Wanting to Adopt AI Is Not a Problem But Lack of AI Fluency in Commercial Teams Is

Speaker A: Foreign. Discussion with one of our expert partners. Hope you enjoy it and feel free to suggest other topics or ideas you would like us to discuss in the future. Hey, thanks for joining us today and being part of this video series. We'll kind of discuss what it's all about in, in a minute. But before we get to that, uh, how, how are things with you guys? We are kind of. I can't believe we're halfway through the year. How's 2026 been treating you?

Speaker B: Really, really well. You know, it's been a really interesting year. I'm sure like everybody you talk to, you know, things are moving very fast. There's a lot of fun and interesting stuff seems new each day. Um, right. To learning to consider. So it's, it's been a wild year. It's been a lot of fun. There's a lot of really cool stuff going on.

Speaker A: Excellent. I look forward and what we'll have to do is we'll have to get you onto our um, predictions that we like to do at the end of the year uh, and kind of reflect back on what 2026 was like. But also we can look into 2027 and see if it's going to be as crazy then as ah, as this year has been. But that, that's for another time. This discussion is really about uh, I mean the whole, the whole shaping of this conversation came from a, came from the thought that everyone was talking about A.I. uh, but everyone, everything is written assuming that, that every hotel is ready to just layer it on and away we go and it's the magic wand that's going to solve all the problems. And the reality is we know that kind of isn't the case but rather than hearing me prophesize about it, we thought well let's get our partners on board and they can actually talk about what they see, what they hear and what they experience as you are the guys who are dealing with hoteliers all the time. So with that in mind let's dive into our, our first question. So all of our audience is within the commercial space within a hotel operation. So revenue, marketing, distribution, sales, G.M. m. Um, so we're talking at it from, we're taking this topic from that perspective. So the first question I want to throw over to you is how do you assess whether a hotel's current commercial operation is genuinely ready for AI Um, and what do you see as maybe some of the most common gaps that you find that maybe people do admit to or maybe don't admit to?

Speaker B: That's Million Dollar Question. Right. Um, so I actually think that that question and kind of that categorization itself needs, needs a little unpacking to really be answered. Well, to me, to an extent, questions like that, it's almost kind of like asking, are you ready for the Internet 25 years ago? Um, because AI covers so many different technologies that solve very, very different problems. So being ready for AI really only means something once you know what you're talking about and what specific thing that you and your commercial team are looking to improve.

Speaker A: Yeah, good point.

Speaker B: Right. So, uh, I mean, the first thing that I would look for or you know, in assessing if people are ready is can they answer that question? Like what, what specific outcomes are you trying to move or are you trying to influence? Because we want to incorporate more. AI is not a problem statement. Right? You can't. I mean, you can solve it, but it's not going to be effective or worth anybody's time. So, you know, not, not responding to the market quickly enough with price changes, or, um, your, your inventory statusing and restrictions, or not getting the right type of spend to the right type of marketing, et cetera, those are actual problem statements. One, they're measurable. You can assess them. And they're also all different types of problems. So once you can understand that, then, and this is one of the gaps that you'll find too, is understanding what type of AI actually fits that problem shape. And this kind of gets into, does your team have a basic fluency and level of understanding in this area? Not expert. You don't need a data scientist. Right. But, um, making sure your team is kind of conversant in AI types of AI, what that actually means, that's a step that a lot of people skip. I mean, not understanding the difference between. Well, first of all, there's a lot of stuff as AI. A chatbot is AI. The, the LLM that can actually create new answers, generate or an image generator. That's AI. Um, machine learning, reinforcement learning, neural networks, like this all falls under that umbrella. They all are capable of very different things, and they all are used for very different things. So that's kind of the second step is you have your problem statements, you know what you want to impact, what is actually the right type of AI to look at for that. So like, again, we talked about LLMs, and I think that's everybody's favorite thing right now. That conversational layer is very different than something that can pull data from a bunch of different systems and compile it into report and especially automate that. And you Know, get that sent out on a schedule. That, that's agents. That's not, you know, the same thing as what you want for asking questions and then something for forecasting and pricing and an actual optimization recommendation engine. Again, very different. Right. So your team really needs to understand the difference and that's a big gap that needs to be addressed.

Speaker A: When you talk about AI, you know, the, the, the man, woman in the street will think about things like chat GPT and those conversational or you know, the, the chatbot as you said, that now comes up when you're, when you're engaging with a website. But that's a very limited scope in a sense of what AI can do. But that's how, that's how people perceive it. And so I think you're, you're uh, highlighting the problem and then what is the tool that you might need and then what is the understanding within the team? Yeah, I think they're very critical factors. What's interesting is that um, you're the first person I've spoken to that didn't use the word data at the starting point of the answer. Ah. So it's kind of refreshing to hear a different angle and thought why, uh, what made you, what made you feel that that way of answering was, was what was front and centering in your mind?

Speaker B: I mean the data is certainly part of it. Um, but also there's only, and that goes a lot to, to the, to the size of your operation. Right. The, the data discussion for you mentioned a GM with a single property, even if it's a lot of rooms, is a completely different situation than if you're talking about an enterprise operation. Um, that maybe even has different systems across their different properties. But they need to roll that all up to an enterprise level. Um, and they're in terms again the, the data governance and security and there's just, it's, it's a very nuanced conversation. But also at the end of the day your data problems may need to be fixed by AI. If it's like, if it's bad quality, you know, it's, to me it's um, when we're just talking about general commercial readiness, I don't think it's where you start.

Speaker A: Yeah, that's a very, very point. And maybe that leads us to question two, which is um, what taking it from another angle rather than looking at gaps. What does good look like in hotels? Uh, day to day commercial execution. So that when that AI element comes into play, we know that it can meaningfully add value on top um, we've touched on data, we've touched on people, we've touched on problems, we've touched on solutions. What's your interpretation of okay, you are in the right place to be able to benefit from AI, however that may be interpreted.

Speaker B: So I think that if your team is already in a place where you can kind of confidently identify where your processes maybe are either falling short or just, you know, could be better, like they're as good as they're going to get, but there's a lot of room for growth that you're already in a good spot because you have kind of defined processes. You can at least measure them to the extent that, you know, if they're not working or not delivering. Um, so you're able to kind of assess results. So that, that's a really good starting point and I think that actually feeds kind of question one about it's like a precondition for it. Right. Um, and then the other important thing is not just the implementation of this new technology, but what your team is going to look like in terms of ownership after the fact. Because this will, it will change processes, it's going to change the way your team works, hopefully for the better. If we're using the right solution and it's, you know, it's giving us, we need, um, so what, what is the plan? Right, that's, that's a new thing because it, it will change the way you work, it may change your team structure. Um, but what, what will that look like and who's going to own that after?

Speaker A: Mhm.

Speaker B: Right. We're day two. We have our new solution. Um, that, that that's part of the plan and that you're ready and everybody is bought in.

Speaker A: Yep.

Speaker B: And that that transition can happen to support these changes that you're making.

Speaker A: It's interesting you say that particularly that phrase about people being bought in because that requires a culture to be open to change, which by nature we're not always very good at changing, particularly if we feel there's a threat which everything associated with AI and jobs seems to be AI jobs threat. Uh, so people can be a little bit nervous about it. How much do you think the, how much do you think being good or in a position of good, if we use that as a starting point, is a balance between systems and um, processes and culture and leadership. And is it like a chicken and egg?

Speaker B: Yeah, 100%. I think it starts with leadership which defines the culture and then leadership and culture therefore are going to kind of create that, that framework around the processes and the expectations and how people do their work. So that's where it starts 100% is with leadership in the culture that they create. And, uh, you make a really good point. Right. There is some kind of natural fear around AI. I mean, we're certainly seeing the upheaval and the changes on the tech side as well. You know, the people that are making all these systems and softwares, um, it's very disruptive. So that's only natural. Um, but I think what people will find when they're identifying these problems is not that it's putting me out of a job, but as they're actually taking stock of what they can improve or what they can fix.

Speaker A: I understand what you're getting.

Speaker B: Yeah. And looking at that ROI and that value creation that it's, I mean, that's a little cheesy and I feel like probably off repeated at this point, but it just, it makes you better at uh, your job and it also lets you spend more time on the areas that the I can't, can't touch. Right. And that's where your value is, should be in the first place.

Speaker A: It's interesting, I was reading an article last week and most people know about the Pareto principle, you know, the 80, the 80, 20, like 20% of your customers deliver 80% of your business, for instance, and someone was, um, positioning it in a slightly different way and he said, right, well, if AI could do 80% of your job, what would you be doing with the other 20%? And I think that's quite interesting because when you were talking about really identifying the problems, sometimes we also look at it from how can it solve a problem or what can it do for our operations? And to your point about looking at how the roles will change, how the culture will change, how our value will change, how are we valued as employees? It's like, well, if AI was doing 80%, what are we doing with that other 20%? And how is that valued within the operation? So it does throw up a lot of questions.

Speaker B: Yeah, yeah, no, it definitely does. And that actually, that gets to, I think one of the long standing, I don't know if I'll call it a gap or a pitfall or. But just things that you see, um, in terms of being ready to really incorporate AI is are you ready for the outputs that can be generated? And it's funny to me because this is a long standing thing. I think before everybody was talking about AI, back when we were only the more sophisticated applications with the optimizations and the recommendations in that perception of the black box.

Speaker A: Mhm.

Speaker B: Right. Is, is the readiness to adopt the output. With LLMs, it's, it's a lot more straightforward. They hallucinate things. Um, but I think people feel more equipped and comfortable to, to flag it and call it out but still be supportive of its use.

Speaker A: Yep.

Speaker B: Um, so that, that to me is still a really interesting gap that is definitely changing. But that's another part from the leadership perspective and culture on down is like we're, we're experimenting. We're also going to trust it and you know, we have this level of buy in to where if it challenges how we've done things or what we have historically thought was the correct approach, we're not going to dismiss it immediately.

Speaker A: Yeah, that makes a lot of sense. Let's uh, just taking a slightly different angle because we're coming at these questions from sort of different directions even though they drive often at the same point. So if a hotel came to you today, so you know, working on that basis of we, we know we spoke in the first question about having the gaps. We've now spoken a little bit about what good might look like. Say uh, uh, you know, you guys are in contact with a hotel group and you're having these conversations and they kind of feel that they might be ready but they ask you for a readiness assessment for want of a phrase. Um, what would be the actual. What do you guys do in terms of evaluating and analyzing? And how do you take somebody through a conversation if you're kind of seeing that their reality doesn't quite match where reality needs to be? And does that necessarily mean that it's, it's dead in the water? Or how do you, how do you take somebody through that, through that process of engagement and improvement?

Speaker B: Right. So it, this all ties together. That's the funny part. Right. I know we're coming from a different angle, but it gets right back to that idea of knowing what you're trying to solve, knowing what's the proper way to solve it, um, and then understanding what's the highest value to solve first. So obviously we're, you know, we're a revenue management commercial optimization platform. We're very clear in what we're here to solve, which is using very sophisticated deep machine learning for optimization purposes. Right. So we're not, we're not building rules, we're not just automating things. Um, we're really putting in place a system that will, will learn and adapt frequently, um, and really personalize to your hotel's performance. So we walk them through you know, exactly. This is what we do. This is why, this is how we approach it. And um, then just to understand their needs from their, from their perspective, because you do sometimes run into usually a smaller operator potentially, um, that they just want to put on a minju when they hit 80% or you know, they, they don't, they're not looking for that level of sophistication and capability. Um, and they're also not comfortable adopting it. They, you know, they're, they're looking for something different, a little more simplistic, more rules based, et cetera. Um, so that's, but to me, that's not really an AI readiness issue. That's more system specific. Right. So it, that comes back to your

Speaker A: point about trusting like that black box thing, isn't it? And if, if you're, if you're letting something go, like a automated driving car where some of us still like to have of it a little bit.

Speaker B: Yeah, well, and it's, it's funny, I, I shouldn't fully contextualize that by any means. Um, because I think, you know, it's really important to just keep in mind that like autonomous revenue management or autonomous AI in general doesn't mean just fully handing over control, but it should mean that you don't need to have routine oversight.

Speaker A: That's a good point. Yeah, yeah, yeah, I see what you mean.

Speaker B: There's that, there's that trust, there's that buy in, you know, that it's doing a better job than you could because there's way too much data to crunch.

Speaker A: Mhm.

Speaker B: You know, um, and we, and you accept that and the performance is there, you see it. Um, but our system, or any other kind of system, again, just really, the readiness piece again, really just gets back to do you know what you're trying to solve? Can you measure it? And what value would you get by fixing it?

Speaker A: And that leads us on to the final question I want to ask you, which is kind of separating real AI from rebranded software. Um, when we were having another conversation the other week with one of my partners, I said it feels a bit like protein. It's like suddenly everything has protein in it now. It means they can charge a bit more for it. Even your cereal now comes with protein or whatever it may be. So there's this for me, I always think I strike this balance between what is, okay, we have, we are an AI company as opposed to actually a real AI, technological, technological solution. So, uh, how, I guess, how do you help somebody to know or how would you advise somebody to know whether what they're looking at genuinely creates value. Uh, and, um, where in the guest journey, AI is actually moving the needle rather than it just being, you know, oh, we are AI enabled or we are an AI company that you see in a bit of marketing or in a pitch deck.

Speaker B: Yeah, yeah. The question of the day, I think for everyone right now, um, I mean, everybody's going to say AI, but the reason is twofold. Right. One, because everybody wants to be perceived as AI, but then also because there is to a large extent in a lot of the market, again, not that depth of understanding or sophistication in AI and its applications.

Speaker A: So back to the point you were making at the, uh, first in the, in the answer to the first question, the, the kind of variety of applications and understanding them.

Speaker B: Yeah, right, yeah. No one's going to say like, we're a reinforcement learning company because that, you know, that may speak to somebody that's a data scientist or somebody that's in that space, but to, you know, your, your average software buyer. Yeah, okay. I don't know.

Speaker A: Yeah, Let me go to Chat GPT and ask what that is.

Speaker B: Exactly. Right. And that. Okay, thank you. That's a, that's a perfect bridge. Thank you. Um, and it's funny because I think that that is language models like Chat GPT or Claude or you know, perplexity, like whatever you want to use. I think that's where this is showing up most clearly right now in a lot of software is bringing in that type of functionality and kind of this perception, both intentionally from the company side, but then also maybe from the consumer side that like, oh, that's AI and that can do all these things that I need AI to do that we already discussed, need different types of AI, because a general language model, it knows a lot. Right. We all talk with those every day. But it learned it from texts on the Internet up to a certain point in time. Right. It was heavily and widely trained, but it was trained on stuff on the Internet.

Speaker A: Yeah.

Speaker B: Um, so when, when you're using it for data exploration or help writing queries, pulling data together, just asking things, discovery, it's fantastic. And that's what it's for. Um, but if you're asking it for strategy recommendations, for example, or, you know, how's my pace? Okay, great. What should I do with my rates? What's my rate? Strategy? It's just generating an answer from patterns it was trained on from other texts. Right. Applying it to whatever numbers it's getting out of your system at that moment. It's not. It's not learning. It doesn't take that into consideration next time. It doesn't incorporate it into any cycle and have that show up later when it's trying to generate more output. Um, and that's, again, that's just a really, really important thing for people to remember when we're looking at these AI companies or. AI.

Speaker A: Yeah, that's a really. That's a really. Yes. Clear way. It's. You see, a lot of people have built AI into their interface to interact with the data that you have in the system and maybe provide a summary or something to that effect. But the degree to which it's actually behind the system driving, you know, it's almost like a. It's. It's almost like your dashboard telling you that your tyres are running a bit flat or the engine might be running a bit hot, rather than it actually being the engine and the drive shaft and the suspension itself that is actually moving. Moving the vehicle. Exactly.

Speaker B: And I don't want to diminish the huge value.

Speaker A: No, it's. It's just differentiating between it, isn't it? Yeah, yeah, yeah. Which is the question we were asking. How do you separate the two out?

Speaker B: Yeah, it doesn't change what's underneath. Right. If they. If you don't already have those type of engines and models that you've been building and refining and compounding over time, because that's how those things improve that, then you don't have that underpinning any of the output, just because you can now chat with your system. And. And, uh, to me, that's one that I see coming up quite a bit. I think with. Especially with some of the consolidation, we're seeing people wanting to earn or, uh, earn. Wanting to own more of just kind of the full workspace across the tech stack. Again, it may be easier to do some of those things across, but it doesn't change the fundamentals of what is underneath and where those pieces are coming from. So those type of capabilities, you know, we're still. I actually really liked your analogy at the beginning. So it's kind of. You have a leak, you have a electrical issue, whatever. Sometimes you may just need a handyman that can, like, rewire a lamp and, you know, fix a leaky faucet, but sometimes you need a master plumber and, you know, a very specialized electrician, because you're talking about, you know, the entire wiring of your house or all of the plumbing and the foundation. Those are very different things.

Speaker A: Yes. Yeah. No, I think. I think that's a, I think that's a very, a very good point and worth emphasizing as you say that, yes, nobody's diminishing any way that AI is being utilized or interpreted. But back to your point, the understanding what you want it to do, um, and understanding why you're looking to bring this into your hotel or your enterprise, there's a difference between whether you want to look at what it's doing on the outside or whether you actually want to lift the hood and see what's going on underneath. And that's where you will maybe see the separation of the application of AI into the solution itself. And then that might help you better understand how that aligns with the problem or the efficiencies you're looking to gain.

Speaker B: Absolutely. That again comes your leadership, your culture. But everybody just go, I don't know, take a quick course on maven or something. Just, you know, baseline familiarity with AI and the types of AI and the applications will do a whole lot. Um, in terms of your. Just your success in making sure that you're building the right stack or if you're. Whether you best breed all in one, whatever, know why you're making that choice.

Speaker A: Yep. Yeah. It doesn't matter which choice you make. That's your choice. But understand why you're making it.

Speaker B: Exactly. Because there's trade offs. Right. There's pros and cons of both. I'm biased. Um, obviously, I think best of breath the way to go. Um, but yeah, you should understand how those platforms are solving the problems that you need to solve and choose based on that and understand why you're picking that solution.

Speaker A: That sounds like a perfect way to end this conversation and a great way of summarizing where maybe the starting point should be. Which takes us right back to where you started at question one. So thank you so much for your time. It's great to talk with you.

Speaker B: Thank you. You as well.

Speaker A: Thank you for listening and I hope the conversation was of interest. Please do feel free to suggest any topics you would like discussed in the future. We're always open to ideas. Thank you once more and I hope you will join us again. Goodbye for now.