On 1 October, the hospitality technology consultancy h2c released its AI Opportunity Study 2026. It is based on 122 responses from 113 hotel chains, which together operate more than 8,200 properties and about 1.3 million rooms. The headline went round the industry within days: 91% of hotel chains already use AI, but only 28% have a company-wide AI strategy.

Most coverage read it as a story about hotels. I read it as a story about the vendors selling to them, because several beliefs hotel tech companies hold about selling AI do not survive the numbers. Here are six, with what the data says instead.

Myth 1: Hotel chains still need convincing that AI is worth using

The data: 91% already use it and another 8% plan to within two years. The share of chains still exploring how to approach AI fell from 29% in 2025 to 12% in 2026.

What it means: the education pitch is over. A demo that opens with what AI can do for hotels tells a chain what it already knows. The question buyers are asking now is how to make AI work across the whole company, and that is the question your pitch should answer.

Myth 2: Putting AI in the product sets you apart

The data: external tools such as ChatGPT make up 60% of reported day-to-day AI use. AI built into vendor systems makes up 27%, and in-house tools 12%.

What it means: your AI feature is competing with a general chatbot the hotel's staff already use for free. That does not make it worthless; 31% of chains now rely mainly on vendor-provided AI, up from 21% in 2025. But the reason to choose you is what a chatbot cannot do: work inside the hotel's own data and workflow, safely, every day. Sell that, not the label.

Myth 3: The return on AI is obvious, so buyers will pay for it

The data: nearly 70% report efficiency and automation gains and 59% say staff can spend more time on higher-value work. Only 13% report a measurable return on investment. Chains rate AI's contribution to overall business performance at 5.6 out of 10.

What it means: hotels feel the benefit but cannot prove it, and that gap stalls renewals and expansion. The vendor who agrees a baseline and a target before go-live, and reports against it, owns the proof the other 87% lack. The same discipline that makes a pilot convert applies here, as covered in how to run a hotel tech pilot that converts.

Myth 4: The buyer has an AI strategy you can plug into

Only 28% have a company-wide AI strategy led by senior leadership. Chains rate their own AI knowledge at 3.4 out of 10.

The data: 56% name lack of expertise, skills or training as their biggest barrier, and 32% say they have no clear AI strategy or roadmap. Most chains run AI through pilots, departmental projects or individual tools.

What it means: in most deals there is no central AI owner waiting to approve you. You are selling to a department, and the people using your product will rate their own AI knowledge low. Training, plain-language setup and a fast first result are part of the product, not extras. Reducing time to first value matters even more when the user is unsure of the technology. It also means expansion to the rest of the group needs its own plan, because nobody at head office is pulling you in.

Myth 5: Integration and data are the hotel's problem

The data: integration with existing systems is a barrier for 38%. Data security and privacy concern 29%, and data governance and access management 28%.

What it means: if your answer to the integration question is that the hotel's IT team will handle it, many chains have no one to hand it to. Show which systems you connect to, what data you read and write, who can see it and where it is stored, before the buyer asks. Integrations already sell hotel software more than features do, a point made in why integrations sell hotel software, and with AI the data questions come earlier and harder.

Myth 6: Chains want AI agents talking to guests first

The data: AI agents are a top innovation area for 70% of respondents. Among chains using or planning agents, 70% are looking at staff-facing uses such as reporting, operational analysis and finance, against 55% for guest-facing uses. On autonomous decisions in guest conversations, 56% worry about unclear liability for errors, 54% about privacy and security, and 52% about accuracy and bias.

What it means: the easier sale is the agent that helps staff, where a mistake is caught internally. If you sell guest-facing agents, lead with control: what the agent can and cannot decide, how a person takes over, and who carries the risk when it gets something wrong.

The opportunity hiding in the booking numbers

One more finding matters for distribution vendors. Seventy percent of chains expect AI-driven booking channels such as ChatGPT and Gemini to increase direct bookings within one to two years. Yet only 22% systematically track how their brands and properties appear in AI-generated results, and 33% have done nothing specific to be found or booked through those channels. That gap is a product and a service for booking engines, CRS, channel managers and digital agencies. Google's own changes point the same way, as covered in today's checklist on Google Search Campaigns for Travel.

A caution on the numbers

The study covers hotel chains, averaging 73 properties each, and the results are not weighted by chain size. If you sell to independent hotels, treat it as a signal rather than a measurement of your market. The direction is still useful: buyers who use AI but cannot prove its value, and lack the skills to scale it, need a vendor who brings the proof and the support with the product.

If your AI story is not turning into signed deals, the cause is usually in positioning or proof rather than the product. A free revenue leak audit shows where in your pipeline those deals are being lost.

Frequently Asked Questions