I spent Saturday morning at a table with some of the travel industry's most successful operators at the Virtuoso Technology Summit in Las Vegas—advisors who've built multi-million dollar practices, owners of luxury agencies, and specialists who command five-figure planning fees. These aren't people struggling with technology adoption. They're sophisticated business operators looking for competitive advantage.

What struck me wasn't their eagerness to adopt AI—that was expected. It was the fundamental disconnect between what they think AI can do and what it actually does.

The Model Training Myth

"We're training our own model on our client data," one advisor told me confidently.

No, you're not.

This was the most persistent misconception I encountered. Elite operators throw around "training" like they're building custom neural networks in their back offices. What they're actually doing is prompt engineering with context injection. Training a model costs tens of millions of dollars—third-party estimates place GPT-4's training around $80-100M. Even fine-tuning requires specialized infrastructure and costs thousands to tens of thousands depending on data volume.

What these operators want is RAG—Retrieval Augmented Generation. Feed your client profiles and destination knowledge into a system that can search and retrieve relevant context, then generate responses based on that information. That's what I showed them with Claude Projects: upload your knowledge base and let the AI search through it to generate personalized content.

The distinction determines whether you'll succeed. You're not training a model; you're organizing knowledge for retrieval.

The Information Problem

Another advisor asked me how to get ChatGPT to know about her latest Bhutan offerings that launched last month.

Here's what most operators miss: These models aren't databases. Their training data has specific cutoffs—Claude Sonnet 4 uses January 2025, GPT-4 varies by version, and GPT-5's knowledge cutoff is September 30, 2024—surprisingly old for such a recent release. That exclusive new resort you're selling? The AI has never heard of it.

"But it seems to know current things sometimes," another advisor protested.

That's because some ChatGPT interfaces can search the web, or because older information seems current. The model itself remains frozen in time.

Stop thinking of AI as a database to query. Think of it as an interface for interacting with information. It's not about what AI "knows"—it's about what AI can help you organize, retrieve, and communicate.

The Discovery Challenge

The conversation that revealed the real disconnect: "How do I get ChatGPT to recommend my new Bhutan offerings when someone asks about luxury Himalayan travel?"

This is the fundamental challenge small operators face: They want to be found when potential customers use AI for travel research.

One vendor at Virtuoso was selling services to get travel marketing data into the Common Crawl, one of the datasets used to train some models. For smaller operations, this offers low predictability and long lag times. Even if your data makes it into the next training run, it would be so diluted among trillions of tokens that it might as well not exist.

The harsh reality? As of August 2025, most mainstream chat assistants either rely on general web search or do not guarantee inclusion of specific business content. When someone asks Claude or ChatGPT about African safaris, it won't know about your exclusive camps that opened last month.

The best current workaround remains traditional SEO, since some AI tools can search the web. But the promise of AI discovering and recommending your specific offerings remains largely unfulfilled for small operators in 2025.

The ChatGPT Lock-In

Even after I demonstrated Claude's expanded capacity for document storage—Projects can handle substantially more files versus ChatGPT's varying file limits—several advisors remained skeptical about switching.

"But all my clients know ChatGPT," one said.

This is OpenAI's notable success: brand attachment in a commodity market. People stick with ChatGPT not because it's better, but because it feels familiar and established.

For business operations, it's costly. Claude's Projects feature changes what's possible with larger-scale knowledge management. One advisor had been manually copying client preferences into every GPT conversation. With Claude Projects, those preferences persist automatically.

The difference becomes stark at scale. Twenty clients with detailed profiles, destination research, supplier contracts, and communication history? That's potentially hundreds of documents. ChatGPT's file handling becomes constrained. Claude Projects handles it at scale.

The Hidden Competitive Advantage

Here's what I didn't say at the table but became clear: These operators are sitting on a goldmine they don't recognize—their insider knowledge that doesn't exist anywhere online.

Every elite operator knows things ChatGPT will never know. Which "luxury" camps have deteriorated but still trade on old reputations. The real transfer times between locations during rainy season. Which exclusive properties actually have availability versus those booked two years out. The suppliers who deliver versus those who consistently disappoint.

This proprietary knowledge becomes weaponized when properly organized. Set up a Claude Project with your insider intelligence—every supplier evaluation, seasonal variation, logistics reality, and hard-won lesson. When a potential client shows you an AI-generated itinerary they found online, you can run it through YOUR AI assistant that actually knows the truth.

"Your ChatGPT itinerary looks beautiful," you can say, "but let me show you why it won't work." Your Claude, armed with real data, will instantly identify that the suggested camp closes in March, the "2-hour scenic drive" is actually 7 hours on unpaved roads, and the "hidden gem restaurant" has been closed since 2023.

You're not being negative, you're demonstrating irreplaceable expertise. The generic AI gave them a fantasy. You're offering reality, refined by decades of experience that no public model can access.

Implementation Path

For elite operators ready to move beyond misconceptions:

Start with Claude Pro and upload your knowledge base—client profiles, destination guides, supplier agreements. But more importantly, document your secrets: the camps that photograph well but disappoint in person, the real logistics that no website mentions, the seasonal realities that marketing materials ignore.

Set up Perplexity Tasks to monitor your specialties with scheduled searches. Use Granola to capture client interactions.

Use metaprompting: Don't write prompts—have AI write them for you. Use voice tools like SuperWhisper to brain dump what you want, then feed that to Claude with: "Turn this into a clear, structured prompt for an AI assistant."

Build evaluation workflows. Never send AI output without review. Use one AI to create, another to verify. Domain expertise remains essential.

The Real Opportunity

The Virtuoso Technology Summit revealed a travel industry eager for AI transformation but confused about what that transformation looks like. The gap between perception and reality is creating two classes of operators: those who understand these tools as sophisticated pattern matchers and information organizers, and those still believing in training myths.

The winners won't be the ones who "train their own models" or get their marketing data into training datasets. They'll be the ones who accept AI for what it is—excellent pattern recognition and retrieval systems—and build workflows that leverage that reality.

For travel advisors: Stop trying to make AI into something it isn't. Start making your existing expertise scalable through intelligent information organization. The competitive advantage isn't in the technology—it's in understanding the technology accurately enough to use it effectively.

The insights in this article come from my session at the Virtuoso Technology Summit in Las Vegas on August 9, 2025, where we discussed practical AI implementation for luxury travel businesses.

The scenarios range from OTAs adding chatbots to autonomous AI agents negotiating directly with suppliers. What they all miss is the fundamental question: where do travelers discover and plan trips today?

Spoiler: It's not on Expedia.

The OTA Fortress Isn't Going Anywhere

Let's get one thing straight—OTAs aren't getting disintermediated anytime soon. Gavira's scenarios 3 and 4, where suppliers bypass OTAs entirely, ignore the brutal economics of travel distribution.

Here's what OTAs actually excel at: single-item booking. One hotel stay. One flight. One experience. The classic "search, detail, price, compare, book" workflow that dominates travel today.

Booking.com commands 69.3% of the European hotel OTA market, while Expedia holds 11.5%. In the US, these two giants account for roughly 93% of the core OTA leisure market. Their scale isn't the result of breakthrough tech—it's operational mastery at brutal scale. They solved the hardest problem in travel: aggregating fragmented inventory while handling payments, cancellations, and customer service across thousands of suppliers.

80% of travelers visit OTAs to research and compare prices, and OTAs spent $17.8 billion on marketing in 2024 to maintain that traffic dominance.

But try to plan a complex, week-long itinerary using an OTA? The systems completely break down.

Even current LLMs break down under this weight. OpenAI's most advanced model achieves only a 10% success rate on complex travel planning benchmarks, while earlier LLMs manage less than 1% accuracy compared to humans at 100%.

Travel Planning Happens in Communities, Not Comparison Sites

Here's what these scenarios fundamentally misunderstand: travelers don't start their journey on OTAs. 89% of travelers turn to social media for travel inspiration, with 75% relying on social platforms to discover destinations.

They start in Facebook groups asking "Where should I go for my anniversary?" They browse Instagram stories from friends' trips. They ask for recommendations in Slack channels and Discord servers. Importantly, 55% get inspiration from family and friends' accounts—far more than the 18% who follow celebrity influencers.

During my time at TripAdvisor, this pattern was crystal clear: most of our traffic was using our reviews to validate travel choices people had already made elsewhere. We weren't the starting point—we were the confirmation step. OTAs follow the same pattern. People aren't browsing Booking.com to discover their next destination. They're going there to book a hotel they've already settled on as part of a trip they've already planned...elsewhere.

What's happening in those "elsewhere" conversations is "anti-planning." Instead of search-driven queries, people are dreaming out loud. They're sharing half-formed ideas, describing ideal experiences without concrete constraints.

"I want to go somewhere with great food and history, maybe Europe, in the fall, with my partner who hates crowds..."

That's not a search query. That's a conversation containing far more valuable intent data than any keyword-based search. 62% of travelers who use social media for trip planning make specific decisions after viewing content.

The gap between this kind of dreaming and actually having a bookable itinerary is massive.

The Technical Reality Check

What strikes me most about these scenarios is the assumption of travel tech infrastructure that simply doesn't exist.

Gavira envisions AI agents from airlines and hotels conducting "complex, instantaneous algorithmic negotiations." Have you tried to change a United flight recently? Some airline legacy components date back to the 1970s. Airlines globally spend $37 billion annually on IT—much of it just maintaining infrastructure held together with "rubber bands and crossed fingers."

These systems were designed around batch processing and physical paper documents. Customer records aren't even in one place. Most hotel APIs can't handle room preference requests consistently.

Yet we're supposed to believe these same companies will suddenly deploy sophisticated AI agents that can negotiate dynamic pricing and custom packages? The technical debt alone would take years to unwind.

The Real Disruption Play

Rather than chasing OTA disruption, the smart play is building AI that works where travelers actually are—in communities and social platforms—and can handle the complexity that breaks current systems.

The technical challenge is significant. You need AI that can:

• Monitor and capture the nuanced conversations where travel dreams are shared

• Break down complex, multi-week itineraries into manageable components

• Manage the combinatorial weight of coordinating flights, accommodations, timing, experiences, and logistics

• Reassemble everything into executable plans that actually work

This isn't just a UX problem—it's a computational architecture problem. You're building a travel planning compiler that can decompose complex intent into bookable components, then optimize the entire system.

Current OTAs can't do this because their systems are built for single-item transactions. Standard LLMs can't do this because they lack persistent state management for complex travel logistics.

But here's the key insight: once you solve this technical problem, the data becomes incredibly valuable. A detailed, coordinated week-long itinerary isn't just a travel plan—it's a high-value product that multiple parties will pay for.

A Two-Sided Marketplace Opportunity

Build a platform that uses AI to create complex travel plans from community conversations, then operate as a two-sided marketplace for that high-value data.

On one side, you're capturing travel intent from communities and social platforms, using AI to transform vague dreams into detailed, coordinated itineraries that individuals couldn't create themselves.

On the other side, you're distributing these plans to multiple buyers. OTAs want qualified leads with higher gross booking values. But more importantly, travel agencies and specialists want access to the detailed planning work—they know how to take a sophisticated itinerary and execute it profitably.

A detailed week-long Japan itinerary with optimized train schedules, restaurant reservations, and experience coordination is worth far more to a boutique travel agency than a simple hotel booking is to an OTA.

You're not trying to replace the booking infrastructure—you're creating more valuable inventory to flow through multiple channels while aligning incentives instead of fighting them.

Why This Actually Makes Sense

The travel industry has always been about intermediation, not disintermediation. The most successful companies succeed by creating new forms of valuable intermediation.

There's a genuine technical moat available to whoever builds specialized AI that can capture conversational travel intent, decompose it into coordinated components, and manage the complexity of reassembly. Current systems simply can't handle this computational problem.

I remember trying to plan a complex trip across multiple Japanese cities with my family. We spent hours coordinating train times, hotel locations near stations, restaurant reservations that fit our schedule, and activities that worked for both adults and kids. Each decision cascaded into a dozen others. No single platform could handle it all.

That's exactly the kind of complexity that AI could solve—but only if it's built specifically for this problem, not bolted onto existing booking systems.

The Real Disruption

While everyone's debating whether AI will replace OTAs, the actual disruption will come from whoever figures out how to use AI to create more valuable travel products in the first place.

That company won't look like Gavira's scenarios. It won't be an autonomous booking bot or an incremental improvement to existing OTA search.

It'll be a platform that meets travelers where they actually are, uses AI to help them envision better trips than they knew were possible, then efficiently distributes that value through the existing travel ecosystem.

Full disclosure: I'm an engineering advisor to Globe Thrivers, which is building exactly this model under CEO Shir Ibgui. They're creating a community-driven platform that captures conversational travel intent and uses AI to transform it into detailed, bookable itineraries. Their B2B2C approach already has interest from enterprise partnerships — proving that this two-sided marketplace model resonates with both travelers and the industry.

The winners won't replace the OTA fortress. They'll render it obsolete by turning inspiration into itineraries before the first search bar is touched.

Want more contrarian takes on where AI opportunities actually lie? Subscribe to HyperDev for practical analysis of emerging tech.

Hotel pricing is one of my favorite subjects because it's so wonderfully arcane. At Tripadvisor, we had an amazing hotels team that not only drove the bulk of the company's revenue for many years but remained by far the most profitable division. We learned the hard way just how complex hotel distribution really is.

Let me break down why this isn't quite the breakthrough it's being portrayed as.

What MCP Actually Is, And Why It's Not The Solution

MCP (Model Context Protocol) is a structured framework designed by Anthropic to standardize how AI models connect with external data sources and tools. Described by Anthropic as "like a USB-C port for AI applications," it provides a standardized way for AI systems to access information from business tools, content repositories, and development environments (Anthropic Docs, Anthropic Blog).

But sophistication doesn't matter when the underlying systems can't deliver what you need. Even the most elegant orchestration framework can't overcome fundamental infrastructure limitations in the hotel tech stack.

The "Real-Time" Pricing Myth

What does "real-time pricing" even mean in this context? Reading between the lines, what the article actually describes is metasearch tools pulling data directly from systems of record: property management systems, channel managers, revenue management systems, and booking engines.

I've seen firsthand how challenging this can be. During my time at Tripadvisor, we explored various approaches to enhance our hotel pricing capabilities, including connecting with GDS services. These efforts revealed the significant complexities in the hotel distribution ecosystem, where GDS systems primarily serve travel agents and offer special discounted pricing intended for "closed user groups" – not meant to be sold as retail rates available to the general public.

More fundamentally, hotels maintain multiple differentiated rates depending on their distribution partnerships. There are retail rates, member rates, package rates, wholesale rates, and corporate rates – each with their own business rules, restrictions, and distribution channels. This complexity exists before we even get to the technical constraints.

The infrastructure limitations aren't theoretical - they're documented technical bottlenecks that any engineer who's worked with these systems can attest to. Most backend systems simply aren't built to handle direct traffic at the scale that Google, Tripadvisor, or Trivago demand.

The Caching Compounding Effect

Perhaps the most significant practical challenge for price accuracy is caching. Caching protects tech companies from runaway volume and unqualified requests that drive up infrastructure costs, particularly when operating in the cloud. But it creates a critical technical trade-off: freshness versus performance.

There are layers upon layers of distribution channels, each adding their own layer of caching that can abstract away price from the source of truth. Add up the compounding effect of each layer, and you quickly get price accuracy issues between the display price on metasearch sites like Tripadvisor, Kayak, Trivago, or forward distribution channels like Google Hotel Ads, Perplexity, or affiliates like Rakuten.

This caching challenge exists regardless of whether you're using MCP, a direct API, or any other integration method. AI can't solve stale data problems.

The Regulatory and Trust Dimension

Beyond technical challenges, there's another critical aspect: "bait and switch" pricing is considered uncouth in the industry. You can't knowingly display a rate that, upon a user progressing through the multi-step shopping funnel, suddenly increases. This practice destroys consumer trust and is increasingly in regulators' crosshairs.

This pricing accuracy challenge is behind newer "total price display" regulations like California's law that went into effect last summer, requiring companies to display prices inclusive of fees, not just the base rate (California Hotel & Lodging Association). Similar regulations are emerging globally.

When cached data becomes stale or doesn't include all applicable fees, companies risk not just disappointing customers but potentially violating these regulations. No AI protocol, no matter how sophisticated, can overcome fundamentally inaccurate or incomplete source data.

Hotel Pricing Is Dynamic, Not Real-Time

Hotels don't typically price rooms "in real time" anyway - they use dynamic pricing tied to seasonality and demand forecasts. Research from pricing technology providers reveals most systems refresh rates on scheduled intervals (hourly/daily) rather than instantaneously, with true real-time pricing remaining rare due to technical limitations and integration challenges (SiteMinder, Mews).

To be fair, major chains like Marriott and Hilton may be pushing for (or already implementing) more real-time capabilities in their own direct channels. But even if the source hotels have real-time pricing, the challenge remains that when metasearch companies send requests for rates and availability, one or more caches along the supply chain may be stale. The end result is the same – what reaches the consumer isn't truly "real-time" pricing.

The pricing engines adjust rates based on booking pace, competitor rates, and other factors - but these changes happen on a schedule, not continuously. Industry-standard solutions from providers like IDeaS and Duetto primarily use predictive analytics with periodic updates, not continuous real-time adjustments.

The companies that benefit most from pricing arbitrage aren't hotels but OTAs and bed banks. They buy inventory ahead of time at the best prices, then leverage their demand models to maximize profit as stay dates approach (HotelTechReport, Cloudbeds). This economic reality isn't changed by adding an AI layer on top.

The Itinerary Complexity Problem

Here's something else the article glosses over: hotel rates are "itinerary chunked" - not priced per room per night but per stay. Hotels often employ Best Available Rate (BAR) models with derived rates applying various rules based on factors like minimum/maximum stay requirements, weekday versus weekend pricing, and dynamic discount structures (PriceLabs, HSMAI).

A two-night stay spanning Friday-Sunday might be priced differently than the sum of individual Friday and Saturday night bookings. Multiply this by thousands of properties, each with their own rules, and you quickly see why "real-time" anything becomes exponentially complex - and why caching is particularly challenging since travelers rarely share identical itineraries.

PMS Tech Stack Limitations Are Real

The scaling limitations of Property Management Systems are well-documented and acknowledged by industry experts. Technical analysis confirms several key bottlenecks:

• Legacy architecture not designed for distributed high-volume queries (Shiji Insights)

• Database contention during peak periods limiting response times

• Synchronous processing models that create performance bottlenecks

• Insufficient caching mechanisms for high-frequency queries

• Limited API capabilities restricting integration options (Hospitality Tech)

A 2024 hospitality technology assessment by Phocuswright found that PMS systems "remain the primary technical bottleneck in distribution" with most designed primarily for property operations rather than as distribution platforms. Even modern cloud-based systems struggle with the volume demands of metasearch platforms (OpsMatters).

Room Type Babel

Let's not forget that room types add even more variance to pricing, and there is no standard for matching room types even within a single chain, let alone across chains.

Room type standardization difficulties across hotel chains are a legitimate industry challenge with no universal taxonomy existing across major chains and independents. The same room category (e.g., "Deluxe") can have drastically different amenities by property, with regional naming variations creating significant mapping challenges (HotelMinder).

A critical confusion point: Room Types and Rate Types are often conflated, making price comparison across sites extremely difficult. Rate types have numerous variations:

• Refundability options (fully refundable, partially refundable, refundable until X days before check-in)

• Included amenities (free breakfast, free wifi)

• Membership rates (loyalty program tiers, AAA, corporate)

Meanwhile, room characteristics create even more variations: ocean view, accessibility features, refrigerator, microwave, greater square footage, etc. These multipliers quickly create multiple different rates for what appears to be the same basic "Double Queen Bed" room type.

When a consumer sees different prices for what appears to be the same room, they might actually be looking at different rate types or slightly different room types with variations in amenities or views.

This confusion extends to "best price guarantees" offered by many booking sites and hotel chains. These are often thinly veiled advertising ploys with layers of customer support designed to ensure customers rarely prevail in their claims. The burden of proving exact room type and rate type equivalence across sites falls on the consumer, creating an almost impossible hurdle.

AI room type classifiers do exist in the market, with companies like Giata and Vervotech offering solutions. Giata provides AI-based room mapping with claimed "99.99% accuracy," while Vervotech offers standardization and de-duplication of room data (Giata, Vervotech). But the technology is still in active development, with full standardization remaining a future goal rather than current reality.

How OTAs Have Already Solved This (Sort Of)

Case studies from major OTAs reveal they've invested heavily in technical infrastructure specifically designed to overcome the limitations in hotel systems, creating competitive advantages through superior technology rather than simply better contracts.

They maintain proprietary caching layers to reduce direct queries to suppliers, develop normalized room type databases, and implement intelligent query management to reduce system load (Vervotech, Koddi).

This is why companies like Booking.com and Expedia can handle massive metasearch volume while direct hotel connections often can't - they've built entirely separate technical infrastructures to work around the limitations.

The Real Problem: Infrastructure, Not Protocols

All this is to say that pricing services in hospitality aren't really an AI problem - they're a fundamental travel tech infrastructure problem. The industry runs on antiquated systems that weren't designed for direct integration at scale.

The real challenge isn't getting an AI to talk to these systems (that's what APIs are for). The challenge is modernizing the core infrastructure that manages availability and pricing.

What Would Actually Help

If we want to improve real-time pricing in hospitality, we need to focus on:

1. Standardizing APIs across the hospitality tech stack

2. Building infrastructure that can scale to handle metasearch-level traffic

3. Creating industry-wide standards for room type classification

4. Developing more robust caching strategies that account for itinerary variations

There are initiatives working on standards, like the Open Travel Alliance (OTA – yes, another overloaded term in the industry). During my time in the industry, I've seen companies utilize some of these standards in building connectivity APIs, including amenity codes and other specifications. Organizations like these could be a source of standardized room type codes – but adoption remains fragmented and incomplete across the industry.

Frankly, AI efforts might be better placed elsewhere in the travel distribution ecosystem. For instance, as an upper funnel technology, AI could do much more to understand customer intent. Understanding which travelers are likely to book (versus just browse) would be far more valuable to metasearch companies than trying to solve real-time pricing problems. Sending volumes of low-quality referrals to suppliers just drives up cloud costs and drives down cost-per-click in a CPC business model.

AI could help travel companies identify higher-intent consumers and help those consumers find the right travel choices at the top of the funnel, before referring them to booking sites. This would provide real value to both providers and consumers in the industry, rather than focusing on forcing real-time pricing into systems that weren't designed for it.

AI will certainly help with associated services, especially in natural language understanding and content classification, but an AI-mediated protocol for "real-time" pricing isn't addressing the core issues.

The Bottom Line

MCP may be useful for what it is - a sophisticated framework for agent collaboration - but it isn't magical middleware that suddenly makes decades of fragmented travel tech infrastructure disappear.

The next time you see breathless coverage about AI solving longstanding industry problems, look closely at what's actually changing. Often you'll find that the fundamental limitations remain unchanged, just wrapped in new terminology.

Real progress will come from unglamorous infrastructure work and industry standardization, not from adding another layer on top of already shaky foundations.

It's clever tech, and frankly, the first real attempt I've seen to crack programmatic AI video matching at scale. But there's a deeper story here that goes beyond the technical achievement. After reflecting on how platforms approach creator partnerships—including companies like Globe Thrivers where I serve as an advisor—I'm starting to wonder if Expedia is optimizing for the wrong metric entirely.

The Technical Achievement (And What It Signals)

What Expedia has built appears, at least on paper, to represent a significant engineering accomplishment. While I haven't been able to test the feature extensively myself, the technical concept—using computer vision and AI to match any Instagram reel to bookable travel itineraries—is genuinely ambitious. That said, the gap between announcement and execution in this space can be substantial.

Early testing reports suggest the feature works: one beta user shared that when they searched for "an A-frame cabin in Maine," it served up a relevant Maine itinerary, and for a mystery beach location reel, it correctly identified Bora Bora.

Currently available for beta access, with full-scale availability for U.S. travelers anticipated in the coming weeks, this initiative marks Expedia as the first online travel agency to incorporate real-time, AI-powered travel planning directly into Instagram.

Strategic Context: Where Expedia Sees the Market

"It's no secret people are getting their news, inspiration and travel ideas from influencers and content creators, with a recent study showing that 80% of millennials rely on social media for travel decisions," said Jochen Koedijk, chief marketing officer at Expedia Group. "We want to be where consumers get their inspiration, which is through social media."

This positioning reveals Expedia's view: meet consumers where they're inspired. But it raises deeper questions about how platforms view the creators generating that inspiration.

The Creator Compensation Question

Here's where the strategic questions become more pointed. According to industry reports, Expedia has stated they won't be offering compensation to creators whose content drives bookings through the Trip Matching tool. This stands in notable contrast to their broader creator economy efforts.

Strategically speaking, this creates an interesting tension. Just last year, Expedia Group introduced Travel Shops, "a way for influencers and other content creators to earn commissions when travelers book their recommended hotels," with 14 launch partners. These are "personalized storefronts that give creators a dedicated page in the Expedia app to curate recommendations and earn commission."

The question becomes: why compensate creators in Travel Shops but not in Trip Matching? The answer likely reveals competing strategic visions within Expedia's approach to the creator economy.

Alternative Paradigms: The Partnership Model

Reflecting on this, I'm struck by the contrast with companies like Globe Thrivers. Our entire model centers on equitable creator relationships, recognizing something fundamental: in today's travel landscape, creators aren't just content producers—they're influential voices that command genuine trust and drive real booking behavior.

Take Shir Ibgui, founder and CEO. Coming from years leading strategic partnerships at AI startups, she built the company specifically to address what she saw as a broken creator-platform relationship. As she puts it, "At Globe Thrivers, we believe planning for travel should be as authentic, seamless and empowering as the trip itself." Her vision creates "a virtuous cycle of value for all players in the travel space—travelers, travel agents, hotels, creators, and more."

This isn't just marketing speak. Globe Thrivers is developing a model that actively compensates creators for their contributions, viewing them as essential partners rather than free content generators. It's a philosophy that prioritizes ecosystem growth over platform capture—creating value with creators rather than extracting value from them.

Platforms like Videreo offer yet another approach. Instead of building walled gardens, they're creating brand-agnostic storefronts where creators maintain agency over their partnerships. Creators get compensated, brands get direct consumer relationships, and the platform facilitates rather than controls the entire transaction.

Strategic Implications for Travel's Future

Reflecting on these competing approaches, I see two distinct visions emerging for travel's future:

The Platform Consolidation Play: Own the entire customer journey from inspiration to booking. Maximize extraction at each touchpoint. Scale through technological sophistication and network effects.

The Ecosystem Partnership Play: Enable transparent creator-brand relationships. Take sustainable cuts of expanding transactions. Scale through collaborative growth and aligned incentives.

Expedia appears to be pursuing both simultaneously—partnering with creators through Travel Shops while extracting value through Trip Matching. The question is whether this dual approach can remain sustainable as creators become more aware of their value and alternatives.

Bottom Line: Winning Battles vs. Winning Wars

Expedia has built impressive technology that demonstrates real AI capability. As the first OTA to integrate real-time, AI-powered travel planning directly into Instagram, they've achieved a notable technical milestone.

But strategically speaking, they may be optimizing for the wrong metrics. In an economy where authenticity and trust drive travel decisions, treating creators differently across platforms—compensating them in Travel Shops while extracting value from their content in Trip Matching—creates an inconsistent value proposition.

The future likely belongs to platforms that align creator incentives with platform growth consistently—not those that selectively extract value from creator content based on the specific tool being used.

From my perspective working with companies in this space, the most sustainable approaches recognize that creators are partners, not resources to be mined selectively. Winning the AI and social commerce battle means little if you lose the creator economy war through inconsistent partnership approaches.

And that war is just beginning to unfold.

Want to test Expedia Trip Matching? It's currently in beta testing—try it by using #TripMatchingAccess in a direct message to Expedia or visit Expedia.com/tripmatching.

Interested in how travel tech is evolving with AI? Follow my coverage of strategic developments and technological innovation at HyperDev.

The Good: AI as the Ultimate Trip Planner's Assistant

ChatGPT (GPT-4o specifically, using a Pro GPT account) did a great job of structuring the itinerary—grouping activities by geography, balancing heavy sightseeing with downtime, and even sequencing major travel legs. It even embedded Google Maps links for easy navigation, which was surprisingly useful.

For filling in gaps, I also loaded some publicly available itineraries into context, which helped it generate recommendations faster than I could Google them. That was a big win. Plus, GPT’s ability to iteratively modify plans—rearrange days, swap activities, adjust pacing—was a huge time-saver. No more manually copy-pasting in a spreadsheet for hours!

The Bad: AI Still Needs a Human Editor (Preferably One Who Knows Japan)

Here’s the thing—if I had zero knowledge of Japan, this itinerary would have been an absolute disaster. While GPT nailed the broad strokes, it also:

• Got some basics completely wrong.

• Missed important details that an experienced traveler would catch.

• Suggested things that don’t actually work logistically (e.g., back-to-back activities that require impossible transit times).

It also struggled with preserving edits. Asking for a change in one section often resulted in weird modifications elsewhere—activities mysteriously moved around, some disappeared, and sometimes entire days got reshuffled even when I explicitly asked it not to. Eventually, I resorted to saving each finalized day as a static file and only reloading the whole itinerary when necessary.

The Ugly: Memory & Booking Limitations

One major limitation: GPT couldn’t edit more than a few days at a time before losing track of details. However, this appears to be specific to editing within the canvas feature and is likely a short-term problem.

And, of course, it doesn’t book anything. Hotel availability? Flight schedules? Forget it. You still need actual travel tools for that. Maybe one day AI will seamlessly integrate with booking systems, but we’re not there yet.

Verdict: Worth It?

Absolutely. Despite its quirks, using GPT made structuring and modifying a complex itinerary much easier. It eliminated the cold start problem and got me to 80% completeness in record time. The final 20%—sanity checks, verification, and fine-tuning—still required human effort, but that’s a much better use of time than staring at a blank document.

The Bigger Takeaway: Invest in Travel APIs, Not AI Itinerary Builders

This experiment reinforced a core belief: investing in full-fledged AI itinerary builders isn’t worth it. The moving parts are too many, and the gaps in AI’s capabilities (live availability, real-time pricing, logistical nuance) make it an uphill battle.

Here's one day in the itinerary we built (pasted as an image because LI's article editor wouldn't preserve Gdocs formatting):

Industry Trends Reshaping Travel

• Discovery is happening elsewhere. TikTok, Instagram, and YouTube have overtaken search engines as primary sources of travel inspiration. Over 40% of young travelers now plan trips via social media.

• Google dominates search. Google Flights, Google Hotels, and Google Things to Do are funneling users away from aggregators like TripAdvisor. AI-driven search will only tighten Google’s grip.

• Experiences are the new frontier. The tours and activities market is projected to hit $300B by 2030. Platforms like Viator and GetYourGuide are aggressively capitalizing on this shift.

• Direct bookings are winning. Hotels, airlines, and OTAs are investing heavily in direct relationships, reducing their reliance on intermediaries like TripAdvisor.

TripAdvisor Today: Strengths, Weaknesses, and Untapped Potential

Strengths: ✔ One of the most recognized travel brands globally ✔ Largest repository of user-generated travel content and reviews ✔ Viator is growing rapidly and a key revenue driver

Weaknesses: ⚠️ Direct traffic is down, with visits at half the level of major OTAs ⚠️ Monetization is lagging; review volume and relevance are declining ⚠️ The U.S. market is eroding faster than expected, leading to missed opportunities

The Plan: Four Key Moves to Drive Growth

1. Transform TripAdvisor from a Static Content Repository into an Engagement Platform

• Surface high-value content effectively. AI-driven personalization should ensure travelers quickly find relevant reviews and recommendations.

• Turn the app into a real-time travel companion. Live itinerary management, location-based recommendations, and seamless booking can transform TripAdvisor into an interactive tool.

• Strengthen trust. AI-powered verification of reviews is necessary to combat fake content and maintain credibility.

2. Fully Realize Viator’s Potential

Viator is quietly TripAdvisor’s most valuable asset, generating $270M in Q3 2024 alone. But long-term profitability remains a question. To accelerate growth:

• Strengthen supply chains. Exclusive partnerships and preferential rates can secure premium experience providers.

• Expand financial offerings. With over $1B in transactions, integrating flexible payment options, insurance, and financing solutions can drive conversion.

• Enhance competitive positioning. Viator must differentiate by curating exclusive, high-quality experiences.

• Prepare for IPO. Spinning off Viator could unlock capital for growth and allow it to scale independently.

Market Valuation Insight: GetYourGuide secured $194M in funding in June 2023, valuing it at ~$2B. Viator’s Q3 2024 revenue ($270M) and gross booking value ($1.1B) suggest it could command a valuation exceeding $2B.

3. The Fork: Strengthen or Sell

Operating separately from TripAdvisor, The Fork’s value proposition is unclear. To enhance its competitive position:

• Diversify revenue streams. Subscription-based premium plans for restaurants can drive predictable income.

• Leverage technology. AI-driven reservation management can reduce no-shows and optimize table availability.

• Expand strategic partnerships. Collaborating with food delivery services or acquiring regional restaurant booking platforms could accelerate growth.

• If it can’t scale effectively, divest. If The Fork can't be positioned as a high-growth asset, selling it to Booking Holdings (which owns OpenTable) may be the best move.

4. Fix the Advertising Business

TripAdvisor’s advertising potential remains underutilized.

• Shift to native advertising. Sponsored content should enhance—not disrupt—the user experience.

• Leverage influencer and creator-driven content. Strategic partnerships with TikTok, Instagram, and YouTube can drive engagement.

• Expand off-platform monetization. TripAdvisor’s rich travel data is an underleveraged asset—data partnerships can unlock new revenue streams.

The Next Move: Two Strategic Paths

Scenario 1: Private Equity Buyout & Restructuring

• Viator goes public.

• The Fork is sold.

• TripAdvisor is streamlined into a lean, high-margin engagement platform.

Scenario 2: Strategic Acquisition

• A player like Booking Holdings, Expedia, or a media-tech firm acquires TripAdvisor for its brand, audience, and data assets.

A Fork in the Road

TripAdvisor isn’t just a travel site—it’s an iconic brand. But failure to adapt has eroded its influence. Without bold action, TripAdvisor risks fading into irrelevance. Incremental changes won’t cut it. The company must decide: lead, pivot, or get left behind.

The next move is crucial. Will TripAdvisor redefine its future, or become a cautionary tale?