Untangling Tourism Tech — Podcast
30 January 2026
Tourism operators often feel like they are swimming in an ocean of change when it comes to technology, but 2026 marks a specific turning point in how we use artificial intelligence. In this episode, Liz and Fab walk through the transition from static AI models to active digital agents: tools that do not just talk but actually get things done.
If you have been feeling overwhelmed by the constant updates to ChatGPT or Gemini, this episode brings much needed clarity. We move past the era of generic chatbots and into the world of digital doers that can help your destination stay relevant, sustainable, and accurate.
The first thing to understand about the AI landscape in 2026 is the difference between an LLM and an Agent. A Large Language Model (LLM), like the standard version of ChatGPT, is essentially a static brain: it contains a massive database of information and waits for someone to ask a question.
An AI Agent is the body: it is a doer that has a goal, a memory, and a specific set of rules to follow. While the brain knows things, the body does things. This shift means AI is moving from a research tool to a functional assistant that can follow a playbook you create for your business.
We have all seen the damage that outdated AI information can do to a brand. Fab shares the example of the Paperbark Walk in Agnes Water, which was closed for several years. Because old AI models relied on web scrapes from years ago, they would frequently tell tourists the walk was open and “amazing,” leading to frustrated visitors and negative sentiment.
AI Agents solve this by connecting to live, structured data sources. Instead of just guessing based on the internet, an Agent can check a specific text file or a volunteer’s update to verify if a location is open today. It provides accuracy that generic chatbots simply cannot match.
A new term every Destination Marketing Organisation (DMO) needs to learn is agent legibility. In the past, we focused on making websites look pretty for humans. Now, the priority is making your local knowledge machine readable so that AI Agents can find it and use it.
This does not require a massive, expensive IT transformation. It involves:
The absolute advantage for regional tourism is the passion and local knowledge of the people on the ground. AI cannot create lived experience, it can only learn it.
Liz and Fab suggest that the race is now on to harvest the local tips and stories from your visitor centre staff and volunteers. Whether you record their voices or take raw notes, getting that information into a digital format ensures your destination’s unique “soul” is represented in the AI answers of the future.
Google has made massive strides in integrating AI into its platforms. Through Google Workspace, AI can now access your business documents and emails to help run your back office more efficiently.
Fab notes that this is not a technology for the distant future: these are pre-built packages ready to be used now. By adopting these tools, tourism operators can spend less time on administration and more time on the human side of their business.
The core message of this episode is simple: do not wait for the technology to be perfectly understood before you start. The AI landscape will continue to change every few months, but starting the process of digitising your knowledge now makes it much easier to adapt later. Take small steps, test the tools, and start building your partnership with AI today.
Tourism Tribe members get access to our regular tech sessions where we dive deep into practical demos of these tools. If you want support, clarity, and a community that is using this technology to grow their visitor economy, check out our memberships and join us.
An Agent is an AI tool that has a specific goal and a set of instructions. Unlike a chatbot that just talks, an Agent can use tools, access specific databases, and follow a playbook to get a job done.
While developers can help with complex builds, many Agent frameworks are becoming “plug and play.” You can often grab pre-built Agents from platforms like Google and customise them with your own data.
Join the club, I don’t know anyone who does. To push past this aversion, start by filming what you love in your region or add a voiceover if showing your face is too scary for now. But over time, showing up as the human behind the business will always connect better.
Google now provides AI summaries in search results, which often give visitors the answer they need without them ever having to click on your website. This is why making your data “legible” to these Agents is more important than ever.
No, it is often more about the time spent organising your information than a large financial investment. Capturing local knowledge in a digital format is a low cost way to start.
AI can handle the information, but it cannot replace the lived experience and human passion that makes a destination special. The goal is to use AI to handle the heavy lifting so humans can focus on the visitor experience.
Untangling Tourism Tech Episode 17, with Fabienne Wintle and Liz Ward. Each line shows who is speaking and when. Times match the video. We removed filler words and fixed mis-heard names.
[00:01] Liz Ward: Hey Fab, how are you? Happy New Year.
[00:05] Fabienne Wintle: Thank you, Liz. So glad to talk to you. I love our brainstorming sessions, especially preparing for our podcast episodes.
[00:13] Liz Ward: This is exciting. Our first episode of Untangling Tourism Tech podcast for 2026, and we’ve got a really punchy topic. This is Episode 17. Welcome to our listeners. A big warm welcome to our Tourism Tribe community, and in particular our Tourism Tribe members. We have, as Fab said, we’ve been brainstorming this topic. We’ve been spending the quieter time over December and January to really sharpen up our knowledge again, layering up on what is happening in the world of AI and digital technology for the tourism industry. Fab, you’ve been doing some awesome research and experimentation, and we’re going to unpack that into a quick story for our listeners today, so that we can bring some clarity to what this ocean of change is that’s going on in the tourism travel marketplace because of AI.
[01:15] So, let’s get started. Let’s start with a little bit of potted history on what’s been going on since 2023 and where we’re at now in ’26. And then we’ll get into an explanation about what is going to be the most relevant technology for regional tourism to remain relevant and sustainable, really sort of hold that important role that they have and add value in their visitor economy. So 2023, we were just starting to understand what AI was, weren’t we?
[01:51] Fabienne Wintle: That’s unbelievable, was it, then? Yeah, ChatGPT was released by OpenAI, and then everyone was talking about chatbots and opportunities for tourism. It was, look, it was the starting point. The chatbots could write a generic itinerary, but it really lacked soul, and we were like, “Oh, well, we all still have a job, don’t we?”
[02:16] Liz Ward: That’s right. And talk… and it still does. Like, we use ChatGPT and we go, “Wow, it can get it so wrong so often.” And, you know, we talk about training it and how that improves it, but we’ve seen some major shifts over that time. What would you say was kind of the big shift in ’24?
[02:37] Fabienne Wintle: Well, I guess we realised that there was some sort of collapse in the model. AI started training on the other AI content. So the chats were just rehashing what other models were saying, and the industry started to realise that there was really a lack of human-led knowledge in that AI content. Basically, it was kind of an autocomplete, as you said, that could be wrong. Something needed to change.
[03:11] And obviously a lot of us, or a lot of people maybe that are not tech-savvy, just saw that, “Oh, we can just chat to it.” But in the background, what was happening were just huge developments in terms of creating platforms, creating template, creating framework, creating code, etc. And we saw that change in Google when really… and we saw, I think operators would have seen, “Oh wow, okay, I don’t just have to go to ChatGPT. Now this information is integrated in the Google search results. That’s going to have an impact on my analytics, on my results.”
[03:56] Liz Ward: I saw a massive shift. 2025 for me was the year that Google woke up and went, “Oh gosh, we’ve got to get into the AI space.” And we started to see a whole lot of things change in our Google platforms. Like, the most obvious ones were the changes in search, where we saw AI summaries start to appear, and that has had a massive impact. Obviously, we know that destination marketing organisation websites, a bit like publisher websites, they have really taken a hit in terms of the visitation to their website, because the AI summary is performing the role of an agent and bringing back a really neat summary that means that the user doesn’t need to go and click on the website. It’s bringing it all together.
[04:48] So that’s been a massive shift, 2025, and really brought into question for a lot of our DMO clients, you know, “What do I do about it? And what, you know, what does this mean for our website? Do we need to restrategise what the role of our website is?” But very interestingly also is what’s going on under the hood, is what Google has enabled for development of AI tools. And since ’23, we’ve seen over 29% year-on-year growth in development in AI. And travel is one of the leading industries in that space, because there is so much opportunity in it, because it is so difficult and time-consuming, and you get very mixed information when planning a trip. So some of the things you’ve been playing with, Fab, tell us about your exploration into the Google Workspace area of AI and also the development playground that you have been in.
[05:55] Fabienne Wintle: Yeah. Well, I guess since… I’ve just been waiting. So, as most of you, we would have all been, you know, used to being a ChatGPT user, but I’m like, I’ve got Google Workspace. Google knows everything about my email, my documents. At some stage, they’re going to just take over, because they will provide me with a smart language model brain to access all the information that I have in my Google account. For example, anyone could say… you’ve had a client for many, many years. You could say, “Give me all the information about Mary. What shall I be doing with Mary based on what you read?” So Google can access much more than just travel data. It can access all the business back office. So it’s got a huge advantage there.
[06:43] So I’m really excited about that part about businesses being able to be so much more efficient, not just about marketing but about running the business leveraging Google. So that piece excited me. Then I was really excited about the shift into Google AI Mode. So if you participated in our fortnightly webinars, you would have seen all the demos. Google… I’m still wondering when Google’s going to shift from the Google standard search results page to the AI Mode, which is basically the ability to have a conversation with Google and AI Mode. And this is where your Google Business Profile is so relevant, because this is Google’s golden goose about tourism.
[07:27] And as we’ll see, especially with agents, in order to have such a great level of conversation, the agents need to be able to access data that have been structured in a very specific manner. So having that database of listings of travel businesses that are written in a machine format, you know, the text boxes where you got to find the information, gives it a superb advantage to be able to return the right information.
[07:54] But what I’ve been playing with lately is agents. So, you know, some of our listeners might know that I’ve got a background in IT and coding, etc., but I much prefer conducting all the different parts of IT to make something relevant for the tourism industry. But I’ve still got that part of my brain that loves playing with the tools, and I need to understand how things work together to form the product of an agent. So, first…
[08:23] Liz Ward: Fab, can I stop you there? I just think for our listeners, it’s really important. There’s been a lot of new language we’ve had to learn, including the term LLM. What are we talking about here?
[08:34] Fabienne Wintle: Okay. So an LLM is called a large language model, and it’s ChatGPT, Gemini. So it’s the brain, right? It’s the content that exists on ChatGPT plus ChatGPT’s engine to be able to respond to a query. It waits for someone to ask a question. It knows a lot, but it can’t do anything without you being there and asking the question. The difference… so it’s kind of static until someone comes and interrogates it. The difference is the agent is the body. It’s a doer. It’s a tool. It’s got a memory and a goal, and it can interrogate that content, but it can follow rules that you can give it.
[09:21] So the LLM is a big database that might have some very basic sets of rules. You know, if someone says on Google, where shall I get information… Google might have said, oh well, where can I find accommodation in Agnes Water? Google might say, “First interrogate Google Business listings and then look at such and such websites.” But with an agent, you can really hone in on exactly giving it a set of instructions or a playbook, so it can go in a specific direction. It’s got a memory, so if someone comes back to that agent, saying… and they’ve used the same WhatsApp number to talk to them, they can continue the conversation as well. So this is where it’s at.
[10:07] Liz Ward: So for 2026, yeah, this is where it’s at. And this is what I think the clarity is for me, is while we’ve been hearing about agentification, this is coming. We might have had a play in agent mode on ChatGPT to go and try booking a restaurant, or watch some videos on YouTube, people doing their online shopping, all sorts of things using an agent.
[10:29] This goes so much more deeply though, and this is the answer to me for regional tourism, is to look at this opportunity to be able to bring the best, the very best of their source of truth information, and get it into the hands… and not just theirs, but curate the very best source of truth information and the local tips, all that. If you think about the knowledge that sits in the heads of visitor information centre staff and volunteers, you think about the passion that sits in there. If you can curate that and make it available to your visitor wherever they are, 24/7. Now the technology to do that, yes, is agents. That is the solution. So that’s the line.
[11:25] Fabienne Wintle: And a key difference. So the agent might use the large language model, let’s say the database of content that has been literally scraped from the internet and put in a big database. They might use that as one source of information. But you are able to instruct the agent to use other sources of information. So if you’re a visitor centre and, let’s say, you have got a lot of volunteers that use “info at Mary Street Visitor Centre dot com”, right? You could actually instruct the agent to go into Google, Office 365 emails and also search for information in there as well. So the agent doesn’t just rely on the large language model. You can connect it to lots of sources. So, as I said, anything in the back end like your emails, but you can create your own database of content from all the knowledge of the visitor centre staff.
[12:26] So you could actually go and film the visitor centre staff, or just take notes, or get them to record themselves. It doesn’t matter, because that knowledge needs to be in a machine-readable format. And I think that’s a big change for how tourism businesses and tourism organisation think about AI. It’s not just about the pretty things on the website. There’s really a race now to digitise the knowledge that might be in people’s head, but it doesn’t have to be pretty, because all that the agent needs is that information to be structured in a machine learning format.
[13:04] Liz Ward: And the problem that this solves, that we see at the moment, is that we’ve got the same content replicated, or the same story or information told in a slightly different way, across many websites about a destination, their experiences, their tours. So you might have a regional area, and that same content is on the Australian Tourism Data Warehouse database. It’s on the local government association’s website. Some of it’s sitting inside Expedia. It’s just the same story again and again. So the LLM can’t learn anything new.
[13:40] And the point of difference, the absolute advantage for regional tourism and those ecosystems that exist there out in the regions, is that lived experience knowledge, because AI can’t create that. It’s got to learn that. And the only way it’s going to learn that is if that information is digitised. So it’s a huge advantage for DMOs to harvest that and to find efficient ways, which of course we can teach them, to keep updating it.
[14:15] And the thing I think is interesting, Fab, is we might think of agents, as I said, about, you know, run and do your restaurant booking for you, and all of that’s possible. But this is a below the surface… This is like sitting under the… this is the iceberg underneath the water, that this technology is not difficult. A lot of it depends on, as you say, structuring and digitising the information, which isn’t hard, and then, you know, making it available for any agent to be able to find or deploy it through… what would be some mechanisms? Through a WhatsApp message or a concierge chatbot on your website, or just available in a private way for your staff to use.
[15:06] Fabienne Wintle: And there’s where we’re at now, is agents talking to other agents, right? Agents are building agents. This is not like 2030 stuff. This is what’s happening now. So there’s a protocol for agents to talk to one another. And when you develop your agent, or shall I say when you go to Google’s Agent Garden or similar from other systems, you don’t have to build an agent from scratch. They’re there. You can already grab this agent, ask AI to improve this agent with such and such, connect another database, and bam, you’re there.
[15:41] So if we were thinking about the case of… in Agnes Water, we have a beautiful place called the Paperbark Walk, right, that’s been shut for about three years, if I remember. At the very start, when there was just the large language models, right, this Paperbark Walk, because no one had updated their website and Google had crawled stuff that was quite old, would return to… it’s a very popular Instagrammable location. So Google would say, “Yeah, no, go there. It’s amazing.” And then there’s all these people rocking up and it’s shut, and it’s really bad negative sentiment.
[16:12] Now that we’ve got an agent, if that was happening now, we could, you know, if it was shut yesterday and there was an agent, we could have connected another source of information, which might have just been a text file saying, “This place is shut. Do not send people there.” So instead of just relying on Google, Gemini or ChatGPT crawling the internet for an update, we could have been very quick at sorting that problem.
[16:41] So I guess that’s sort of the difference that the agent opportunity brings, is not just going into the LLM, but connecting lots and lots of resources that were available before via API, but we can talk to it like a human without having to program everything. And it can read that this beach has got pets, or there’s a photo with this person on it, therefore the beach is pet-friendly, and I’ve checked on Google Maps, it’s open. Return it to you now.
[17:08] Liz Ward: So a real headline out of this story that we’ve told today in this podcast is about agent legibility. This is a new term we’ve got to get our heads around, about, you know, building knowledge bases that are really up to date and easily accessible by agents. And that needs to be the focus moving forward for DMOs, for destination marketing organisations, and also for… and it’s achievable, right? It’s not… we’re not talking about a massive investment in AI transformation. Yes, it would take time, but the technology is accessible.
[17:51] Fabienne Wintle: Much easier than putting it on your blog post, you know, making it pretty, interviewing the staff. You can just put it in the database.
[17:59] Liz Ward: Just raw information. And I think that’s the marriage, isn’t it? That partnership with AI is, you know, you use the AI tools, find a developer, work with someone like us as the strategist to actually embed this AI strategy into your digital marketing strategy to maintain that relevance for your DMO. The tools aren’t that hard to use. Find the right developer, which we can provide guidance on. But we provide that framework so you can make sense of it and everybody can understand it. But I like this model very much. I think this framework is absolutely right, because it’s that partnership with AI and it’s the human focus as well, bringing it together, which is the missing piece.
[18:52] Fabienne Wintle: The tools are ready, sharpened. You know, the big ecosystems and framework like Google have got them ready, in a shop, free, available to developers. It’s literally plug and play. It reminds me a little bit when… before AI, when we’re talking about websites, and I’d be running workshops and I’d have web developers in the room saying, “Oh, I don’t want that WordPress stuff. I just build everything from scratch.” Yes, you can, of course, if that’s what you fancy doing, but you’re actually making it much more expensive for the person that invests in you to get a great positive outcome. And that’s where things are at now with agents. They are pre-built packages of cereals on the shelf, ready to go. You just need to find someone that understands your business and might be able to advise, “Go this way, because you’re going to be faster to go to market. It’s going to be more sustainable.”
[19:47] But one thing is sure, is just you can’t wait. I’ll say that every time. We can’t wait, Liz. I had to roll my sleeves and… I’m a geek at heart, so I love playing with that things. That’s amazing for me. But we got to jump at this opportunity. Yes, it might change. It will change two, three months down the track, but at least we’ve gone through the process of gathering that information and turning into knowledge on how we can use it for our business, our destination. So it’s much easier to adapt to the next step than, you know, just waiting and trying to be proactive. Then it’s just going to be an overwhelm. I mean, I’m already overwhelmed now.
[20:24] Liz Ward: But that’s what we’ve said all the way through for the last couple years with AI. Take small steps and progress, rather than waiting for it to be a perfectly understood opportunity. It’s just not going to happen. This is going to keep changing and keep moving, and we’ve got to implement what we can and test and see how it goes. And I think that this is the first opportunity for 2026, a marriage with an AI… with AI agent technology to bring to life that human-centred content.
The free Pocket Rocket app gives you a personal AI marketing coach, website audit, weekly action plans and 5-minute tips. Built for tourism operators.
Get access to all our training, coaching, and tools built for tourism operators who want to grow with confidence.
Explore membership
