Skip to main content

Unlocking New Guest Value with AI-Powered Websites Part II: The Pinwheel of Feedback Data

Just as a pinwheel gains speed as the wind increases, when a website is provided with user data and is also oriented with guest needs, the more likely a guest is to spend time on the website and interact with it, creating a positive feedback loop.
a multicolored pinwheel
Advertisement - article continues below

In Part One, we started off with a centuries-long view of when certain innovations cross the chasm to become general purpose technologies and how this then not only leads to exponential growth of certain industries but also changes in the very rubric of social life. That’s a high bar! 

To help explain just what’s possible in the here and now for hotels to meet these lofty expectations, we recruited Frank Reeves, Chief Evangelist at SHR Group, to discuss the company’s next-generation website and integrated booking engine (IBE) platform called What’s fascinating and groundbreaking about is that it’s able to reconfigure pieces of content displayed on a website based on what it knows about the user and all previous interactions with the website.

And based on this reconfiguring and A/B testing of what promotes specific behavioral outcome, the machine learning (ML) engine within can better discern what’s motivating conversions and how best to personalize the website and IBE for future user interactions. This is what we call the pinwheel of feedback data, whereby the pinwheel gains in turning speed as the amount of wind increases. In this case, the wind is the user data, wherein the better a website or IBE is already oriented to a guest’s specific needs at any given time, the more likely a guest is to spend more time on the website and interact with it more, thereby supplying it with more data and more ML – a positive feedback loop.

Hence, the more a user returns to the website, the more the AI can optimize the configuration of content based on how the prospect interacts. This also applies to the IBE, where past reservation searches are stored for the guest to pick up where they left off once they revisit the website. And, of course, the IBE can be set up to continuously A/B test various offers for even more insights about what drives bookings and what drives ancillary spend during the prearrival stage.

Refinement after refinement at each customer click, the pinwheel of ML learning and feedback data means that a website can automatically cater the content to what each individual wants to help move them down the sales funnel and, ultimately, into a finished reservation. Managers can also set up rules to incentivize results, such as having the website display a direct booking discount or F&B credit when a user returns for, say, the fourth time onwards or after two weeks have passed since the first viewing. 

Finally, it goes without saying that all these content optimizations help immensely with a channel shift away from the high-commission OTAs. No third party will ever be able to offer as personalized and as pleasurable a user experience, resulting in reduced costs per acquisition and greater net revenues.

As a digression here, it’s important to consider why many customers prefer the OTAs in the first place. Yes, there’s the convenience of aggregating all properties in a location, but it’s also because of guest frustration with Website hyper-personalization via an AI backbone fixes this issue, really bringing the hotel’s hospitality into the online realm.


Websites for Beyond the Sales Funnel

The value that AI-powered websites will add to the prebooking phase is by itself remarkable because of how it can help brands to micro-segment customers within the sales funnel, so much so that marketers can now accurately separate the upper funnel from the lower funnel or even the ‘upper-upper funnel’ with specific insights at each partial phase on what motivates guests to move in the right direction. 

Why stop at using websites solely to drive guestroom and package sales? What can an AI-powered website do to boost ancillary spend (TRevPAR) or keep the relationship going after departure?

“Take a London hotel that we work with, for example,” commented Reeves. “We observed from both past guest behavior on this brand’s website and from the totality of interactions across all hotels using the platform that travelers who had already made a direct booking and who originated from the United States have vastly different needs leading up to arrival over those travelers coming in locally from within the United Kingdom. Monetarily speaking, past US travelers highly favored F&B content, so prioritizing the display of various dining options for incoming guests from the US resulted in more prearrival F&B revenue per guest and more on-premises utilization.”

To expand on this example a bit further, the true power of ML comes not only from how past users interact with your own website, but also from how all users across all websites have behaved. It’s this combination of macro (systemwide patterns) and micro (feedback on your own that has allowed to attain a minimum efficient scale for any new property to leverage past learnings.

Such a blend of macro and micro opens a wholly novel website functionality: reducing cancellations. Because the system knows who is more likely to cancel a reservation during the prearrival phase based on past cancellation data from all travelers at all properties on the platforms, managers can use this information to proactively send out cheerful reminders in the days leading up to arrival or even send out additional incentives to those ‘risky’ guests such as F&B vouchers.


Mapping Value for the Whole Guest Journey

The analog of a pinwheel spinning across the entirety of the customer journey to unlock value at each stage really works here. Each website or booking engine interaction provides feedback to fuel improvements for future interactions. This can be broken down as follows:

  1. Discovery: apply learnings from past website visitors and systemwide patterns to make a better first impression with a new user, encouraging them to revisit versus book through an OTA
  2. Early Prebooking: rearranging content based on dream phase interaction and use ML to deduce what content best optimizes for continued engagement and conversions
  3. Late Prebooking: further personalize website and IBE content, possibly adding a booking incentive to further encourage a direct sale versus one made through a third party
  4. Early Prearrival: optimize the display of add-ons based on what’s known about a guest in order to maximize revenue on the books, which also helps with staff scheduling
  5. Late Prearrival: defend against possible cancellations through website personalization that accentuates how great the onsite experience will be combined with other one-to-one offers
  6. Onsite: act as a virtual concierge to relieve onsite teams by displaying most relevant information as well as show the most relevant add-ons to further amplify more property usage and TRevPAR
  7. Early Post-Departure: show a thank you note and other departure information to amplify how the onsite experience was perceived, meaning higher guest satisfaction and better reviews
  8. Extended Loyalty: after some time has passed, entice return visits by highlighting what’s new onsite in relation to the context of the guest’s past stay


Optimization Increases Valuation

To close, consider how a hotel can get ‘more juice from the squeeze’ at each point in this chain by leveraging the latest in artificial intelligence and machine learning. Every optimization works to increase automation so that teams can be more productive per unit time while also providing those teams with more accurate insights to inform other operational improvements.

At discovery and prebooking, we’re talking about increasing the conversion rate versus the competition and driving more traffic away from the OTAs, enhancing both gross revenues and net revenues that flow through to net operating income (NOI). At the prearrival stage, we’re talking about TRevPAR and fewer cancellations, both of which also augment NOI. Then for onsite and beyond, we’re talking about how the website can lead to better satisfaction scores which in turn enhance word of mouth and return visits to further lower future customer acquisition costs. 

Taken together, all of this acts to boost the overall economics of a property and its long-term asset valuation. If the static brochure websites of yesteryear are like steam power, then ML deployments truly are the electricity upgrade your hotel needs for the decade ahead.



Together, Adam and Larry Mogelonsky represent one of the world’s most published writing teams in hospitality, with over a decade’s worth of material online. As the partners of Hotel Mogel Consulting Ltd., a Toronto-based consulting practice, Larry focuses on asset management, sales and operations while Adam specializes in hotel technology and marketing. Their experience encompasses properties around the world, both branded and independent, ranging from luxury and boutique to select service. Their work includes seven books: “In Vino Veritas: A Guide for Hoteliers and Restaurateurs to Sell More Wine” (2022), “More Hotel Mogel” (2020), “The Hotel Mogel” (2018), “The Llama is Inn” (2017), “Hotel Llama” (2015), “Llamas Rule” (2013) and “Are You an Ostrich or a Llama?” (2012). You can reach them at [email protected] to discuss hotel business challenges or to book speaking engagements.


This article may not be reproduced without the expressed permission of the authors.

This ad will auto-close in 10 seconds