

8 views0 comments