Half of hoteliers now use AI on price. One in ten would let it decide

AI has arrived in hotel pricing, but almost nobody is letting it sign off. That is the short version of RoomPriceGenie's 2026 State of Revenue Management Report, out on 20 October, based on 450 senior people with a say in revenue decisions across the UK, US, Canada, Mexico, the DACH region, Spain, Australia and New Zealand.

Some 53% now use AI in some form to support pricing decisions, with 16% describing their use as extensive. Another 35% are piloting it or weighing it up. Only 11% say they are not considering it at all.

Trust is the more interesting number. Some 77% trust AI to at least some degree. Just 10% would let it make most pricing decisions on its own. The rest treat the output as a recommendation to be checked, which is a long way from the automated pricing the category has been selling for a decade.

The finding underneath the AI story

Strip out the AI framing and there is a plainer result in here: 43% of the hotels surveyed have no revenue management system at all.

Of those, 14% price by instinct or by fixed seasonal rates. The other 29% work from spreadsheets and historical data with very little automation. That is a substantial share of the market setting rates with tools that cannot see tomorrow's demand.

It shows up again in how hard the information is to get. Only 30% say gathering the pricing data they need is quick or easy. Among hotels that already have an established system and a strategy to go with it, that rises to 47%. The tooling gap and the information gap are the same gap.

What hoteliers say drives their pricing is conventional enough: 52% rate demand forecasts as very influential, and 51% say the same of competitor rates.

Who actually does the pricing

Some 85% agree revenue management is essential to profitability. Rather fewer have anybody dedicated to it. A revenue manager has primary responsibility for pricing at just 24% of properties, and 58% say a dedicated role is very or absolutely essential.

The report puts the median point at which a hotel hires dedicated revenue expertise at 146 rooms, and says properties become significantly more likely to do so once they pass 50.

Hiring is difficult where it is attempted. Some 47% struggle to find qualified candidates, rising to 66% in DACH and 56% in Canada. At the same time 70% believe revenue management expertise is growing across the industry, which is an odd pair of findings to hold together, and the report does not resolve it.

Chas Scarantino, RoomPriceGenie's chief executive, said "adoption alone isn't enough" and that margins are "tighter than they've been in years". David Heigl, the chief product officer, argued that automation should keep hoteliers in control and that AI should explain its reasoning in plain language so a whole team can follow it.

Read it with the source in mind

RoomPriceGenie sells a revenue management system to small and independent hotels. A finding that 43% of the market has no RMS is a market-sizing statistic and a sales case at the same time. That does not make it wrong, and it matches other recent research, but it is worth knowing whose survey you are reading.

The report does not publish its methodology, sampling approach or margin of error, and the way the AI adoption figures were reported in trade coverage this week did not consistently reconcile, so treat the individual percentages as indicative rather than precise.

Two other studies point the same way. H2c found 69% of hotels still enter guest preference data by hand, against 19% using AI extraction. The State of Distribution 2026 report from NYU, RateGain and HEDNA found more than 80% of commercial teams spend one to two days a week producing reports manually. The pattern across all three is that the data work has not been automated, which is a reasonable explanation for why so few people trust an automated price.

The full report is on RoomPriceGenie's site.