ÐßÐßÊÓƵ

Transcript

Introduction
 
Hi, everyone. You are back with Matt Talks, and I have Connor Winders. Winders or Winders? Winders. Winders.
 
There you go. Yeah. With me today. And Connor works in our tech team, and I always keep getting confused every time I speak to Connor, because I was like, is he a developer?
 
Is he a product person? But he's recently been promoted. So now he is both on the development side, but also on the product side of a number of our systems. So maybe you wanna give yourself a short introduction.
 
Absolutely. Yeah. And, maybe one of the nice things about that for me is I don't really see the distinction between tech and and product. Ultimately, everything that we're doing is to solve problems for customers, solve problems for guests, and technology is just is just a neighbor of that, which is why it's maybe confusing chatting to me sometimes on which end I I come at it from because I come at it from the, I believe, the the customer side of things.
 
No worries. Because you sound like you actually talk to customers, which for developer isn't always the most natural thing to do, but you sound like like you've had a lot of conversations with customers.
 
Lots of conversations with customers and lots of watching recordings of conversations with customers and talking to people internally who talk with customers a lot as well. And I find, to be honest, like, throughout my whole career, like, the products that I've built that I most enjoy are when I can really relate to the customer. And it's one of the reasons why I love hospitality and why ÐßÐßÊÓƵ was so attractive to me is that I can really relate to the customer. I worked in hotels when I was a teenager. I worked in restaurants when I was in college and bars, And I travel a lot as part of as part of the job as well.
 
So I can Did you already know that you worked in hotels before?
 
What did you do?
 
Oh, I did, you know, in in the kitchen.
 
I was, working on the floor, working behind the bar.
 
I did I did breakfast chef for summer, which was a fascinating experience as well. I dropped, but but it was a very, very small hotel. So I was cooking breakfast and then delivering them to people's rooms and stuff and had, yeah, an interesting summer that year.
 
Nice. And what do you do at ÐßÐßÊÓƵ?
 
So, yeah, I lead the product and engineering teams, for what we call the back of house division. And so the kind of personas or archetypes that my teams really care about the most are the likes of the GM or the revenue manager or people working in HQ staff.
 
So I have multiple teams underneath me. Each one is focused on specific parts of that of that journey. But, ultimately, it's all about how do we optimize for a hotel or for our customers revenue and for for efficiency.
 
So those are the most heavy duty enterprise features from a b two b system. Right? So they're the most complex user personas that we're trying to satisfy.
 
Yes. Yeah. Absolutely. And and and they overlap so much as well, which means that you constantly, you know, have to make sure that you're not just caring for one persona, but you're caring for all of them in inside of that world. So, it makes for a lot of interesting discovery and then a lot of interesting products to build as well.
 
So, like, the topic of today's conversation isn't that actually. It's it's about AI, and, AI sits also in your remits because it's it's, sitting I don't even know what that team is called. What's the team called by the AI team sits?
 
Well, the kind of internal phone name, was Jarvis when we, when we started, as a nod to, to Ironman, I believe.
 
I love that. So yep. So was that a different name now or still the Jarvis team?
 
It's still, it's still pretty much referred to as Jarvis by by everybody internally.
 
So just to warm up a little bit. So I know that you're going on a really nice vacation far, far away. Have you used AI in any way to plan any part of this trip?
 
Yes. But I but I think I could answer that to pretty much any question that you'd, that you'd ask me. I use AI extensively. I, I'm never really out of ChatGPT. So anytime I'm traveling, I tend to, like, screenshot my itinerary, screenshot my flights, and things like that, chat to, or put it into ChatGPT and have a back and forth on, you know, preparing for, you know, avoiding jet lag, what I should do when I get there, and and so on. And then any sort of app that will give me intelligent recommendations on, you know, trips to take or or or that sort of stuff, I'm I always always use.
 
Love that. Like, I, I use it every day. And, like, some of the questions when you read back some of the questions you've asked, like, the the the AI are pretty stupid, like but you wouldn't ask them to a human. But I'm very grateful to ask them to a computer because, you know, they don't know who I am.
 
But the other day, I screenshot it as part of a handover to someone in the team. I screenshot it whatever the output was, and I've forgotten to remove the sidebar. And it was literally me asking about self tanning to salons in Amsterdam and that. Well, I'm like, oh, god. I should really just try and not share all of my secrets.
 
Yeah. But, you know, there's, like, some benefits to that when you're not sharing it with, with with with other people. Like, for example, I'm a I'm a huge whiskey fan. I know, like, a lot about Irish whiskey, and I

Chapter
AI and humans in hospitality
 
love learning about different whiskeys and spirits and stuff when when I travel.
 
So when I go places, I'll often, like, have a chat with ChattyBT on what sort of distilleries there are or, you know, what the local spirits are and that and that sort of stuff. And so so the trip that I'm taking, actually, I'm going tomorrow to New Zealand. Like, one of the recommendations that Chat GbT gives me, it already knows that I'm interested in distilled spirits and whiskey and stuff. But that's what's able to recommend what I should look at when I'm in, when I'm in New Zealand.
 
So, you know, there are benefits to me personally, but I'm not looking up self tanning salons and sharing that with everybody.
 
The other day, I asked it to draw a photo of me based on my search history, and it pretty it was pretty good, like, what what it came up with. I was like, you know, it's really got something to know me over the last few months. I guess so if you think about travel, and you work at a travel company and we look for people that have a deep craving to improve travel, What's been an experience you've had in a hotel or on the on the road where you thought this is the future of hospitality? This is really good.
 
I I want to cheat and give a couple of examples, and and some of them are a little bit biased, related to technology I was working on a long time ago, maybe maybe ten years ago where I used to work in software for, airlines and travel management companies. And some of the technology we were working on then was pulling in lots of data from different airports and different airlines and things like that. But one of the really, really cool things that I loved was we we built a widget that would tell you what the security lines were in whatever airport that you were going to be traveling through.
 
And just like that tiny little micro interaction of getting a notification and a widget on your screen that gives you that heads up. Okay. You need to leave ten minutes early or you can leave ten minutes later. Those sorts of things are, like, what I love about travel tech is all these little micro interactions that you can have that can, like, genuinely take stress out of somebody's day and make the travel experience much more much more pleasant.
 
So so for me, like, when it comes to, like, airlines and airports and stuff like that, it's, you know, apps like the easyJet app and Etihad and and and those apps where they do that that I absolutely love. And then when it comes to the hotel side of things, like, it won't shock you that I'm a huge technologist. I love playing with technology. I'm like an early adopter of absolutely everything.
 
But definitely, like, the first time I stayed in a citizen m property, like, just blew my mind. Like, the ability to go in and control everything from the little iPad in the room, like, from the light colors to the TV to everything was like to me, I was like, what this why doesn't every single hotel do that? And then I realized that I can actually do it all from my phone, from my personal phone as well by downloading the CitizenM app. And it was like this, like, amazing moment of I often think the best technology is built.
 
Like, if you build it for, like, self-service, and even if it's staying on a self-service key a self-service kiosk or whatever inside of a property, then just give that to the guest. And so to see the way that CitizenM have built the exact same app that's now on my phone. And when I'm, like, arriving to a property, I I always stay at CitizenM now, if if there's one available and and convenient to where I'm going. Because I I I can do I can set up my preferences on the phone and have the room exactly the way I want to when I when I walk in.
 
So for me, like, it's those sorts of experiences in a hotel that are that are incredible. And I and I think about the future of, like, it's cool that I can do it all on the phone now, but in the future, that will just happen automatically.
 
Doesn't stay relevant. So now I, for example, check into citizen and I'm like, I just want to close the curtains as fast. I don't wanna open an app and then go into the app and figure out what the curtain button is. And same with my Apple Watch. I find that, actually, I don't use it anymore. I just need time. And everything else, I don't really care about my fitness status and stuff, and I'm wondering whether it's gimmicky or whether it really solves a problem.
 
I so so I think that the further evolution of some of this technology does solve more of those problems. So so like I was saying, like, it is cool that I can do this stuff on on the app, but it should learn. And it should learn like, this is where AI and machine learning and stuff are gonna really play an incredible part in in the future is that it should learn exactly what my preferences are based on why I'm there, the time I'm there, where I'm coming from, what I'm doing, and start to change those preferences automatically. So when I walk into the room, it's exactly how I want it to be, and I didn't have to do anything except stay there previously.
 
So so it's that kind of evolution that I think is, like, super exciting, but I still enjoy playing with it with it right now. And and I thought I was the same with the Apple Watch until I went a few days without it because I forgot my charger. And and I actually really, really missed it. I really missed, like, little things like the, like like the simple reminders to get up and stand and to get up and move and and things like that.
 
So all that off immediately.
 
That was so annoying.
 
I well, I see, I I have a water bottle here that glows when I should drink. So I'm all about, like, technology helping me be, more efficient and healthy.
 
Yeah. I love it. So now being in the hotel space, is there one thing that you would love to fix? Like, something that really frustrates you in hotel that you're like, if I can get my hands on that thing, I would go and fix that tomorrow.
 
I so despite the fact that I work in hospitality and and and I do love hospitality, I also like, I'm a huge introvert.
 
And and it's really, really hard for me to have to interact with people sometimes.
 
So the more and more I have to do things like pick up a phone to get room service or, you know, if I go into a room that's dirty or whatever and happen to pick up a phone or go down to reception, interact with with people, like, I I find it really hard. Like, it's really, really difficult for me for me to do that. It's not just that I don't like it. It's it's it's challenging.
 
Like, so so the more I'm not an introvert, and I don't wanna be off that phone.
 
My husband always, like, I know you called them. Like, you called the room service person. And then I just tell another human, like, on the phone what I wanna eat and then, like, the guilt that comes with my me ordering a burger in my room. It, like, I Yeah. I totally recognize this.
 
So so so it's all of those sorts of things. Like, I think it's great to have the option to interact with with with other people to interact with with humans.
 
But if that can all be digitized, then you have then you have the option. Right? And and and you can take a lot of a a lot of pain, a lot of stress, or a lot of embarrassment out of, out of those situations. Like, whether you're traveling for business or you're traveling for for leisure, like, staying in a hotel is the opportunity to be a really, really nice experience.
 
It's one of the reasons I love working in this industry. Like, it can be such a great experience, but anything that gets in the way of that. And for me, that's often those kind of awkward human inter interactions. Anything that gets in the way for me is something that, like, I'd love to see technology start to start to solve.
 
And and, thankfully, you know, that is exactly the sort of thing that we're that we're doing is find finding ways to make it more seamless, taking friction out of it for guests, for staff. So Yeah.
 
And I think it's often discussion that I have with hoteliers that they're like, yeah. But we want to create these really personal interactions. And I'm like, yeah. But that's what you want.
 
But what is it that your guest wants? Because, yes, I get it. You're a luxury hotel, and you want to create these really curated personalized experiences. But maybe your guests are looking for a digital experience.
 
They're looking to be served in a different way, and I think that's the difference of providing choice versus forcing, service, I guess.
 
Yeah. Like, that that optionality, I think, is really, really important because different people want to experience your property or want to experience your service in a in a in a different way. And so having that choice and having that optionality, I think, is is really important. But, also, like, where where technology can really shine is, like, making sure that those personal interactions and those in person moments are exceptional.

Chapter
The concept of invisible hospitality
 
You know, like, if if if I have to pick up the phone and talk to somebody, like, that to me, that's really, really awkward. But if I can, you know, walk into a room that automatically, you know, you know my preferences, you know the types of meals that I get, you know the types of pillows that I get, and it's there. But to me, that's just a much, much nicer experience. And then when I walk down to the lobby later on my way out, you can ask me, did I enjoy the the the, you know, the the the pillows or whatever it might have been that you left in my room as a as a result? That to me is an amazing personal interaction as opposed to me having to phone for it.
 
It's the invisible hospitality.
 
Like, when you don't even know that they're deeply thinking about your experience and then you get these little moments instead of the obvious, like, in your face. Like, we are luxury. We have we must have a Belbin at the door who's gonna carry my luggage, which I don't really care about because then I have to wait for the bell band to come to my room. And the the little small things that they do, underlying it all, that they don't have to put in their face that they've done them for you, like, that's what, for me, luxury travel is.
 
Yes. Absolutely. I am I am a hundred percent on board.
 
So let's talk AI.
 
Yes.
 
What have we done in news? Have we done anything significant in use with AI?
 
We've done a lot. We've we've we've done a lot. And what's really interesting, I was reflecting a lot on this, recently.
 
I believe, like, we've probably experimented with AI and learned a lot about AI, maybe more so than people than people realize. So today in the product, there's certain things that you're gonna see, that are driven by AI. So we have what we call smart tips, which appear on guest profiles, and they're starting to appear throughout the rest of the product as well. But they'll tell you things about that guest that allow you to have that moment of, that moment of insight or creating those delightful experiences. And what we're doing there is basically, you know, looking at all of the notes and everything that we can find within your system that that talks about that guest and giving you a little summary of here's here's something you should know about them. Here's something about their preferences. Here's what they did last time and so on so that you can then interact with them in a really, really nice personalized way.
 
But we'll start to extend that as well and and and look at your actual business performance and where the opportunities are within, you know, how you use ÐßÐßÊÓƵ, how you've set up your property, the products you have and so on. So we can recommend, hey. This would be a nicer this would be a nicer product for, for you to sell at this time of year and and and so on. But all of that comes from, you know, what what might look like a small little interaction within the product comes from a huge amount of work and a huge amount of data that we've been processing over the last couple of years as we started to really understand the power of AI. So we invested a lot in terms of how do we how do we understand what users are doing within the system.
 
And a lot of that interaction happens through the search bar.
 
So we experimented a lot with how could we automate things and intelligently perform actions based on what a user is doing through the search bar. And we have a lot of case studies where that works flawlessly, but we also have a lot of case studies where that doesn't work. And so we're able to learn enough about the technology to know that that doesn't solve the problems we thought it would solve right now, but in doing so landed on that idea of smart tips that we now have rolled out throughout throughout the product. So, it's that sort of experimentation because AI is such an it's still an early stage technology for this type of use case.
 
It's that sort of experimentation that's been invaluable to us to really understand how it can add value. You know, AI for a long time has been a solution without a problem. And through that, you know, through that experimentation, we're starting to find what the problems are that it's really, really good at at solving. So that's one example.
 
We, you know, we we have others, like, when you chat to our our support widget within the product, the first thing it's gonna try to do is attempt to resolve your problem automatically using AI by taking information from our knowledge base, by taking information from other support case and so on. I'm trying to guide you through what you should actually do in order to answer the question that you're asking. And only if it can't, then bring in a human to talk to directly. And and and even through that, like, again, almost like experimentation, one of the things that I found personally is if you're trying to configure ÐßÐßÊÓƵ, I'm trying to set up ÐßÐßÊÓƵ, the single best way I found to do it is chatting to that chatbot by saying, hey.
 
I want to configure this type of product, and I want to be able to charge this for it. How would I do that? And it starts to bring in all of that information and give you step by step on how to actually do it. And and and, again, like me personally, I found being able to set up ÐßÐßÊÓƵ from scratch the fastest way to do it by interacting with that AI chatbot, which it was never even created to do.
 
So it's, again, those sorts of experimentations, are becoming invaluable in learning how we're gonna use it more and more in in the product.

Chapter
Problems with AI adoption
 
I think you think about this deeply, and you've embraced AI in your day to day life. We found a lot of resistance from hotels once we started introducing this bot on the support side, especially especially in the French region. The French hotel is like, no. I wanna speak to a person.
 
I must speak to a human. And I was like, yeah. But the bot's actually sometimes better because it's trained on everything we have in terms of data, whereas people aren't always fully, fully trained on all of the topics that you're looking for. So, actually, the bot performs incredibly well.
 
But there's this Yes. Journey that we have to take hoteliers on to embrace the bot because, actually, it it's pretty good.
 
Yeah. I outside of hospitality, I give an example. I, I use a red light therapy belt for for some of my running injuries, and I bought it off this company called, oh my goodness. The name the name escapes me now, but that's not that's not important.
 
The no. No. It wasn't. But the belt the the the the belt stopped working. And so I emailed their support to say, look, this belt stopped working.
 
I bought it a few months ago. It's really good. I wanna get it fixed. I wanna get it replaced.
 
And I had a couple of emails back and forth, where it ended up with them saying, you know, keep the belt that you have. At least it works some of the time, but there's a new one in the post on the way out. They sent me the tracking number and and all of that. And it was only afterwards that I looked at the email again, and the entire conversation had been with a bot.

Chapter
AI in customer support
 
And I think this is where it's gonna get even more powerful. Right? It's like we're used to interacting with or we're starting to get used to interacting with some of these bots and open, ChatGPT and so on in a very conversational manner, but they're not yet taking actions on on our behalf.
 
This support bot did. It ordered the replacement. It's you know, sent me the tracking number. It did everything that a human would have done.
 
And the way the tone in which it spoke to me was, like, some of the best service I've I've ever had. And so we, you know, we might still be somewhat early in general in this idea of chatbots, but I've seen firsthand that they are capable of serving a human, at least in that context, better than I've ever been served by an by another human. So it's a journey to go on and there's trust to be built. But I'm fully of the belief that that that this is that this is possible.
 
Because these AIs, they're they're processing so much information. Like you say, sometimes a human just isn't trained. It's not feasible. Even with years of experience to know everything, but it is feasible for an AI very, very quickly to have all of that information. And when we start unlocking more of the actions that it can take, that's a far more valuable, way to or far more efficient way to to actually resolve, a a lot of support cases in in my opinion.
 
Yeah. Like, I stayed in the luxury hotel the other day, and and then before arrival, they sent me this questionnaire, this, like, Typeform type of questionnaire, and saying, what kind of pillow you want? And then they had, like, ten pillows, and I had to click on all of the options. And it was this endless questionnaire, and I go through it because I, you know, I want to just see what the tech does and if they actually deliver my crazy request.
 
And it just felt so outdated now because this is all like, this could all I've been a really easy chat where I just said, hey. I'm allergic to feathers. Can you replace the pillows or something? But, like, we're so far away today in hospitality from getting that right at this point. Is this something that you're like, you're seeing this in your life? Is this something that you're thinking of fixing Immune or helping build Immune?
 
Absolutely.
 
And and I would say, like, some of it in ÐßÐßÊÓƵ and some of it, you know, we work with partners who are, like, doing incredible job at, you know, this sort of thing as well. So, you know, we work with the likes of DOO, with the likes of OKI and so on. And some of their technology, like, is is really incredible in terms of how it's able to understand the guest, how it's able to help tailor guest experiences based on what they understand about the guest. There's a lot of this that, you know, is based on data that exists in Muse, and there's really interesting things that we'll be able to do in Muse for sure. But I think, like, at least for the foreseeable future, like, the way to solve this for us is, like, really great partnerships with people who are already doing a great job, a a great job in us.
 
As you look at the ÐßÐßÊÓƵ ecosystem of of products, like, there's certain things that today we're looking at, you know, dietary requirements for our POS for, you know, food and beverage. That information being shared and being intelligently used between the PMS and the POS is incredibly valuable and something that we can unlock very, very easily for for for our customers.
 
But, I mean, in short, it's absolutely an area that I think we we have a really, really big role to play. There's just so much context and so much information inside of ÐßÐßÊÓƵ that when used the right way can be can be transformative.
 
And, like, we have to almost stimulate integration partners to give us more data because, you know, we've not never really embraced people to send those data into these notes fields. So we have, for example, guest review, which is this integration to guest reviews. And if you leave a review on Google, triggered by guest review after your stay, we post those reviews back into ÐßÐßÊÓƵ so that you can absorb those into those tooltips again. So we're just, like, looking for more data.
 
Yeah. Always hard to get a hold of, and we need to convince integration partners. So if you're a hotelier and you have a great system, talk to them. Get them to integrate more data into our system because we can absorb pretty much anything.
 
Right?
 
Absolutely.
 
And and you kind of touched on one there that I think is a really, really interesting use case that that we're experimenting with and researching as well, which is the correlation between reviews, public reviews, and revenue.
 
So so this is one of the reasons why, like, Atomize is a match made in heaven for us, the way that they've been thinking about some of these problems, which is basically over time, the more that you're able to drive five star reviews, what does that do to the rates that you're able to charge in the future? And when you charge too much, what does that do to the rates? When you charge too little sorry. To the ratings, when you charge too little, what does that do to the ratings? And there's all sorts of really, really interesting mathematical models, and AI that we're looking at in this area to help hoteliers be far more thoughtful about, okay, revenue has a direct impact sorry, has a very real impact on star rating, which has a very real impact on revenue.
 
And because we're able to, you know, work with partners to bring some of this data in because it's not all going to live in ÐßÐßÊÓƵ, we're able to look at some of these really, really interesting use cases that otherwise, like a PMS on its own, could never solve for.
 
Yeah. Yeah. So if you talked about Optimize and and the AI that they have is different. It's like machine learning AI. Yeah. How do you how do you explain what they do with machine learning versus the large language model tips that you talked about earlier? What's the difference between those two models?
 
So I guess if if if I could sum up what we do with Atomize, and I'll I'll oversimplify it for the sake of this. Like, Atomize is an incredibly sophisticated mathematical model that takes in a lot of different data sources and performs calculations on it. Whereas with LLMs and generative AI, that's more about in you know, taking in vast amounts of data and information to have a conversation or eventually to take some particular action. Whereas Atomize is much more specialized in kind of the the the mathematical side of things.
 
And over time, learning what different data points and different, different inputs will do to that model based on what the outputs are coming out the the the far end. So, I mean, what what I think is incredibly exciting about Atomize is not just that model, the and and the fact that that model gets smarter over time and and and and so on. It's actually when you look at the customers who are most successful with Atomize, they're the ones who turn it on in autopilot mode, which is then the kind of the next level of automation and then AI on top of that, which says, we trust that this model is so smart and continues to get so smart that it's just going to manage our rates for us.
 
And, you know, it takes a while sometimes to get to that level of trust, but being able to step back and let the system actually generate more revenue for you and show you that it's generating more revenue for you is, like, an incredibly powerful thing. On top of that, what we're doing also with AI in the atomized world is explaining why we're setting the rates that we're setting or recommending the rates that we're recommending. Because the mathematical model is so complex and it evolves over time, building that trust can be hard. Right? Often, what you just wanna do is look at, well, what are my competitors down the street charging? Maybe I should be doing that without factoring in the hundreds of data points that are going into the atomized model. So now what we're doing is using AI to actually start to explain why we're making the recommendations that we're making to build that kind of trust level up, as opposed to blindly, blindly assuming that it's right.
 
Yeah. Like and and there's a lot of pieces that are currently ongoing between the different teams about how we're thinking about driving better businesses. If you're running on ÐßÐßÊÓƵ, we want to help hotels drive a better business. And revenue management seems like a no brainer, but the majority of hotels still today do not use a revenue management system.
 
So, our accounts team is right now building data insights that we can visually show people that the traffic part, the total revenue per available room in your hotel versus your benchmark hotel is actually far behind or your traffic pan, like, per square meter, for example. And we're trying to really tell the story of, like, you could make a lot more money or you could do run a much better business if you're thinking more holistically about driving revenue through these amazing AI tools that we have. And it's this I feel like we're reaching this full circle moment finding that as a hospitality industry, we've been waiting for for so long.

Chapter
AI and revenue management
 
Because we've still today, we still talk about RevPar and these antiquated metrics.
 
Whereas, you know, there are much more powerful data points that we have that we can feed to you at the right moment so that you can strategize on top of that in some way. Okay. So I think people will start to see some of those pieces come together in the next few week few months.
 
Yeah. When you when you you nearly stepped and said few weeks, I was, just gonna say, Matt, I've got some idea.
 
Weeks is fine. Weeks month.
 
But there's but but for sure, there's incredible, incredible opportunities and things that we're looking very, very seriously at and making progress with. So so atomize and the revenue side of things is is is one piece, but we're also investing a lot in how we do, like, provide intelligent insights on your overall, the holistic data within within the product. So as you say, you know, not just that, hey. You're earning this amount or your rev power is this or your t rev power is this, But, actually, here's what you could change in order to increase it.
 
Or if you change this, it's going to decrease and so on. So, we are building an intelligence layer into ÐßÐßÊÓƵ at the moment that will start to surface all the way through the product, not just in one dashboard or not just in one report. We'll start to show you through the product building on that smart tips idea what you should be doing in order to optimize your bid your business. And and to me, this is, like, the most exciting thing that that you're gonna see in the next while.
 
I love that. And I think that's really there's this fear from revenue managers that that these AI tools replace them. But, actually, I think that there's a, an evolution happening in the job description of what a revenue manager is, and it's total revenue of the hotel. And you have to Yeah.
 
Not spend your time in Excel sheets trying to figure out why we had such a pickup yesterday. It should actually be, well, what are we doing tomorrow? Like and how are we driving more revenue tomorrow? And it's quite a shift that's taking place, but we're trying to put the pieces in place that help serve them so that these tool tips that you talk about saying, you know, maybe that's an area that you're not performing particularly well in.
 
Here's some suggestion potential integration partners that you could look at. And and that makes that job of a traditional revenue manager actually much more interesting than just data crunching.
 
A hundred percent. And and, actually, if I draw a parallel to even software engineering or product development, like, it's not uncommon for there to be a nervousness around AI and, well, what does that mean for my job, or what does that mean for the expertise that I currently have? But, actually, the way to look at it and the way we look at it is, like, yeah, what does that mean to my job? Because I'm now going to be doing much more high value, high impact, interesting work.
 
And, I shared on our our company, All Hands, last week, we've rolled out an AI code assistant for every engineer at ÐßÐßÊÓƵ. And what this, code assistant is really, really great at is taking a lot of the kind of grind work and the grunt work out of software engineering. So as an example, thinking q four, we had over twenty three thousand code suggestions from this, AI assistant that we then went and implemented. And they're all things that, otherwise, an engineer is sitting down and writing for minutes or hours at a time as opposed to solving the actual problems that our customers have.
 
So it's not that we suddenly need less engineers. It's that the engineers that we have are doing much higher value work as a result of not needing to think about that. And the same will be true in so many industries and in particular in hospitality. It's not about how you need less people.
 
It's about maximizing the value that those people bring and the way that they're able to think about their jobs will change dramatically.
 
And I think positively. Right? So if you're not job.
 
Would I just get to do more of the things that I really enjoy, which is a deep thinking and making a change, I'm more likely to stick around. And we have such turnover in hotels where people are entering hotels and they become disillusioned because we go up thinking, oh, I wanna work in hotels. It's such a fabulous industry. And then you end up entering data into a system. Whereas I think with AI and some of the exciting insights that we can provide to this atelier is saying, here's an insight on this human that's about to arrive. Do something fun with this insight that we give you. And if you can create that into actionable fun experiences, I think people will be more likely to stick around.
 
Yeah. I mean, I know I certainly would rather be doing that than blindly looking at a screen and bashing it bashing the keyboard, right, which is a lot of what the manual work ends up being today. And I think we can remove that and create much more interesting interactive, and valuable experiences for sure.
 
Nice. So, we shared our four big mega trends for this year earlier this year where we said, okay. These are the things that we're focusing on. Productivity is one of those where we've seen or the data shows, and this is an industry trend. This is not a news trend, where hospitality has not seen a gain in productivity in the last couple of years, whereas other industries have seen a significant gain in productivity.
 
Underlying that, is it an AI problem that we have, or is there a bigger innovation problem that we have to tackle before we start to reach for AI to solve all of this?
 
I think, like, AI can be the solution to many things for for sure.
 
But I think some of it is embracing technology in in in general.
 
And we shouldn't, like, always conflate, like, technology with with AI. There's a lot of great technology and great automations out there that can solve problems, make people more efficient, and make things go make things go faster without the need to actually layer AI on on top of it. So so for me, some of it is actually a maybe resistance is too strong of a word, but a lack of embracing technology and different technology in order to be to be more efficient.
 
AI will just further accelerate that in in in the future, but there's a lot. Like, simple simple things like or what might seem like simple things like like check-in. Like, if you need to have people checking in long lines at a at at your hotel lobby, like, that is there's no objective measure that says that is more efficient than allowing a number of your guests to check-in online.
 
You know, and, yes, it's still very uncommon, at least in my experience, to be able to check-in online.
 
And and and and a lot of properties that I stay at, was they offer online check-in. I still then need to go and queue up at the desk because they haven't actually fully embraced the technology. So so for me, it's actually embracing the technology first. And that's going to unlock a lot more with AI in the future.
 
Like, I still get photos from friends of mine who are, like, in a queue at a reception desk and they take a photo of it, like, here's a hotel you need to go and fix. And I'm like, yeah. Like, I don't think it works like that, unfortunately. Like, we have a very long journey ahead of us, but it's a real frustration.
 
People do not like the queue. And if they you give them the digital option, they actually end up using it. The the thing we've seen is a shift in its the kiosk, for example. Like, COVID was great, and you now have kiosks in your supermarkets and all of your shops that you go to.
 
And I quite enjoy ordering my McDonald's, order on a kiosk rather than telling someone else that that I want three burgers. Like, I don't know. Tell a person that that's what I'm eating.
 
And and that's led to much stronger adoption of kiosks. So you can divert some of these these, queues that are happening in hotels, but we've completely diverted from AI. I don't know. I'm right there.
 
There's this one worry that people have, and I've seen it on my LinkedIn post where I talk about AI and these tooltips that we have, which is data security.

Chapter
Data security concerns in AI implementations
 
Because I don't want you to, like, use my personal data in an insecure way. How are we protecting that personal data?
 
Yeah. So so there's a number of different ways and number of different things that that that are involved here. But, like, first off, we only use data within the bounds of ÐßÐßÊÓƵ. So our entire AI infrastructure, which is built on a technology called called Databricks, is, confined to the world of ÐßÐßÊÓƵ.
 
We're not taking that data and sending it off externally. We're not sharing it with third parties. We're using it internally and as part of the ÐßÐßÊÓƵ dataset. So it's already bound in terms of, you know, where it's even available.
 
Then, like, from a from a personal, or or personnel, perspective, like, once we need certain data in order to to do things in the system, like, it's limited in who's able to access that data. And then that same principle holds through to what the technology actually is able to do. So when we're passing stuff into a model or training a model or building a model, we're only putting in the data that's absolutely needed for the context that we're trying to solve a problem for. So guest data for Matt, for example, is only going into a model that is building a recommendation or a smart tip from for for Matt.
 
It's not going into a model, that builds a recommendation for a different hotel, for a restaurant, somewhere else. Like, so we're very, very specific about how we use the data for different for different models.
 
And then, ultimately, things like what we're doing with the Atomize product, for example, where we're actually then starting to explain why the algorithm is doing the things that it's doing and why the AI is making the recommendations that it's that it's making, will also help long term as well because you'll be able to better understand how the data is used and what the data is used for and why it gave the result that it that it did.
 
Nice. Yeah. So, like, I was thinking sorry. I've completely lost my train of thought there.
 
We I mean, we're on on the topic of of data, I guess the other thing that I'd say is, like, really important is everything within the ÐßÐßÊÓƵ world is encrypted at the point that we send us anywhere. So so data from one point to another into an into a model, out of a model, etcetera, we use industry standard encryption technology to make sure we're not just blindly passing around raw raw data for anybody to, for anybody to take. So we're passing just the right data to just the right place, and it's always securely encrypted.
 
So the question I was gonna ask was around if I stay at a chain hotel and I have information in the one hotel and the next week I stay in a different city within that chain. Is there an information sharing that takes place? Does the AI tool tip pick up on some of my preferences about allergies that I have at the one hotel and then tells the people at the next hotel?
 
So that is possible, but it's not something that we're doing at the moment.
 
It's definitely something that, like, is a really, really interesting use case. But I think there's also there's elements to that that you want to understand. Like, does a guest actually want that, or does that creep a guest out?
 
It creep a guest out.
 
Right. So, so so for sure, the technology is is is possible. But right now, what we use is bound to the actual property.
 
Nice.
 
Integration partners. So we can't do everything ourselves. So are there any integration partners that do something that you're like, that's quite interesting, actually. That that's an interesting angle to it.
 
I think I I mentioned a couple of them earlier with Duve and Okie and then HiJiffy. So I don't know if I I mentioned. So so very much around the kind of guest experience, guest journey, upsell, experience for guests. I'm re I I find that really interesting because like I said earlier, being able to kinda put yourself in the shoes of the customer, because I travel so much, because I I I stay in hotels so much, I'm always amazed at how bad hotels are at upselling me on things that I would actually buy if they just made us if they just made it easy. And so when you look at something like OKIE, and the algorithms that they're building in order to make intelligent recommendations based on, you know, on on your profile, where you're going, I think that's really, really interesting.
 
I'm I'm one of those people I like. I'm more than happy to spend more money at a property if it's something that I want. Just make it really easy, and make sure that, like, you're reminding me that the thing that I want is is available.
 
So so I find that, like, really, really interesting. And then I think, like, some of the stuff that HiJiffy is doing, some of the stuff that Duve is doing around how they like, their their messaging capabilities with guests, I think, is really interesting as well. So being able to almost like that experience we talked about earlier where, you know, a support bot is able to intelligently answer and speak to to to to a customer. Being able to do that with a guest, I think, is really interesting, as well when it brings in, like, high quality translations and localizations and things like that as well, being able to serve guests much more easily where they're coming from, and in the way that they want to be served.
 
Right? Whether that's through WhatsApp, through iMessage, through email, in a language that they understand. Being able to do that with technology, with AI, like, is really, really exciting because previously, you need literally teams of people to do it or or somebody who's able to, like, be very, very multilingual, and operate multiple keyboards at once. So so I'm really interested and excited by that type of technology as well.
 
Yeah. Like, I was talking to the founder of Runner dot ai, which is this chat app that starts to communicate with guests through, WhatsApp, I think. And I like the use case that they've integrated it into a way they can read the reservation. They can see if breakfast product is attached.
 
And if it isn't, then it can actually do that end to end upsell journey, like the example that you gave earlier about your fitness belt. You can actually engage this entire chat. And then if they sell breakfast, they will attach the breakfast item to the reservation automatically. No humans involved.
 
And I think those are some of the use cases that are painful because, you know, people want to eat breakfast, and they wanna buy that, and they don't wanna go to the reception. Often, I go to the restaurant and say, can I have breakfast? I don't have it included. And they're like, you gotta go back to the reception to add onto your reservation.
 
I was like, oh, god. It's so annoying. And I actually don't know if you can solve this right now.
 
Yeah. And and like other use cases, I don't know why it came into my head, but, like, anywhere I go, if I if I if I could get a massage, like, I probably want to take it. But I can't remember the last time I did because, like, I'm not actually being intelligently upsold that type of product.
 
But if I was receptionist who has a brochure that you can't take with you, and then I have to go back to call them to order that thing.
 
Yeah.
 
And and then you have to have that interaction when you're booking it where they ask you what time you wanna have it. And then you say I want you'd say the time and say we don't have that. Is there another time that suits? As opposed to, like, why can't the technology just tell me these slots are available for these types of treatments than the last time you chose this one and and recommend it recommend it that way.
 
Like, there's so many of those types of experiences that I think can be massively leveled up through through technology and through AI. And where AI and machine learning is playing a role there is, like, really understanding that Connor, who's a business traveler, is likely to want this package based on the fact that we know other profiles like Connor want this package. So I think there's a there's there's a huge opportunity there. And like like you say, there are there are partners already in this space, but I think it's it's something that should be adopted everywhere and, in my experience, is not nearly adopted widely enough.
 
Last question.

Chapter
Advice for hoteliers who want to implement AI tech
 
So it sounds like you've embraced AI, and I have embraced AI in our lives deeply.
 
Like, my I don't even Google anything anymore. I go straight to AI. Unless I explicitly know what I'm Googling, like, the website I'm looking for, I will go and ask my questions in AI today.
 
But we do feel a lot of hesitancy from hoteliers who are, like, not sure how they should use it. What's your piece of advice that you'd give a hotelier about what to do tomorrow, do differently?
 
I think it's the same advice I give anybody, to be honest. When I was you know, back when I was learning to code and learn to build software, there's this concept of the rubber duck. And everybody has a rubber duck on their desk that you like when you have a difficult problem, you say it out loud to the rubber duck. And just that experience of being able to say it out loud actually helps you solve solve the problem.
 
And I find, like, the best way to on kinda on ramp into learning to use AI is to treat it like a rubber duck. If you have an idea, whether it's ridiculous or you think it's great or whatever, so literally, like, chat to something like ChatGPT or one of those, one of one of those products and just get an opinion. And and that turns into a conversation. And through that, you may be completely disregarding everything that the AI is saying, but it's actually helping you frame your own thoughts.
 
And I find, like, a lot of the use that I have with ChatGPT is actually that. It's ideas. I I wanna bounce the idea around. There's there's nobody else here.
 
I'll bounce it off ChatGPT, and that will help me formulate the idea. And through that sort of behavior, I've just become so much more comfortable with AI in general. And so whether it's a hotelier or any industry, I actually think that idea of treating AI like a rubber duck is really, really helpful to start to build a way, and comfort with how you interact with it.
 
Like, especially in areas where you're uncomfortable. So some of the areas that are not native to me would be, for example, a financial statement that I have to read and some metrics I might not understand. So I throw it in the eye and say, like, hey. Explain this thing like I'm a five year old, And it's so good at simplifying some of these concepts.
 
Because if I were to ask our CFO to explain something like I was a five year old, he's like, well, that you're you're clearly in India. Right? But AI, you can ask all these dumb questions too. No judgment.
 
Yeah. I I use it, for for all sorts of things. Actually, I do a lot of running, and I was training for a for a marathon last year.
 
Strava now has a feature that gives you AI insights within it, which is amazing. But, actually, even before that, what I was doing was just taking a screenshot from the Strava app, putting it into ChatGPT, and getting advice on, like, how should I adjust my training? What went well? What didn't what didn't go well? And, like, it's incredible that actually just interpreting any sort of information like that and giving you ideas.
 
Unlike that kind of idea of it, you know, sometimes in product development, we'll be looking at vast amounts of data and trying to understand what it actually tells us. It can be really, really helpful to just screenshot graphs and say what, like, what insights would you take from this graph? And then that starts a conversation where you actually start to frame your own thinking because you're getting these little these little nudges.
 
That is so nice. Connor, I could talk to you for hours, but you have a flight to Perth to New Zealand. So I'm gonna leave the conversation here. But thank you so much for joining me today.
 
No problem. Thanks for having me. It was great.