AI Search Visibility Is Rewriting the Hotel Booking Funnel with Kin Sio

Podcast Summary and Highlights

🏨 Meet Kin Sio

Kin Sio is the CEO of Lights On Digital, a hospitality commercial strategy firm that runs revenue and marketing operations for more than 40 independent and boutique hotels. He spent six years in product management at Microsoft, then helped build Coinbase’s international exchange, before acquiring Lights On Digital in early 2025. Most of his work this year centers on AI search visibility, testing whether ChatGPT, Perplexity, and Google AI Overviews actually name a property when a guest asks where to stay. He also hosts the Lights On Podcast and splits his time between Hawaii and Seattle.

🏗️ Where Hotels and Short-Term Rentals Actually Diverge

Kin’s world runs on scale. A single hotel can carry the same distribution complexity as an entire short-term rental portfolio, because one property has to sell tens or hundreds of rooms every single night. That’s why hotels lean on RevPAR (revenue per available room) as their central metric. It normalizes performance across a 30-room boutique and a 500-room property, and Kin sees larger STR operators starting to adopt the same benchmark. Worth checking your own numbers against, especially if you’re managing multiple units in one market and can’t tell whether a slow week is noise or a real problem.

💵 The Rate Parity Leak Nobody Notices

Ask Kin where most hotels bleed direct revenue, and the answer isn’t marketing. It’s price. He’s seen properties where the direct booking price on their own website was higher than what showed up on Expedia, purely because of fragmented systems between the property management platform, the channel manager, and the rate distribution layer. Guests notice instantly, and they book wherever it’s cheapest, usually not direct. Before spending another dollar on ads or content, Kin’s advice is to check that your rate is identical across every channel your property lists on. It’s the cheapest fix in the entire conversation, and one CraftedStays customers rarely have to think about since PMS-synced pricing keeps that layer unified.

📈 Ads Buy You Time, They Don’t Build Equity

Kin doesn’t dismiss paid ads, he’s seen returns north of ten to one within two or three months when they’re run well. But he’s clear about what ads are actually for: buying time while you build the channel that compounds. Organic and social, done right, becomes the highest-ROI channel long term, but it takes real patience, typically three to six months before it moves the needle, sometimes six to twelve before it becomes a consistent revenue driver. Ads fill that gap. They don’t replace the need to build one.

📱 Why Unpolished Content Is Outperforming Pretty Rooms

One of the sharper points in this episode: hotels are overproducing content, and it’s hurting them. Kin’s read is that polished shots of pretty rooms and pretty pools have become interchangeable across the entire industry, and travelers have stopped responding to them. What’s converting instead is user-generated, first-person content, ideally from a micro-influencer trading a stay for authentic footage, or from the property owner simply walking through the space on camera. If you’ve been holding off on video because it isn’t “professional enough,” Kin’s data suggests that’s backwards.

🧭 The Destination-First Content Play

Kin’s framing here applies directly to short-term rentals: almost nobody starts their trip by searching for a specific hotel or property. They start by choosing a destination, then work backward into where to stay. That means content built around the destination, not the property, is what actually intercepts a guest early in their planning. Local guides, itinerary posts, “what to do with kids in [destination]” content, these aren’t just SEO plays anymore. They’re also exactly the kind of structured, specific content that AI models pull from when answering travel questions. (Note: Kin references a McKinsey study on guest touch points during this discussion; the specific figures haven’t been independently verified for this writeup and are worth confirming before citing publicly.)

🤖 Inside the AEO Shift: Who AI Actually Trusts

This is the part of the episode worth relistening to twice. Kin’s AI search visibility research across roughly 20 hotels found that third-party sites, TripAdvisor, OTA listings like Expedia and Booking.com, and earned media from outlets like Travel + Leisure, are getting cited by AI models far more often than the hotel’s own website. AI tools break a single guest question into what Kin calls a “query fanout,” multiple sub-questions answered by pulling from wherever authority already exists on the web. That makes public relations and third-party mentions more important to AI visibility than most operators realize.

He and Gil also worked through a common misconception: traditional backlinks are losing weight as an SEO signal, but that doesn’t mean your presence across the web matters less. What LLMs weigh instead is entity reputation, how consistently and accurately your business is described across every place it’s mentioned, not how many links point back to your site. Buying backlinks won’t help. Getting mentioned accurately, repeatedly, and consistently across the web will.

🧩 Structured Data and the Multi-Engine Problem

Since AI models read structured data rather than raw page text the way a human would, clean schema markup on your website reduces the odds of an AI hallucinating incorrect details about your property. But Kin’s bigger point is that you’re not optimizing for one search engine anymore. Google, ChatGPT, Perplexity, and Claude all weigh signals differently, and some models (Claude, for instance) update on a training cycle that can run six months to a year behind, meaning a content change might not show results for a while even if the strategy is correct. Patience matters here as much as technique.

⚡ Rapid Fire with Kin Sio

Book recommendation: Extreme Ownership by Jocko Willink. Kin credits it with reinforcing that taking full responsibility, rather than blaming the system or the situation, is what actually gives you the ability to fix things.

Mindset advice for starting something new: Have a growth mindset. Kin picked this up during his time at Microsoft under Satya Nadella and puts it simply: be a learn-it-all, not a know-it-all.

One tactical move to start this week: Get on camera and walk through your property. No script, no editing, no blog post required. It’s the fastest way to build the kind of authentic content Kin’s data shows both guests and AI models are responding to right now.

🔗 Connect with Kin Sio

Website: lightson.co
Podcast: The Lights On Podcast
LinkedIn (his most active channel): linkedin.com/in/kinsio
Email: kin@lightson.co

🚀 Listen to the Full Episode

This writeup only scratches the surface of the AEO breakdown Kin and Gil get into, including the specific mismatch between what hotels are told about SEO versus what their own AI visibility data actually shows. Catch the full conversation on the Booked Solid Show for the complete walkthrough. And if your website is still working against you instead of for you, whether that’s rate inconsistencies, missing structured data, or a site that simply isn’t built to convert the traffic you’re already getting, CraftedStays was built to fix exactly that. Start your free trial at craftedstays.co and see what a platform built specifically for direct bookings can do.

Transcription

Kin: That, to me, is one of the tools in the tool belt. When a hotel has a good volume, we’re talking tens or hundreds of rooms, you usually need multiple channels or multiple levers to get all of that inventory sold on a daily basis. That’s definitely one key part where you’ll see return on investment the fastest. Usually, whether it’s paid Google Ads or paid Facebook ads, any of those, when we deploy that kind of advertising, within two or three months we’re able to see more than a one to ten return. Meaning every dollar of ad spend, we get ten plus dollars back in booking revenue. So it’s definitely the fastest way to see ROI.

Gil: Before we bring on our guest, I wanted to talk a little bit about something I’ve been hearing a lot from hosts. I keep hearing the same thing: “I know my website isn’t converting, but I can’t afford eight thousand dollars on an agency to rebuild it.” Here’s the thing, you’re learning all these marketing strategies, driving traffic, putting it all to work, but if your site isn’t built to convert, you’re basically lighting your energy and money on fire. And even if you could afford an agency build, every time you want to test something or make a change, you’re paying them again. You can’t iterate, you can’t test, you can’t improve. You don’t need a custom ten thousand dollar website to get the conversion rates that matter. You just need the right platform. That’s why I built CraftedStays. It’s purpose built for short-term rentals and designed from the ground up to help you drive more direct bookings. You can finally turn that traffic into bookings, and you can keep testing and improving as you learn, all on the platform. You don’t need to learn something new. So if you need help or want to get started, go to craftedstays.co and start your free trial. Now let’s bring on our guest and dive into hospitality and marketing.

Gil: Hey folks, welcome back to the Booked Solid Show, the show where we bring in top operators to discuss hospitality, operations, and direct bookings. On today’s show I have Kin from Lights On. He’s usually focused in the hotel space, but today we get to dive deep into the contrast between hotels and short-term rentals. He walks us through how hotels define their booking funnels, we dive into direct booking strategies in the hotel space, and we get into our independent findings on AEO research. So apologies in advance, today’s conversation gets tactical, but hopefully you’ll hang on tight and get a lot of value out of it. Without further ado, let’s bring him in. Hey Kin, welcome to the show.

Kin: Hey, thanks for having me, Gil.

Gil: This is interesting because we actually recorded together several months back on your podcast, Lights On, so it’s good to flip it around and get the chance to interview you. You come from a very different side of hospitality, the hotel side, and that was actually the contrast on our last episode. When I recorded on your podcast, I was giving your listeners a glimpse of what we see on the short-term rental side. I want to flip it around today and get a sense of what you’re seeing on the hotel side, and whether we’re seeing the same commonalities or a divergence based on stay type. So maybe to start us off, Kin, do you mind giving folks an introduction on who you are?

Kin: Yes, this is Kin. I own a company called Lights On Digital, based in Hawaii. What we do is help boutique and independent hotels drive more bookings, get more guests, and optimize each booking’s value. This is basically the same conversation short-term rental operators are having, just on the hotel side of things. Very much like residential versus commercial real estate, each hotel typically has a lot more rooms to deal with, a lot more inventory to think about, versus owning one unit or a handful of units. So the problems we’re trying to solve for hotels are very similar in principle. It’s just a matter of scale, when you’re constantly selling tens or hundreds of rooms every single night, the magnitude and distribution complexity is quite different from owning a single unit.

Gil: Yeah, and even some of our property managers have multiple units in the same market, sometimes hundreds of units in the same market, but it’s different than having one property with a hundred-plus subunits you need to sell. There’s a lot more scale and, I’d say, sensitivity, because anything you do or don’t do gets exacerbated. If your marketing is off, you’re not seeing a dip on one property, you’re seeing a dip on a hundred units. So I can see why you’d dive deep into helping clients with that. Given that, I’m curious about the services you offer and what have been the biggest pain points you’ve seen across the hotel space right now.

Kin: I’ll boil it down to two key points, then explain how that resonates from a guest’s journey. The two biggest things we’re solving are, first, how do we get more eyeballs on a particular hotel by driving more guests to the top of the funnel. More people discovering your hotel means more likely booking business. Once you have lots of eyeballs and guests knowing about your hotel, the next question is how does somebody actually book your hotel at the optimal rate. That’s the revenue management side. So the two arms of what we do at Lights On are, one, digital marketing, everything from organic search, paid search, and social media, we talk a lot about AI search too, in terms of getting more people to know you and book you, and two, revenue management, making sure the stay feels worthwhile enough for someone to book at the optimal rate, maximizing profit on the back end.

If you think about those two problems in the context of a guest journey, planning a trip, McKinsey did a study back in 2018 showing it takes on average 46 touch points over roughly 36 days for someone to book a trip. So starting from discovery and trip planning, it takes a while and a lot of research from the guest’s perspective. The question becomes, how can a hotel, or in your case a short-term rental, show up as many times as possible to take mind share from the guest until they’re ready to book. Once someone is ready to book a hotel, they’re looking at Expedia, Booking.com, your direct website, Google Hotels to compare pricing, and comparing you against competitors. With that many options, how do you optimize price and incentive to steer the guest toward booking your hotel through the channel that gets you the most revenue, ideally direct.

After booking, between the booking and the guest showing up, what does pre-arrival communication look like. What is the hotel doing to minimize cancellations and provide upsell opportunities, whether that’s room upgrades, early check-in, late checkout, or packages, before the guest arrives at the check-in counter. After the stay, making sure the guest has a way to leave reviews to boost trust signals, which becomes a flywheel for more bookings down the road. And finally, making sure the hotel captures the guest’s contact information so we can do email marketing down the road to nurture the relationship and drive repeat and referral business. That’s the full end-to-end guest journey, and we offer many different services across that journey to drive more revenue for the hotel.

Gil: So the thing I picked up is that it’s actually a pretty complex system, you mentioned 40-plus touch points. A guest searches for you, maybe finds you on social media, looks at you on Google Hotels, and you’re trying to figure out how to push them down the right channel while managing revenue at the same time. There are a lot of touch points and services. How do you simplify or bucket them, what’s the framework you use to guide clients on which KPIs to hit, and how do you make sure that, from a hygiene perspective, out of everything you could do for clients, you’re focused on the things that actually move them toward their targets?

Kin: Usually the starting point is goal setting. In the hotel world, the golden metric is revenue per available room, or RevPAR. I think some of the bigger short-term rental operators are starting to use that KPI too. It’s a way to normalize how you measure revenue generation across a hotel with 30 rooms versus one with 500 rooms, getting down to the same unit of measurement for revenue-generating ability. So RevPAR is first and foremost what we use. From there, we usually start with what’s closest to generating revenue, which in the hotel world is revenue management. Is the hotel set up so that when you go on Expedia or Google, the price you see is attractive enough to book. Most of us are travelers, you’ve probably had the experience of searching a hotel on Google, seeing Google Hotels, and seeing a list of pricing across different channels, Expedia versus direct, and thinking, I could book this hotel for two hundred dollars on Expedia, but direct booking is two hundred fifty, why is that.

That’s usually the first thing we find that drives guests away from direct booking, which, on the back end of the hotel’s math, means they’re paying another 15 to 25 percent commission, very similar to how short-term rental operators pay a similar fee to Airbnb. That’s why we always talk about driving more business to the direct website to save the commission. In the hotel world, the starting point is usually making sure your price is at parity across every channel. When a traveler sees the same price across multiple channels, they tend to book direct. So that’s the starting point before any other marketing or social media, because if you’re driving demand and traffic without capturing the business at the most optimal price, you’re wasting the money spent driving that traffic.

Gil: That’s interesting, because on the short-term rental side we rarely see a case where direct pricing is more expensive than the OTAs, and you gave the example of 200 on Expedia versus 250 direct. Why do you think that’s such a prevalent problem for the clients you’re starting with? Is it the systems, is it how Expedia handles discounts, where do you see the biggest choke point in these revenue management discrepancies across channels?

Kin: Usually it starts with the systems. You and I both came from tech backgrounds, so we can appreciate this. In hospitality, a lot of the technology systems are very fragmented. When a hotel starts running the business, it deals with many different systems, the core property management system, and a channel manager trying to distribute inventory and rates to maybe 17 different channels where you can sell rooms. In that complex process, with one inventory and one rate being distributed to many suppliers, there’s a lot of friction between systems that can cause discrepancies. Just using Expedia and Booking as an example, there could be a scenario where Expedia has a room priced at 200, Booking.com has it at 180, and your direct booking setup shows 220. There are a lot of moments where things fall through the cracks in that distribution process. For the most part, having a more expensive direct rate isn’t intentional, it’s usually a mismanagement of distribution and failing to keep the rate unified across all channels. That’s a complexity many hoteliers miss, especially boutique and independent owners who didn’t come up professionally in the corporate hotel world, so these principles and skill sets are foreign to a lot of newcomers.

Gil: Yeah, that’s very interesting. I know the hotel space has a lot more fragmentation. On the short-term rental side, the PMS is usually the central node that manages distribution, it’s also the channel manager, and dynamic pricing usually plugs directly into the PMS. It sounds like on the hotel side you have a lot more systems playing together, and often more legacy systems too. A lot of the tools we use in short-term rentals have only existed for the last five years or so, so there’s more modernization and more checks in place that probably don’t exist in some of the more legacy systems you’re working with.

Kin: Exactly. One thing I appreciate about the short-term rental side is, to your point, it’s a newer segment, so a lot of the tooling is more modernized than what we deal with in the hotel world. The other good thing about short-term rental operators is that most people coming in are a lot more comfortable using technology. I’ve noticed that most short-term rental operators I talk to are very comfortable using price management or revenue management tools, whereas in the hotel world it’s not as common as people would think when it comes to changing rates using software or machine learning algorithms. We still see occasions where revenue managers are manually updating and adjusting rates on a daily basis. So, for lack of a better term, it’s a more old-fashioned industry in hotels versus short-term rentals, which are more emerging and more comfortable with modern tech.

Gil: Yeah, I can definitely see that. Talk to me about ads. You mentioned ads as one of your focus areas. Why do you focus so much on ads? Is it because you see the quickest wins, or does it drive the most leverage? Why are ads such an important part of the marketing stack specifically for hotels?

Kin: To me, it’s one of the tools in the tool belt. When a hotel has a good volume, tens or hundreds of rooms, you usually need multiple channels or levers to get all that inventory sold daily. Ads are definitely one key part where you’ll see return on investment the fastest. Usually, whether it’s paid Google Ads, paid Facebook ads, or ads on OTAs, when we deploy those, within two or three months we’re able to see more than a one to ten return. Every dollar of ad spend gets ten-plus dollars back in booking revenue. So it’s the fastest way to see ROI. The downside of ads is that when you turn them off, you don’t get anything more, you don’t build brand equity over time. So to me, it’s a tactic most hotels need to deploy, not just for short-term revenue but to buy time to build the infrastructure needed for long-term branding. Ultimately, over time, organic marketing is still going to be the best channel for building your brand and capturing the guests you want. By organic, I mean social media, when done right, that should be the key channel driving most direct bookings, tying back to those multiple touch points when someone’s planning a trip. When a guest finds you on social media and sees your experience, especially through an influencer who stayed at your property, that’s usually the first touch point that gets someone curious about staying at your hotel or destination. But it takes time and the right strategy to do that well. If a hotel has no social media presence and starts building one, they’re not getting bookings the next day. Even the best social media managers need at least three to six months minimum before organic marketing becomes a major channel. Many hotels don’t have the patience to wait three to six months for an uptick in bookings, so in the meantime, there need to be other levers, in this case, ads, providing good compensation while you build that channel.

Gil: You mentioned social media as an organic path that takes time, and influencers as a great way to leverage and build that. As you think about different content types, which have you found most high leverage, and which have you tried that worked well in other industries but don’t work in ours? Give us insight on the content types that really resonate for driving traffic.

Kin: In the context of social media, two things are pretty prominent right now. One is video, Instagram Reels or TikTok. That kind of content works best, but it’s also the most expensive to produce. Second is user-generated or UGC-style content versus professionally created content. We joke about how most hotels do social media today, everybody’s showing professionally edited pretty rooms, pretty pools, pretty lobbies. It gets to a point where it all looks the same. Travelers aren’t that interested in that content anymore, versus big social media influencers creating first-angle, point-of-view videos that feel more authentic and actually show the true experience of staying at the hotel. My argument, even in the world of hotels right now, is that not enough hotels are willing to take the risk of creating less polished content. They’re stuck in the mindset of, let’s create very professionally produced content. But those intents aren’t working as well as inviting a micro-influencer over, maybe trading a free stay for it, and letting them create authentic content. That ends up working much better for driving reach and the intent for someone to actually book, far more than showing a pretty room on Instagram.

Gil: So it sounds like it’s really about leveraging other audiences to be the megaphone for hotels, less about building your own follower count, though that matters too, but the real draw is if you can get other people’s accounts posting organically and tagging you. Those are the ones you’ve seen get the greatest adoption.

Kin: Exactly. And over time, as a social media account matures, you still want to diversify content type. Yes, you’ll have some collaboration content with influencers, which is always great, but you won’t get that every day. So you still want to supplement with in-house content, whether that’s behind-the-scenes staff videos on site, or still posting some pretty room shots, you just shouldn’t be creating only pretty rooms every day. You want to mix up the content so people see everything, which gives them more reason to follow you and stay curious about your content. By the time a hotel gets to that stage, they’re already pretty far ahead in the game and should be seeing a lot more direct booking intent through that channel. But getting to that point, I’ve rarely seen any hotel accomplish it in two or three months. It usually takes six to twelve months, sometimes twelve-plus, until the channel becomes something that consistently drives revenue. But when done right, that should be the channel that gives you the highest ROI in the long run.

Gil: I’ve even seen some short-term rental operators focus on becoming the expert in a specific market. In vacation rentals, I don’t know if it’s much different than hotels, but we do get a lot of repeat visitors, especially in markets like the Smoky Mountains. We get annual visitors every single year, the same family visiting the same cabins, spending their holidays there. A lot of times they focus their content less on their own properties and experiences and more on the types of stays, local attractions, and nearby things to do. I don’t know if you see the same thing on the hotel side.

Kin: It’s the same. A big misconception in the hotel world is that when you’re in the industry, sometimes you think people are coming specifically to your destination to stay at your hotel, when for most travelers, we don’t start by booking the hotel. We start by looking for a destination. Say I want to bring my family on a ski trip this year, let’s look at ski markets, which ski destination do I want to go to. Once you start planning that trip, then you start looking at accommodations. Unless you’re one of those ultra-luxurious hotels where the hotel itself is the reason people travel, and I’d argue less than 0.1 percent of hotels fit that category, for the most part you’re not the first consideration when someone’s planning a trip. So how can you be first in their mind? That’s where building content related to the destination and trip planning comes in, even some of the classic search engine optimization approach, people writing blogs, which is still effective, by the way, creating content that positions your accommodation as part of the trip plan. For example, “I’m going to the Smoky Mountains, here are the ten things I’d do traveling with kids,” and layering in why people should stay at your accommodation as part of that trip. That makes people willing to read your content and start seeing your accommodation, building mind share upfront. That worked in the early days of people using the internet, and it’s also useful now that a lot of content consumers are AI models and LLMs. That kind of content works both ways.

Gil: I’d argue maybe more so now than before, that traditional SEO blog content, landing pages, and local guides are more important than ever. I don’t know how closely our listeners are tracking organic traffic, but we’ve seen a pretty significant decrease in organic traffic as Google AI Overviews and other tools change how people get information. People are clicking on actual links a lot less now because they’re getting answers in different ways. But at the same time, we’re seeing the inverse in terms of growth, we’re starting to see LLMs drive more traffic to our website. So I think that’s continued momentum toward creating top-of-funnel, blog-style content for people who aren’t even thinking about which destination they’re going to yet, they’re still in the infancy of building their itinerary, figuring out where they want to go, how to plan, where to stay, what attractions are nearby. That kind of top-of-funnel content feels more important to invest in than ever before.

Kin: Absolutely, we’re seeing the same trend in hotels. Across the dozens of hotels we manage right now, we definitely see a big drop in organic search traffic. Funny thing is, booking revenue doesn’t really move as much. That’s the whole phenomenon of zero-click marketing, when the LLM intercepts a lot of that intent and serves up answers before anyone needs to click one of the ten blue links on the first page of Google. People are still getting those answers, still seeing hotels recommended by AI. I’d argue that right now, on your traffic report, you’re starting to see an uptick in AI referral traffic, but that’s only a fraction of the actual AI search influence. The AI referral traffic you see in reports only measures direct redirection, someone in a ChatGPT session finds a hotel, clicks the link, and that gets captured. What doesn’t get captured is someone searching and asking questions in ChatGPT today, getting a handful of hotel recommendations, and for the most part not taking action the same day. They might come back three weeks later, remember doing that AI search, remember those hotels, and do a direct Google search of those names. So the pattern everyone should be watching is, yes, you’ll see less organic search traffic this year versus last year, but if you’re doing things right in terms of building the right content, you should see an uptick in direct traffic to your website. That should be a good proxy that, even without a perfect way to track AI search visibility, you’re likely showing up in AI search one way or another.

Gil: We’ve seen the same thing on our side, a downward trend in organic search links as AI filters out where people land, but we have seen an uptick in overall direct booking growth. I’d also argue that people who click through are usually much higher intent, so you may get fewer impressions, but the conversion rate is probably orders of magnitude different than someone doing a blank Google search.

Kin: We’re seeing the same thing. Expedia actually did a study earlier this year showing that intent coming from AI search converts much better, that’s one. And two, which is actually an advantage for hotels and short-term rental operators right now, many travelers are still hesitant to complete direct bookings within the AI search experience itself. We’re all getting comfortable asking AI for information, but we’re still less confident doing a direct transaction through AI. All the different OTAs, Expedia and Booking, are trying to build direct integrated booking experiences with the AI labs, but we generally still have trust issues passing AI our financial details. So right now you have a set of much easier-to-convert visitors coming to your website. That’s a good thing for accommodation providers if they know how to convert that traffic into actual bookings.

Gil: I had a discussion earlier this week with Jessica Gillingham and Richard out in the UK, and they had a contrasting view. Richard leans toward wanting AI to book all his stays, but those tend to be more commodity stays. When you’re doing real trip planning with your family, you want to feel the experience. Jessica, from Abodo Worldwide, argues that we don’t actually want AI to automatically book our trip, because it’s part of the booking process, part of the discovery of new places. I mentioned before the show, I visited Seattle and was able to discover different places, figure out how to get around, plan our days. AI helped with that, but I wouldn’t want a one-click AI experience that does everything for me, gets me on a plane, lands me somewhere, and drives me around. That’s not the experience I want. I think it’s part of the booking process. AI and LLMs will make planning easier and might get us further down the funnel toward checkout, but I don’t think it’ll bifurcate the same way we’re seeing on the e-commerce side with things like Google’s universal commerce protocol. I don’t think it takes away from the traditional way of planning trips, because overall it’s an experience.

Kin: I think trip planning and discovery are an enjoyable experience in themselves, not something people will give up. The part where you have to compare several sites for pricing, the logistical part, that’s the natural place for AI to help. But for the right kinds of trips, I don’t think people will give up the fun of it, a lot of people scroll social media all day just dreaming about and getting inspired by different travel content, because dreaming about your next vacation is fun. I actually had a conversation this morning with someone building an AI startup focused on using AI to drive direct bookings, this person came from Google and TripAdvisor, and they said the same thing. When it comes to booking a stay or a trip, the emotional stakes of something going wrong are much higher than buying a pair of sneakers. You don’t want to hand everything over to AI, and if something goes wrong, you’re the one facing your family and figuring out who’s responsible. It’ll be interesting to see how traveler behavior shifts over the next few years, but I don’t think the discovery part goes away.

Gil: Speaking to that point, we talked a little before the show about AEO, and how that changes the type of content or strategy you should invest in that maybe wasn’t a priority before. As you walk clients through their AEO strategy, based on your research, what should people be paying more attention to that they haven’t been?

Kin: Focus first on the part you have control over, your website. I know you’ve been building a really good product at CraftedStays for the short-term rental accommodation side of things. On our podcast I always say I wish at some point you’d expand that to hotels too, because smaller hotels with limited budgets for a new website could use similar services to really get their own website buttoned up. There’s a lot of technical work needed so the site is ready for LLM models to extract information from it. That’s the starting point. The other part is, unlike traditional search engine optimization, your website isn’t necessarily the biggest factor in whether AI recommends your stay or hotel. What we’ve seen on the hotel side is that third-party websites are taking on a lot more authority. For probably over 20 hotels we’ve done an AI search visibility assessment for, we find that TripAdvisor and OTA listings like Expedia and Booking.com tend to become the top citation sources for AI when it comes to travel-related answers. Something like, “I’m traveling to Hawaii, going to Waikiki for three days with a family of three, I need to look at accommodations.” For the most part, AI is breaking that one question into maybe seven different sub-questions, we call it query fanout, looking up different information across the internet. Usually it starts with TripAdvisor, because they naturally have a ranking of the best stays in Waikiki, then Booking.com and Expedia, then some third-party earned media, Travel and Leisure, Condé Nast, editorial websites with recommendations. That kind of content tends to influence AI recommendations much more than the hotel’s own website. So what we’re advising a lot of hotels right now is that a public relations strategy has become much more important than ever, because getting mentioned in more authoritative earned media tends to influence AI more than producing the exact same content on your own website.

Gil: That’s interesting, because right before the show we actually found somewhat of the inverse, but you corrected me that it’s not really the inverse. One thing we found in our research was that backlinks are much less important now than they used to be. You clarified that backlinks might be less important as a metric, but with LLMs, they’re not searching for one link pointing to another webpage, it’s really about the entity being cited and what that entity looks like and what the sentiment around it looks like. I think that’s the differentiator. What I wouldn’t encourage folks to do is go on Fiverr and buy a bunch of backlinks, that won’t help, if anything it probably hurts more than it helps, it deteriorates your authority score. But building your reputation across the entire web, having people reference you, we’ve seen it in our own business at CraftedStays. Even when partners, providers, or agencies write blogs about us, and don’t necessarily link to us, we still come up in search. It’s surprising coming from the traditional SEO world where backlinks were the metric, if you got a backlink from a source with a higher authority score than yours, that was usually a positive. But for LLMs, that means close to nothing, the weights and framework LLMs use to evaluate websites are very different from traditional SEO.

Kin: Correct. Aside from traditional web content, we’re also seeing a lot of other content types get used by AI for generated answers, whether that’s Instagram, Facebook, or YouTube videos. I’m looking at a dashboard for one of our hotels right now, and social media content isn’t as prominent in terms of how often it gets cited by AI, but it probably still matters, combined, maybe 20 to 30 percent of the time. So it’s less relevant but still useful. My recommendation to the hotels we work with is that you don’t necessarily need to create net new social media content just for AI visibility’s sake, but there are ways to be more deliberate when writing captions, making them more relevant to your unique selling points. That becomes a secondary signal to AI as well, because it’s not just about showing up in one location, AI captures many different trust signals across the internet and aggregates them. The more you show up in as many relevant places as possible, the higher the likelihood AI trusts that data, which actually leads to another problem, reputation management. Because showing up everywhere also means you have to be mindful of keeping all your data and information consistently correct. Once you have outdated information from a few years ago and no proper way to manage that, AI has a chance of citing or hallucinating information you didn’t intend. So managing the accuracy of your data isn’t just about discoverability, it’s also about reputation management.

Gil: That converges with something we’re seeing quite a bit, how important structured data is on your website. We’re seeing that more and more, not just in our industry but in e-commerce too. When you give very structured data on your website, and your properties are amenitized correctly, AI can crawl it structurally. As you mentioned before the show, that allows them to hallucinate less, because they’re not reading the content on a page like a human would, they’re reading the schemas feeding it. You can be very deliberate about how you want them to interpret what’s available at your property, what isn’t, and the FAQs they can read. I know Google is trying to phase out FAQ schemas, but I think it’s still relevant for all the different crawlers out there.

Kin: You actually pointed out a big point there. Traditional search engine optimization was all about Google, people barely paid attention to Bing. But in the AEO world, with multiple AI platforms out there, you’re not optimizing for just one chatbot. Google and Gemini might look at something completely differently from how ChatGPT, Claude, Perplexity, or Grok look at things. So even though Google publishes a set of criteria for what they’re still looking at from a search engine optimization perspective when it comes to AI, that principle isn’t necessarily the same way OpenAI evaluates things. Actually understanding and trying to optimize from one engine to another is kind of a headache.

Gil: It is. And even the way they’re optimized differs. Claude, for example, is trained primarily on models, that’s how it gets the root of its information. Whereas if you’re optimizing for Perplexity, Perplexity is really good at research because it’s always pulling live resources from the web, most of its queries go out and get fresh data. So if you’re optimizing for Claude, you might not see a difference until the next model iteration, which could be anywhere from six months to a year behind. That’s a signal worth remembering, you’re going to see different periods where some LLMs surface your content and others don’t. You might be optimizing for Claude and not see any difference, but that doesn’t necessarily mean you’re doing something wrong or that you should switch strategy. You just need to know that each of these chat engines behaves very differently.

Kin: Exactly. We obviously talk a lot about how messy the state of AI search is right now, but I also don’t want to overwhelm your listeners. For short-term rental operators with only one, two, or three rentals, chances are you don’t have the mind share or budget to deal with all of that. So thinking about priority, what makes the most sense for you first, is important. Gil, you’ve been building a great product with CraftedStays, and to me it’s a defensive play, making sure you don’t have to worry about the technical aspects of your website to start with. Have a good foundation, then figure out the best use of your time producing the right content for the right engine. If you look at monthly active users across different AI platforms, Google is still by far at the top, quickly followed by ChatGPT. So from that perspective, I’d say produce the kind of content we mentioned, thinking about trip planning, destination content, where you are. Doing videos on YouTube that you can distribute to Instagram or TikTok is a really big bang for your buck, because, back to what we said, you don’t need something super professionally edited. Just showing your face, walking through your stay, walking through your destination, that’s easy enough content to make, and it forces you to produce something authentic, which AI models tend to favor. As more AI slop content floods the internet, the AI labs are trying to optimize for content that’s more humanly generated and authentic. That’s why I started my podcast earlier this year, that’s why you have your podcast, everything we’re talking about right now is completely authentic, not AI slop. I find building videos to be the easiest way to produce content that could matter. The other big point about video is that, as of right now, YouTube videos and Reddit content are two of the places AI tends to reach back years for. If you made a video three years ago, AI will still cite it, versus user review data on the hotel side, TripAdvisor reviews, Google reviews, Expedia and Booking reviews, those tend to have AI pulling fresher, more recent data. So when you know a piece of content you put out today could serve you for years, that’s a very high ROI, especially when each of us has such limited time to produce content.

Gil: You can only do so much out of each day.

Kin: Yeah.

Gil: It’s been a huge pleasure diving deep into this with you. We usually end the show with three questions. First one, what’s been a book that has inspired you on your journey?

Kin: I’m looking at my bookshelf right now, I’d pick “Extreme Ownership.”

Gil: Is that Jocko Willink?

Kin: Yeah.

Gil: Why is that one so important to you?

Kin: I think the first thing is, to do anything in life, you as a person have to take responsibility and extreme ownership in everything you do, because that really gives you the internal drive to achieve more and be better. I think a lot of us, more or less, get into unfortunate situations at some point in life or career, and the natural default behavior is to start blaming, blaming other people, blaming the situation, blaming things that aren’t controllable by you, and then you have no way to pick yourself back up and get to a better place. Just having that free agency to be able to control your destiny no matter what situation you run into, I think that’s the most liberating career lesson you can use to get yourself and your family to a better point.

Gil: I love that. I read that book maybe five years back, at the peak of my management career, when I had a team of product managers I was working with, and it was really helpful in figuring out how to take ownership over my team’s performance. Because if the team messes up on something, it’s ultimately on me. If they missed a deadline and I didn’t check in on them when I should have, that’s still on me.

Kin: It’s on you, yeah.

Gil: Kin, second question. What’s one piece of mindset advice you’d give to someone starting something completely new?

Kin: Have a growth mindset. I learned that back in my career at Microsoft, when Satya Nadella had just taken over as CEO. He was highly recommending a book on the growth mindset. I can’t remember the exact author, but really the premise is, don’t be the know-it-all, be the learn-it-all. When you assume you know everything, you get a fixed mindset and you’re less willing to improve, learn something new, or admit you’re wrong. Once you’re in that mindset, that’s when you start falling behind. The world is changing every day, every moment, especially with AI right now, so you have to keep an open mindset and realistically look at everything you do every day, what are the things you don’t know about, and be willing to be wrong. Because when you admit you’re not doing something right, you naturally start finding ways to get better. That’s probably the biggest thing I took from my career at Microsoft, and it still influences me a lot.

Gil: Nice. All right Kin, last question. For any listeners looking to get started on direct bookings or amplify what they’re already doing, what’s something tactical they can take away from today’s episode and put into practice right away, something small enough that they won’t sit on it for weeks?

Kin: Start creating content. We talked about video content, you don’t even have to write a blog post or anything, just get on a camera, shoot yourself, and start walking through your property. Just start talking to the camera. That’s the fastest way to build momentum. It’s not going to get you direct bookings overnight, but it’s the fastest way to build momentum, and once you’re consistently creating that kind of content, before you know it, you’ll see travelers coming in and staying at your place because of the content you created.

Gil: I love that, awesome. Okay, it was a huge pleasure having you on. Where can folks find out more about you, how can they follow you and everything you’re diving into?

Kin: Definitely check out our website, lightson.co. I’m also a LinkedIn junkie nowadays, I post every day about hotel and commercial strategy and AI search, so just follow me on LinkedIn, I give out a lot of value every day when it comes to tactics and tips.

Gil: Thanks again, it was a pleasure recording with you again.

Kin: Nice chatting with you again, Gil.

Gil: All right, bye.

Kin: Bye.