How Hotel Revenue Management AI Sets the Price You See Every Second

The price you see on a hotel listing is not set by a human. It is set by a revenue management system (RMS) — an AI model that re-forecasts demand every few minutes and adjusts the room rate accordingly. Three vendors dominate: IDeaS (SAS-owned, used by 30,000+ hotels), Duetto (used by luxury groups), and RateGain.

What the RMS actually does

It ingests 40+ signals in real time: current booking pace, competitor rates, search volume, flight arrivals into the destination, local events, weather forecasts, historical comparable periods, length-of-stay patterns, and channel mix. Then it optimizes for Revenue Per Available Room (RevPAR), not occupancy, not price.

Why your search can change the price

If 50 users search the same hotel for the same dates in one hour, the RMS sees demand pressure and nudges the rate up within minutes. This is why hotel rates appear to "rise while you are deciding". They literally are — because your search is part of the input.

The rate elasticity sweet spots

RMS models have blind spots. They are optimized for BAR + length-of-stay. They under-optimize for:

• Advance-purchase rates (non-refundable at 21+ days out)
• Bedbank wholesale inventory (outside their pricing loop)
• Package rates (flight+hotel bundled)
• Last-minute distressed inventory

These are precisely the rate types where the biggest savings live.

How to beat the AI

You do not beat the algorithm by searching more. You beat it by searching in a channel the algorithm does not actively re-price — which is what wholesale-feed-aware tools like Vacayos do. The bedbank net rate is refreshed on a slower cadence and does not react to your session signals.