How autopilot AI pricing lifts revenue per square meter by 13%

When a tech vendor claims its product makes hotels more money, that claim is notoriously hard to prove. A hotel rarely changes one thing at a time: a new booking engine, a management change, a shift in the market can all move the needle at once.  

On the latest episode of Matt Talks Hospitality, Mews CEO Matt Welle explores one of the most powerful tools available when it comes to boosting revenue: Mews RMS. Alongside some colleagues, they explore how it works and crunch the numbers on its real impact for hoteliers. 

Finding a fair comparison 

Richard Smithies, Mews's Director of Data, set out to solve this problem. Working alongside Conor Winders, Mews's VP of Product and Engineering, Richard analyzed more than 6,000 hotels to isolate the effect of a single system: Mews RMS.  

Rather than the simple before-and-after comparison most vendors rely on, which mistakes seasonality or market momentum for product impact, the study used causal inference. Each hotel that adopted Mews RMS was matched against a hotel that hadn't: similar market, similar type and a revenue trend that tracked in parallel before adoption. That match is what lets the study isolate the system's effect from everything else happening at the same time. 

The magic number nine months in the making 

Looking at all Mews RMS customers together, the uplift was real but not statistically strong enough to stand behind. And the last thing hoteliers need are more exaggerated claims and false promises. 

They dug deeper to isolate the one variable that mattered: Autopilot. This is the mode in which Mews RMS sets prices without a human approving each change. Richard is confident in what the data showed: "Hotels with the autopilot feature switched on over a period of nine-plus months had a 13% increase in revenue per square meter over an 18-month period versus hotels that didn't have an RMS." The odds of that result happening by chance are seven in ten thousand. 

Nine months isn't when the benefit starts – it's just when Mews can prove it. Revenue takes time to show up on the books after a pricing change. Some hoteliers ease into the feature gradually rather than switching it on immediately, and the algorithm itself needs a learning period, running thousands of pricing experiments across markets before it reaches full strength. 

What Autopilot actually changes 

The scale of that experimentation is hard to match manually. Hotels in the study made roughly 120 price changes a month before adopting Mews RMS. Reviewing and manually accepting the system's recommendations pushed that toward 500. Full Autopilot took it past 1,700, roughly 14 times the starting point, and for the top quartile of adopters, more than 4,700 changes per month. 

Richard is careful not to discount the importance of revenue managers, who understand market context an algorithm can't. But, as he put it, "it's almost impossible for a human to be monitoring the market every five minutes." 

That doesn't mean Autopilot runs unchecked. Hoteliers set the limits it operates within. You can have minimum boundaries, maximum boundaries that will keep the algorithm in check with what you, as a revenue manager, want it to do. 

Why RevPAM, not RevPAR 

The study measured revenue per square meter rather than RevPAR. Mews RMS lifted average daily rate and occupancy at the same time, an unusual combination, because it avoids throttling either one. It can push rates up when demand is genuinely there and hold occupancy when it isn't.  

More guests in the building means more spend in the restaurant, bar and spa, not just the room. It's also the metric that scales with where Mews RMS is headed: group pricing, meeting rooms, parking and eventually anything in a property that carries a price. All priced from inside the same system a hotel already runs on rather than a spreadsheet on the side. 

Closing the trust gap 

Only 55% of Mews RMS customers currently run full Autopilot. Conor says there’s one clear reason for that: "Ultimately, a product like this does come down to trust."  

To help build trust with revenue managers, the system now shows its reasoning. It surfaces whether a recommendation was driven by a competitor's move, a demand surge or a seasonal pattern, so hoteliers can see the logic before they hand over control. 

For existing Mews customers, getting started now takes a matter of hours. You no longer need months of data validation that an RMS once required, because the pricing engine runs on data that's already inside the same system rather than one sitting on the side.  

It’s perhaps the easier way to power up your pricing and boost revenue without adding to your workload. Want to learn more about Mews RMS and how it works? 

Watch the full episode