AI’s Reckoning in Hospitality

At last week's HITEC conference in San Antonio, more than 6,000 hospitality executives—representing brands like Wyndham, Hyatt, IHG, and Starwood, along with hundreds of independents—confronted a central question: is the industry's AI investment paying off?

The gap between ambition and reality remains stark. While 25% of hospitality businesses report actively scaling AI, fewer than 10% qualify as "AI future-built" per BCG, meaning they have AI embedded across core operations, a solid data foundation, and demonstrable returns. "A lot of companies are throwing spaghetti at the wall to see if it sticks," said Dale Gomez, associate teaching professor in hospitality technology at Florida International University. "They want to see ROI."

Alongside the AI conversation, a broader operational shift is emerging. Many operators still run on financial infrastructure that can't support the automation and interoperability AI promises. Payment systems long considered "good enough" are now driving measurable revenue losses, as inefficient tech shifts from a minor annoyance to a genuine reason guests don't return. Across four days and more than 50 meetings, four trends stood out.

Direct Bookings Are No Longer Won by SEO Alone

For years, hotels fought OTA dependency through search engine optimization—investing in content and rankings to capture guests before they landed on Expedia or Booking.com. That approach is eroding. Salesforce principal solution engineer Jack Wang cited striking data: 65% of Google searches triggering an AI Overview end without any website click. On mobile, that figure jumps to 78%. Industry-wide, traditional search traffic is down roughly 25% as AI-generated summaries replace the ranked link lists SEO was built to win.

Winning inclusion in an AI answer demands a different playbook. SEO rewards keyword density, backlinks, and page authority. AI models reward accurate, machine-readable structured data: room types, amenity details, policies, local context, cancellation terms. A property can rank well in traditional search yet remain invisible to an LLM—over 90% of accommodation sites are undetected by AI models, per Nokumo research.

The behavioral shift is already measurable. Phocuswright reports that 56% of travelers have used AI for trip planning, booking, or in-destination help in the past year. For operators, the first step is an audit, not a budget line: can the LLMs your prospective guests consult accurately describe your property's room categories, amenities, policies, and local context? If not, that gap is costing bookings today.

The Fragile State of Hospitality AI

HITEC's conversations repeatedly surfaced an uncomfortable truth: much of the AI being deployed across hospitality is brittle. Most companies are scaling AI without strategic clarity, a sound data foundation, or the operational architecture to support it.

The root cause is almost always fragmented data. property management systems, CRM, loyalty, food and beverage, and payments each hold partial views of the same guest, and AI outputs mirror those gaps. The same data problem degrades personalization, inflates reconciliation time, and leaves guest profiles incomplete.

Salesforce lead solution engineer Amanda Sharp framed the challenge as AI operationalization, not adoption. Building AI features is now cheap, but running them reliably in production—integrated into workflows that trigger real actions—is hard. Companies that do this well connect clean data to decision points while there's still time to act. Delta Air Lines, for instance, embeds a live AI concierge in its mobile app, using SkyMiles profile and operational data for context-aware support. At Wynn Las Vegas, revenue managers get predictive alerts when performance trends below target, with recommended actions attached.

For most travel operators, the bottleneck is data connectivity, not model quality.

Payments Friction Carries a Price Most Hotels Haven't Priced

Payments have historically sat at the back of hospitality's mind—a cost center to manage and minimize. HITEC's payments conversations reflected a real shift: payments are now a competitive lever, and the stakes are explicit.

Stripe-commissioned survey data from nearly 400 hospitality executives underscores the issue. 90% said payments matter for growth; 37% said a lack of payment options has the greatest negative impact on the guest experience. 58% reported their fraud systems block legitimate transactions, and 74% said fragmented systems waste excessive time on reconciliation.

The structural math favors OTAs, who can fund large payments teams because their revenue justifies the headcount. Independent hotels can't match that directly, but a lean team on the right infrastructure can now support payment methods across dozens of countries at a fraction of the cost of a big in-house operation.

The coverage gap itself loses bookings. "The moment we don't support a payment method is the moment this guest goes elsewhere, to a platform or channel that supports their preferred way to pay," said Sebastien Leitner, VP of strategic partnerships at Cloudbeds. A property that doesn't support the dominant payment method in a target market isn't just creating friction—it's routing that booking to an OTA that does.

Invisible Tech, Measurable Loyalty

Guest tolerance for broken technology is near zero. "There is zero empathy for technology that doesn't work," said Tanya Pratt, global VP of strategy and product management at Oracle Hospitality. "If it's not working, it's going to cause more frustration than if there's a line at the front desk, because people are used to that." When tech fails, guests rarely complain; they simply don't return.

The measure of success, then, is when technology works so well guests never notice it. Starwood Hotels CIO Denise Walker described the vision: a returning guest walks into a room pre-set to their preferred temperature, with their TV channels and pillow firmness already configured. No announcement needed. "It doesn't have to be delivered in a way that says, 'How did you know that?'"

Shannon McCallum, VP of hotel operations at Resorts World Las Vegas, sees an evolution: "We're moving from 'I told you this, so you know it about me' to 'I didn't tell you anything, and now you're predicting it.'"

That invisible personalization requires connected data—tech that integrates across the existing stack and consolidates guest information into one system. It enables a brand to recognize the same guest browsing the website or standing at the front desk, and it's the foundation on which memorable human moments are built.

The Stripe Angle

As guests increasingly discover properties through AI assistants rather than search engines, and bookings shift to agents, the revenue differentiator will be payment experiences that convert and financial systems that interoperate. Stripe Data Pipeline connects payments data with booking and customer systems for unified revenue visibility.

Stripe's infrastructure supports hospitality on three fronts:

  • Drive direct bookings. Payment experiences tailored to local methods and markets keep bookings on the direct channel. Built-in fraud detection and payment tokens let agentic commerce transact without exposing guest credentials.
  • Increase trip spend. Ancillary revenue from dining, experiences, and partnerships requires payments infrastructure that works across the property and with external partners. Cloudbeds, for example, saw 15% revenue growth for businesses using Cloudbeds Payments and a 14.8% average revenue increase for those expanding payment methods through its Stripe partnership.
  • Cut costs. More efficient B2B money movement and fraud protection reduce reconciliation work and limit losses without adding headcount.