Where Agentic AI is Actually Earning its ROI in Hotels
Not guest chatbots. Not autonomy for its own sake. Otel's Paul Ryan on where agentic AI is quietly earning its ROI in hotels right now.
As part of Revfine's industry panel, 12 hospitality tech leaders were asked where agentic AI is delivering measurable ROI, and how hotels are rebuilding trust and accountability around it.
Paul Ryan, co-founder of Otel, argues the ROI isn't in replacing judgement; it's in closing the gap between what a hotel's systems already know and what its people have time to act on. His full answer below.
Revfine Industry Expert Panel Question:
“As agentic AI moves beyond the chatbot era into truly autonomous decision-making, which operational areas are delivering measurable ROI today? How are forward-thinking hotels restructuring their teams to manage accountability and trust these systems?”
Paul Ryan's answer:
Walk the HITEC floor and you’d think the future of hotel AI is a better conversation; guest chatbots, voice concierges, smarter messaging. Meanwhile, the measurable ROI is landing somewhere far less glamorous: in the gap between what a hotel’s systems already know and what its people have time to act on.
The real gap: signal vs. execution
Consider the strangest fact in hotel technology. Nearly every hotel has bought a system whose entire job is to price the hotel, and nearly every hotel still pays a person to second-guess it, daily. Ask a revenue manager what they actually do each morning and you get a list of judgement calls no RMS answers:
Am I correctly priced across the short, medium and long-term windows? What are my competitors doing — and why? Is the market moving with me or against me? Are restrictions in place that shouldn’t be? Has my booking curve shifted this year? Which of these bookings will wash?
The RMS produces a number. The human produces the reasoning, and re-derives it from scratch, every single day, by logging into five systems and stitching the picture together by hand.
That’s where the ROI is today. Not in replacing that judgement, but in arriving with those questions already answered, so judgement gets spent only where it actually changes the outcome.
With our customers, that shows up in hard numbers: ADR gains on soft days that would previously have gone unchallenged, booking-curve shifts caught in week two rather than at month-end review, restrictions quietly killing bookings flagged the day they should have come off.
And the pattern isn’t confined to revenue. The same gap between signal and action sits in F&B and procurement; menu items priced below their input cost, supplier price creep compounding for months, and in portfolio oversight, where owners make decisions without knowing which property has the biggest pickup gap. In every case the signal already existed. Agentic AI’s ROI is collapsing the distance between insight and execution.
Autonomy isn't the goal; traceability is
Which brings me to the assumption buried in the second half of the question: that more autonomy equals more value. The most forward-thinking hotels we work with are doing the opposite; building human-in-the-loop workflows where an agent earns autonomy the way a junior analyst does. It starts by recommending. It proves itself against reality. Only then does it graduate to acting within guardrails a human has set.
Why "the algorithm said so" doesn't survive an ownership meeting
There’s a reason revenue managers intervene daily in a system they paid six figures for: its reasoning is a black box, and “the algorithm said so” doesn’t survive an ownership meeting. Trust isn’t earned through accuracy alone, it’s earned through traceability. Source, timestamp, logic, the options considered. When the reasoning is inspectable, adoption follows.
What restructuring actually looks like
And that’s the real restructuring story. Nobody is removing humans from the loop; they’re moving up the stack. The revenue manager stops being the person who assembles the morning picture and becomes the person who interrogates it and makes the call. The two-hour morning scramble becomes a focused ten-minute review, and the other hour and fifty minutes is where the ROI compounds.
Read all 12 industry responses on Revfine.
This is what Otel builds toward every day. See how it works.
Continue reading
Related articles
AI Risks in Hotels: What Can Go Wrong, and What Good AI Looks Like
An honest look at the real risks of AI in hotel operations, from black-box pricing to alert fatigue, and what traceable, human-in-the-loop AI should look like.
Paul Ryan·21 Jul 2026
What I Saw at HITEC 2026 That Nobody Else Is Talking About
Most startups at HITEC were building guest-facing AI. The problem eating the first two hours of every GM's morning went almost unaddressed. Here's what Otel AI CEO Paul Ryan saw.
Paul Ryan·2 Jul 2026
Valuemaxxing. Hotels have always done it. They just called it Monday morning.
What does AI value look like when the margin for waste is zero? For hotel teams, the answer has always been the same: did the work get done?
Paul Ryan·17 Jun 2026