AI, Branding, Digital/Social, News / Aug 12, 2026

Marriott Solved Ads Attribution. Here’s Why Everyone Else Can’t

Nick Wheeler

Sr. Director, Digital Strategy

Walk into any hospitality marketing conference in 2026 and you'll hear the same conversation on repeat: campaigns underperforming, remarketing that misses, offers that vanish into the noise. The instinct is to blame the channel, the creative, or the algorithm. The actual problem usually sits one layer deeper, in the data pipeline that's supposed to connect a booking to the ad that drove it.

For much of the industry, that pipeline is broken.

The Main Chokepoint: Booking Engine Data

A handful of global brands, Marriott, Hilton, and a few others, built their own booking engines years ago. That decision quietly became one of the biggest competitive advantages in hospitality. Their digital ecosystems are closed-loop: a guest clicks an ad, lands on the site, books, and the data flows straight back to the marketing platform. It's why you rarely see these brands at industry conferences asking for help with attribution. They solved it by owning the infrastructure.

Everyone else, boutique hotels, independent resorts, destination marketing organizations (DMOs), soft-branded properties, rely on third-party booking engines and referral platforms. Every one of those third-party handoffs is a place where the customer journey breaks. A guest sees an ad, but the booking engine that closes the sale doesn't report back to the platform that ran it. The result is that most of the industry is spending marketing dollars without being able to prove, or improve, what those dollars actually did.

The truth is in the reporting. Independent hotels now route roughly 63% of bookings through online travel agencies (OTAs) rather than direct channels, and in some markets that share climbs closer to 80%, according to Cloudbeds' 2026 State of Independent Hotels Report, which analyzed 90 million bookings across more than 180 countries. Every one of those bookings represents a customer journey the property's own marketing data may have never fully captured.

Some booking platforms do offer data back to properties, but that "solution" comes with its own costs. It adds fees on top of already razor-thin margins in a market where price-first shopping and referral commissions eat into revenue. Because most of these platforms weren't built with ad-platform data specifications in mind, integrating what data is available is often slow and manual, a serious liability when advertising in 2026 runs almost entirely on real-time signals.

Why Impressions & Clicks Aren’t Important Success Metrics Anymore

For the first couple of decades of digital marketing, hospitality could get away with measuring performance on impressions and clicks. That era is over. Ad platforms are now hyper-sensitive to data quality: the accuracy of what you feed them determines whether they find the right audience, remarket efficiently, or serve your offer to someone who will never book.

Feed a platform broken or incomplete data, and it doesn't just perform worse, it actively works against you. It ends up wasting spend on the wrong audiences, breaking remarketing sequences, and burying offers under noise the algorithm can't resolve. That risk is only growing as AI-driven discovery takes on a bigger role in how guests find and book properties. Skift's coverage of the shift has noted that traditional last-click attribution can't capture the value of moments that influence a traveler weeks before they book, which only makes weak, fragmented data more commercially damaging, as agentic AI tools move from generating insight to taking action on it.

3 (Not Simple) Existing Fixes

The good news is that this problem isn't unsolved. The bad news is that every existing solution comes with a real trade-off in cost, accuracy, or complexity.

1. Probabilistic modeling This approach tells ad platforms the story of the sales funnel without real-time data, by assigning an estimated average value to specific website actions. It requires full access to booking engine data, access some engines don't even grant their own customers. It's also the least reliable option: from our experience with dozens of DMOs, revenue attribution accuracy can swing anywhere from 60% to 95%, which makes it a shaky foundation for forecasting or budget planning. A booking path that starts with a hotel's own campaign but closes on an OTA often shows up in neither the ad account nor an accurate commission ledger, so the model ends up guessing at a gap it can't actually see. It also requires a reactive approach to seasonality and fails to provide clarity on the impacts of major regional or global events such as political discourse, wars, natural disasters, weather, and more.

2. API integrations This approach relies on an application programming interface (API) connecting the booking engine directly to a central data hub. It's a stronger option than probabilistic modeling, but that central hub is the catch: if your web analytics stack leans on an unstable industry standard, building one gets expensive fast. Accuracy improves meaningfully with this approach, but the cost, technical expertise, and development time required can stall or block it entirely for smaller properties and DMOs.

3. Third-party connectors Equally accurate to API and the fastest option, and unsurprisingly, the most expensive. Vendors that specialize in hospitality connectivity know exactly how valuable this data problem is to solve, and pricing often reflects it, with markups running 10x, 20x, or more over the underlying cost of the technology.

It's worth naming the stakes here directly. OTA commissions already run 15 to 25% per booking, and can push past 30% once preferred-placement fees and promotional tools are factored in, an expense HSMAI has flagged as a customer acquisition cost few hoteliers can afford to leave untracked. Layer a 10 to 20x markup for the connector meant to help you compete with OTAs on your own channels, and the economics for independent properties get punished fast.

What Sustainable Change Looks Like

No single current option provides a complete solution for the hospitality industry as a whole. What is truly required is a hybrid approach, one that combines existing methods while eliminating predatory pricing driven by artificial scarcity. Achieving this outcome hinges on pursuing one or both of two distinct paths.

Booking engines open up their data. Booking engines must either evolve toward genuinely open connectivity, including architectures that function natively within a property or DMO's website, or, at a minimum, make API access available without imposing enterprise-level pricing. Legacy attribution models often fail to map the full customer journey precisely because a modern guest can interact with five or more touchpoints and multiple devices before booking, a pattern Google's own travel research has documented as running weeks longer and requiring dramatically more touchpoints than a single-device journey, and disconnected systems simply can't follow them across it. Data accessibility is table stakes in modern digital marketing almost everywhere else; hospitality is one of a few industries still catching up, and closing that gap is what would let small and mid-sized operators compete with the giants on equal footing.

The industry builds its own connector. A neutral, industry-sanctioned third-party connector, architected like the existing solutions but without the price-gouging, would let the entire market benefit from infrastructure that today only the largest brands can access. Private vendors have little incentive to build this, since it undercuts their own margins. That leaves trade associations and industry societies as the more plausible builders: an investment that benefits everyone except the vendors currently profiting from the gap.

How The Brand Leader Can Help

None of this makes the fix simple, but it does make it diagnosable. If campaigns are underperforming, remarketing feels erratic, or nobody on the team can say with confidence which channel actually drove a given booking, the root cause is rarely the creative or the media buy. It's almost always upstream, in how, or whether, booking data reaches the marketing platforms at all.

That's the layer we spend most of our time in: helping hotels, resorts, and DMOs untangle their data pipeline so marketing spend can finally be measured, trusted, and improved. Every property's setup is a little different, and so is the right fix.

The ultimate next step: Connecting Marketing Data to Economic Impact (Blog #2 - Coming Soon)

Nick Wheeler

Sr. Director, Digital Strategy