Canary Study: Hotels With AI Messaging Grew Upsell Revenue 54%

Hotels using AI-powered messaging grew upsell revenue by 54% on average during their first six months with the technology. That figure anchors the first study of its kind on guest-AI interaction, released by Canary Technologies, which analyzed more than 10 million AI-powered interactions across calls, text messages and webchats at thousands of hotels in over 100 countries.

The report, titled "From Answers to Actions: What Millions of Guest Interactions Reveal About AI Inside the Stay," is Canary's third on AI in hospitality. The previous two examined how hoteliers perceive AI and how they plan to adopt it. This one looks at what guests actually do when an AI answers.

The short version: they book, they ask, and they rate the stay higher.

What did the study measure?

Canary, which sells an agentic AI platform for hotel and guest management, examined interactions across three channels: voice calls, SMS and webchat. The sample spans thousands of properties worldwide, from independents to major brands. Canary's platform is trusted by more than 20,000 hotels in 125-plus countries, including Marriott International, Four Seasons, Choice Hotels, Wyndham Hotels & Resorts, BWH Hotels and IHG Hotels & Resorts.

Three findings stand out for operators:

• Personalization at scale. Six in 10 AI responses go beyond general information and address the guest's specific context — the property, the booking, the situation.
• Satisfaction. Guests who engage with AI are 11% more likely to leave a five-star rating. The mechanism is simple: AI replies in seconds and eliminates the wait for a human on hold.
• Revenue. The 54% average upsell revenue growth in the first six months, across stays managed with AI messaging.

The personalization number matters because it counters the canned-chatbot stereotype. Sixty percent of responses carrying guest-specific context suggests the systems are pulling from live property data, not a scripted FAQ.

Where is the upside?

The report argues the biggest gains remain ahead. Nearly two-thirds of guest conversations — 65% — include a request that goes beyond information, such as making a reservation or fixing a broken air conditioner. Of those requests, 45% are revenue-generating activities like bookings.

Read that again: nearly a third of all AI conversations contain a booking or similar transactional request. Every delay or dropped hand-off in that pipeline is lost revenue or a disappointed guest.

Fulfilling guest requests typically spans several teams and systems — front desk, engineering, F&B, the PMS. Each hand-off invites error and delay. Canary positions agentic AI (systems that act across tools autonomously, rather than merely answering questions) as the fix: connect the teams and systems, automate the hand-offs, complete requests quickly and reliably.

Gurtej Gill, VP of AI Operations and Adoption at Canary Technologies, framed the finding in one sentence: "More than 10 million guest conversations across thousands of hotels globally told us what the future of hospitality looks like."

He added: "Guests want their specific requests taken care of quickly and effectively. Delivering on that takes agentic AI that works across every team and system on property, so hotels can deliver great service, improve operations and drive new revenue."

What are the commercial terms?

The study itself does not break down pricing. What it does establish is the vendor relationship behind the data: Canary sells the platform producing these interactions, so the sample consists of its own client base. The company reports managing tens of millions of guest and staff interactions end-to-end autonomously each year, covering touchpoints from discovery through post-stay.

The technology replaces, in effect, the traditional switchboard-and-front-desk queue — the delayed callbacks, the unanswered texts, the missed upsell prompts at check-in. Who pays varies by property; the study measures outcomes, not contract structures.

What comes next?

Canary is a semifinalist for the 2026 HTNG/AHLA TechOvation Award and will present additional findings from the report at The Hospitality Show in Miami, November 2-4, at booth 2713. The full report is publicly available through Canary's resources portal.

For operators, the operative question after 10 million interactions is no longer whether guests tolerate AI. It is whether the property's systems can close the loop on the 65% of conversations that ask for action — because the revenue, per Canary's data, sits precisely there.