Welcome to Travel Again presents the weekly travel Roundup, covering the headwinds and tailwinds impacting the business of travel. Please welcome our hosts, Mike McCormick and Ed Silver. Hello Mike, how you doing? Good, how are you doing, Ed? I’m doing good, Mike. Oh, stand by. You want to start over? No, let’s just keep rolling, man. The music loves me, so let’s just keep going. Sometimes the music just wants to keep going.
Mike, this is a special edition of our podcast today. Are you aware of that? I am aware of that. All right, this is called a Spotlight Edition where we focus on a specific product, person, company, innovation, or in this case today, we’re going to focus on some new research. Today we’ll be reviewing new research out from PhoCusWright entitled “Travel Innovation and Technology Trends 2025” report. What do you think of that?
I think it’s great. What I like about our Spotlight product is that it’s a nice way to have a balance between a full-length episode where you have to devote a little more time to something where we pick out something like, in this case, a terrific piece of research and really drill in on it a bit. It is something that we help you tell what it means, we give it context, and hopefully help you make better decisions out there in the marketplace. I think in this case, it is a perfect example of how the Spotlight product is really helpful for how we deliver media and deliver content. I’m excited to do more of these and certainly do more of them in partnership with PhoCusWright. Our new partnership is fantastic. Bringing their terrific research, insight, and events and putting it together with our advisory capabilities and helping companies put this intelligence into action is a great combo. Today is another way we can highlight that.
Well said, Mike. With that, let’s bust right into it. This research report was written by Michael Coletta. Michael is the senior manager of research and innovation at PhoCusWright. He has served in a variety of high-level roles spanning 25 years in the travel industry at startups, corporations, and digital agencies. At PhoCusWright, he oversees the company’s technology and innovation research efforts, recently concentrating on the impact of AI and digital identity on the business of travel. Mike, please welcome to the stage Mike Coletta.
Mike, nice to see you. Thanks for joining us. Thanks, yeah, thanks for having me and looking forward to working with you on the partnership. That’s great. Jumping right into it, tell us a little bit about this research you’ve created. I know a lot of time and effort goes into every piece of PhoCusWright research. One of the things, again for our partnership, is my past experience with PhoCusWright years ago. I’m really proud. I think the research that you guys do continues to be best-in-class, and this is another great example of that. So tell us a little bit about the research you created and a couple of the headlines. If you were to pull something out, what would be your headlines from your work that you’ve created?
Sure. It’s pretty much fully focused on the impact of generative AI, perhaps not surprisingly, but of autonomous AI agents in particular. I think everybody’s probably hearing a little bit about them these days. If you’re not familiar, this would be AI tools that can not just provide information to you but go out on the internet and take action, do things for you. It’s about that really specifically, as well as the convergence of generative AI and these agents with digital identity.
We did put this report out for free. A lot of our or most of our reports are for our subscribers, for our clients behind our paywall, but we wanted to put this one out there for everyone to read. We really thought it was important as a way for everyone to get up to speed. I tried to make this report insightful for people who are quite familiar with generative AI and digital identity already, but accessible for people who don’t know that much about it yet. There’s some history, reference links, and some research for context. What I tried to do is just really explain why this all matters so much, maybe for anyone who’s still skeptical. Just really quick, to read it, you can go to phocuswright.com under research at the top, just click free research insights and it’s right there.
Mike, I’m also going to make this research available at our website as well with the permission of PhoCusWright, so either PhoCusWright or traveladvisory.com to make it easy to find.
Thank you. So the headline as a synopsis: we covered basically five trends in the report that we think are the key areas to watch this year and beyond. Four of them are the specific ways that generative AI and AI agents are going to impact the industry. We carve those up as the impact on travel company operations, and then the in-destination experience, travel distribution, and travel marketing. We really focused on the ways that generative AI and agents may transform those areas. Then the fifth trend essentially is how generative AI and agents will converge with digital identity.
The first thing to start with is that agents, these autonomous agents, they’ve been around a while; they just never worked very well. But they’ve become quite capable due to the latest breakthroughs in the models with computer vision and reasoning abilities in particular. So they can understand what’s on the screen, how to navigate the internet, how to break tasks down into goals and accomplish them.
An important milestone, I think, is good for context. I talked about this milestone in the report: OpenAI’s latest model, o3, scored an 87.5 on the ARC-AGI benchmark. That’s a metric that tests its reasoning and problem-solving capabilities. 85 was the pass, so it passed. I think that’s a big milestone as far as explaining why this is all relevant right now. We’re crossing that threshold. As your audience is aware, OpenAI launched Operator in January. With all these launches, travel is always one of the use cases that they provide. The other day they launched their developer tools—this was yesterday as of when we’re recording—for people to build agents on top of their technology. So now is the time that we have capable agentic AI tools in people’s hands. We really will be seeing a lot of activity this year. I’ll say, I’ve tested some of these tools; they’re still a little bit disappointing. Maybe they work well in other industries, but in travel, they just struggle with date pickers as they’re trying to navigate websites.
For example, we saw Operator working with launch partners like booking.com and Tripadvisor. They’re obviously working through these issues. We’re almost there. Even the new one that was announced the other day and got some attention, Manus out of China, from what I’ve seen it struggles with travel as well. I do post a lot about my experiments with a lot of the stuff on LinkedIn if anyone wants to follow there.
There is so much nuance, as you know. Being in the industry, there’s an incredible amount of nuance in travel both in terms of everything in the intent planning and the inputs in, but also very much what’s behind the scenes to execute ultimately a travel booking. Still to this day, it’s a spaghetti maze of interactions and dependencies that make it very difficult. I think certainly one of the most challenging industries, but at the same time, what’s exciting is that it also touches virtually everybody. Travel is the place where everybody wants to play ultimately. The industry people want to be in, the investment that gets poured into it still… but I think a lot of times from an outside perspective people are like, “Oh why is the industry not moving faster? Why does this not…” and the realities are it’s because it’s highly complicated behind the scenes.
When you were doing this research, were there surprises? Was there anything that jumped out at you that you just didn’t expect or didn’t expect to hear? Some of it is probably for us in the industry as expected, but any surprises in there?
Well, this report is what I’d call desk research. It wasn’t based on surveys or market sizing like a lot of our core research is at PhoCusWright. It did reference some of our research that we’ve done recently. In particular, last year we did a technology study with consumers and we surveyed consumers on various cutting-edge technologies. One finding that I think was really significant was that regarding generative AI, we found 39% of all travelers in the US used it for, in general… active travelers who took a trip in the past year, 39% used generative AI in some form. That could have been for entertainment or food or health. But when compared against other kind of measurable technologies like virtual reality and augmented reality, it officially leapfrogged those technologies which have been around a lot longer. That’s kind of a milestone. It just points to how heavily it’s being used and how fast it’s growing.
Just for context, the overall rate of use for travel specifically was 18%, so not mind-blowing quite yet. But we’re going to keep measuring that and I’m sure it’s going to continue to grow rapidly because travelers also told us they’re very happy with the results for pretty much all travel use cases that they’re using it for.
Planning on a personal level? On a personal level. My family uses it in the planning. I’m going with my family on spring break. When we’re there, it’s a place we haven’t been before, what can we do? Give me ideas, give me a sample itinerary. It’s like that next level of not just the fundamentals of “we got to get flights” or “where I go,” but a lot of it is those kinds of things where it’s even just to have a starting point. Now, again, this will all evolve as that matches up with intent and the tools get better. It is interesting and it’s really rewarding. It does in a way, it’s like the reason why you’d go personally to a really good travel agent that you’ve used over the years, because they would know all the places to go and things to do. But that industry can’t support on a broad scale that level of expertise being out there, so maybe this takes us there in a whole new way. But it is interesting.
Why do you think every time something new comes out like Operator, why do you think travel is always one of the first use cases shown? I was always curious because there’s plenty of other shopping methodologies out there. Why is travel always way up there in the use case?
Exactly what you talked about: it’s complicated but it’s fun, I guess to put it really simply. I think everyone sees it as a challenge, a good challenge that it still struggles with really. You have your hallucinations in there. It struggles with space and time. If you have an itinerary and something gets delayed and everything else needs to move, I think it would struggle with that. But it’s getting better all the time and I think we’ll get there. I think it’s just a matter of a fun challenge that all these companies seem to want to tackle. Everybody does it, so it’s popular. Everybody shops, everybody needs health advice, but travel I guess is the more fun one.
Just what you were saying, kind of it acts kind of like a travel agent to jump ahead a little bit, I do see it going more and more in that direction. I think our interactions with computers will evolve and particularly with these agents. Ideally for me, I would be having a verbal conversation with my agent like I would have with a real travel agent. I shouldn’t have to type everything; I’m sick of typing for sure. We type too much. So just talk, and I can go out and research either in real time or kind of go do research and come back, but it should be able to bring me what I need to know, what I need to see—prices, photos, videos, maps.
I think dynamic UI will be a big deal and one of the next shifts that we’ll see. It’s generative AI. It should be able to generate content in whatever format works best for me to understand it and make decisions for travel, but really for everything that I’m trying to make a decision about.
I was leaning towards that. That was the next place I wanted to go—a little about where this is all going. As you were saying that, it actually triggered something for me because when I use generative AI in other ways, not for travel but just for other means, one of the things I find is I get more of the prompting of, “Hey you just asked me for this, do you want me to create this? Or do you want me to go here or here? Here are some other questions that you might want to ask” kind of thing. With the travel applications I’ve seen so far, there’s not a lot of that. It’s just kind of like I ask a question, I get an answer. It doesn’t always seem to know or at least very often very little of getting more of that, “Hey now that you asked me about getting the best ways to get to this remote destination, would you like to have… what more would you… anticipating? Well do you already have a place to stay? Where do you want to go? Are you going to do some other things while you’re there? Is this over a weekend?”
I have to feed that information in to get that answer. I think with travel that’s one of those things too where travel is very iterative. You go, you do a lot of… it’s not linear or just with like a few forks in the road. A lot of times in the travel planning process, especially for leisure or personal use, you could have a lot of different ways you could go that might change your whole thought about the trip you’re taking. I think that’s also harder to… it’s harder, to your point, it’s more complicated. But I just see that was one area to me that really jumped out that was like, “Ah, you know just, huh, we’re not there yet.” And that’ll help the adoption. But other things you’re seeing like next as this rolls out? Things are moving quickly in some ways, but as with travel, not always the fastest. What do you see in the next couple years? Anything jump out at you in terms of where this will go or more tangible things we’ll see in the experience with travel?
Just on that last point, that’s the threshold that we’re crossing. We’re going from it spitting information at you to becoming more of a conversation and potentially verbally, not just as you said, not just typing typing typing.
Another interesting finding I wanted to share was that 40% of travelers told us that they are interested in booking using generative AI. It’s really not here yet, but to me that’s a significant number. They are interested in doing it as it becomes available, whether that’s inside the AI platforms or using these autonomous agents. That interest is really there.
You’re asking about what else? Just what else is on the radar for the next few years? Yeah, what else do you see looking ahead? It will on an IT basis keep changing and evolving. Any other big milestones or things you see ahead that we should be looking out for? Any other trends we should be paying attention to here related to this?
Well, related to the five trends in the report: company operations, the in-destination experience, distribution, marketing, and then the convergence with digital identity. We could dive into any one of those. They’re all covered sort of high-level in the report; I just tried to hit the main points. But then we’re going to be doing a series of online events this year through the spring into the fall where we’ll be bringing in a panel of experts to really dive in and get some outside perspective. And those are free; these will be public webinars.
The first one is on the impact on company operations for good reason. I think it’s going to be one of the most impactful areas short-term, as it’s the most obvious use case for so many people that it seems to be where they start. Well for most travel businesses, yeah, they should be thinking about it.
I think agents in particular, and a couple of quotes I’ll share that are in the report, I just thought were really telling. Sam Altman, CEO of OpenAI, just said in January: “We believe in 2025 we may see the first AI agents join the workforce and materially change the output of companies.” And then another quote from Jensen Huang of Nvidia, I think was also in January: “In a lot of ways the IT department of every company is going to become the HR department of AI agents in the future.” And Salesforce is all in on agents and Google launched AgentSpace. So it’s just, I think, big questions for how these agents will be integrated sort of as digital coworkers and what that’ll mean for the companies that embrace them and figure out how to use them and increase their automation and their productivity versus those that don’t. So really looking forward to diving into that from people and hearing from people who are currently doing this in their own companies.
I’m really… what you just said about the quote about technology becoming the HR department for AI agents, it’s like, “Well will they start asking for time off, benefits, and…” No they won’t. That’s right. No they won’t.
Mike, do you think these tools are coming as a replacement for OTAs and meta search engines? Because they seem like parts of them could replace some of that, but are they going to replace all of it in time?
Certainly not. I think there will be a shift in the power dynamics. I guess the question is how much? I think all the current players will continue to provide value. There are ways that could sort of grow or diminish. We’ll be diving into this specifically in the online event where we discuss distribution—the impact on distribution—which will be in May.
The crux of it would be: you have these horizontal generic AI agents like from Google and OpenAI who you may start to trust as your daily assistant and doing a lot of daily tasks for you. Then you might ask them about travel, and I guess the question would be, are they going to just go to the source of travel, go to the suppliers, or would they go to intermediaries where they can get quick and easy answers? You can see already they seem to be favoring intermediaries where they can get quick and easy answers because that does make sense.
But in the long term, I think it’ll depend on the economics. Will the same model continue of these platforms monetizing on advertising and commissions where OTAs are strong, or are there other ways for them to monetize or be subsidized? Could travelers wind up paying a subscription? I’m not paying $200 a month right now for Operator, but I am paying $20 for ChatGPT. So if I’m paying upfront, kind of like an Amazon Prime model, how much do they need to monetize on the backend? Can that make them be more agnostic in terms of where they book rather than favoring say OTAs who are paying them?
I think that’s one question. Then, would I go to my Google or OpenAI assistant for travel, or would I go to Expedia or some startup that built something amazing? It’s just travel is complicated and I know it’ll work better. Google can personalize things for me probably given everything it knows about me. Google’s sort of in a unique position given the integration with flights and hotels, maps and everything; it really could become that much more of a really excellent travel planning and booking environment.
But these things should be able to say, “Here’s really the best place for you to book, here’s the best loyalty program for you, here are the best perks maybe that this supplier or this intermediary is offering on top of the base rate.” I really would want my agent to be agnostic, and so that’s a question.
Then meta search is really interesting. Focuswire just published an article the other day that I offered a quote for everyone to go read. The story was on the impact of agentic AI on meta search. What I said is they really have a very similar function to what could be handled by these agents in terms of searching for prices and passing the user to an OTA or supplier for the actual booking. So given that is currently meta searchers’ role, could they be that much better at facilitating that experience, an agentic experience? Or Operator, let’s say, can get those prices through an OTA, through the GDSs, or other channels, act as the meta search essentially, and then diminish meta search value. They’d have to get the data from somewhere. Maybe it would be from meta search players; maybe that involves an evolution of their business model. So definitely some questions there.
A lot of play. Yeah. I think it’s fun to speculate on these things and try to get visibility, but certainly nobody really knows. It’ll depend a lot on the traveler, I think, how they want to use these tools and how they’re monetized—the traveler’s willingness to pay or how much data they’re willing to give up, which is generally a lot as we’ve seen, but we’ll have to see how it plays out.
No, that’s great. Well, Michael, again, good great insight, terrific report. Well done, and I know a tremendous amount of work that goes into those. To get to the point you were able to get to and distill it down to what you did in terms of the big trends… it’s a good read for everybody, certainly in the industry, and appreciate your insight and taking the time with us today.
Of course, yeah, thank you again for having me and giving me the chance to talk a little bit about what we’re up to at PhoCusWright. Thanks for your time, Mike. Michael is the senior manager of research and innovation at PhoCusWright, and again thank you for your wisdom and time, Mike. We appreciate it. Thank you. Cheers.
All right, Mr. McCormick. Well, again great, really good. I really enjoyed the interview. I think it’s one of those where it’s hard because you really do have to… aside from reading and digesting the report, there’s really a lot to think about. He and his comment were really funny; it’s true. It’s like fun to speculate at this stage because it hasn’t… businesses are evolving and we’re trying to work towards it, but it hasn’t disrupted travel yet. But you can feel the changes coming on the horizon. Just that note about the role of technology in companies, the roles of how this is going to work, you start thinking about how evolved that can be.
But I’ll challenge too, because then it still comes back to companies—even the big OTAs, the big travel agencies, everybody in the space, flyers themselves—really kind of stepping back and saying fundamentally what are we really about? What are we really here to do? Tools aside, what value are we going to bring? An OTA brings choice, a third-party comparison. Well, okay, which AI agent am I using? Which trusted source? Those things are going to really matter to me as much as the technology itself, because if it’s slicker but if I don’t feel like I’m getting a real view of the market, then that credibility is lost. I don’t care how great the technology experience is.
Yeah, what’s really crazy is how fast these advancements are coming. He talked about Operator, which you can’t even get access to yet unless you’re paying one of the higher levels of fees. You see some demos, but very few people have actually hands-on used it. Then he talked a little bit about Manus that just came out this week, literally this week. Keeping up with these things is almost a full-time job.
It’s a great space, it’s fun, but it is a little challenging to keep up with. I’d encourage our listeners to try different tools and try their own use cases. Think about what you want to do from a travel perspective. Go to your favorite OTA or meta search, try it, but give it a whirl within one of the AI agents and see what happens. As a call to action though, Mike, I would suggest everyone should go read this report. There’s a ton of great detail in it. Mike and the team did a fantastic job. We will make it available on our website so people can read it since it is free.
Then if you want even deeper research that PhoCusWright has, go and subscribe to their Open Access or download a particular report. Some of them have individual fees. It is really great deep research that is available at PhoCusWright. And not for nothing, look if you read the research and if you’re thinking about AI, you want to put a framework together for your own initiatives, Travel Again is here to help you with that. We know travel and we know AI very deeply. We’re now partnered with PhoCusWright to get this level of research, so reach out and we will help you take this from research and data into action. That’s what we’re here to do, Mike.
Sounds good. And on our end, our part of the partnership and why we’re excited about the PhoCusWright partnership is that aside from everything you just said about the value they bring, then for us the advisory piece of saying we can help any company, any brand in space, or that’s trying to in some way sell into travel… we are here to help put a framework together for how to use the tools, how AI can impact your business competitively or otherwise. We do today help companies with a framework that we’ve developed to really help put that down and have a true product plan, a real approach to how to do this, how to integrate your business, and how to deal with the competitive pressures that this all brings. Again, it is a great way to showcase our relationship and why we formed it with each other because we really do complement each other as we get into especially areas like this that are so transformative for our industry.
All right, Mike. That is it for our special Spotlight edition of the podcast today. We have so many great guests for the rest of the season. I can’t wait to continue to release episodes throughout the rest of the season. That is it for our show. We will see you back here next time. Thanks, Mike. Thank you.