An early look at new attribute-based shopping for hotels

Attribute-based shopping (ABS) is a hot topic these days for hotels. 

At a recent panel at HITEC, panelists discussed a survey sponsored by NYU and StayNTouch of more than 1,000 travelers. Among others, the survey addressed two of the questions we hear most often: Will travelers use it? And what will it do to hotel revenues?

The survey had some good response data for both, but it did not address the inevitable final questions: Where can I see ABS in action? And who is using it in production today? 

GauVendi is a small company based in Germany with a PMS-attached internet booking engine and CRO module - both are ABS-ready and in production at a small number of properties. 

This product represents a point-solution for just a single property at a time rather than the brand-level solution the CRS majors are building. But its early data could provide early guidance to the questions about whether customers will like/use it and whether hotels will see increased revenue from it. 

For reference, here are some quick facts about the data gathered for this analysis:

• The data set includes 14 properties with 1,893 total net bookings made and consumed during the period of April 1 through August 31, 2022. 
• All properties are located in Europe
• The booking tools enable travelers to select either an ABS path or more traditional options of best available rate, most popular (actually, a rate the hotel wishes to promote) and other room type/rate plan combinations.
• All data is reported by GauVendi from its proprietary database.

Is this real attribute-based shopping?

 ABS has the following characteristics:

• It lets shoppers define the room experience they want by selecting the specific room-level (e.g., view, bedding, room size, high-floor, late check out, etc.) attributes they want to buy
• It guarantees that the room they purchased will have the attributes they bought - the attributes are not “on request”

In the image above, we can see there are several options a hotel shopper can choose, including attributes at the top followed by lowest price, most popular and other room type/rate plan combinations. 

Note that this particular implementation of ABS is different from the shopping cart metaphor often described in discussions of ABS. Here, the shopper selects all the attributes they are interested in before the shop occurs and sees one price for each recommended bundle instead of seeing prices for each attribute as it’s added to a cart.

When the shopper chooses attributes, the shopping engine examines all the rooms to find those that have at least some of the attributes required to find one or more “best fits” for the shopper. 

The shopper can see which options most closely match their list and choose from among them and, as is crucial for ABS, they are guaranteed to receive what they paid for. 

One other significant aspect of this user interface is that it does not present a price for each attribute individually. Instead, hotels are testing providing bundles of attributes they believe the shopper is most likely to choose - with an associated total price for each bundle.

As ABS moves into production with other systems and is tested further, we'll likely see many different implementations as travelers get used to the new way of shopping and hotels collect actual conversion and revenue data on different designs. There is no one-right-way to present ABS options to travelers.

Results to date are encouraging

Now that we've described the customer experience, we can discuss some of the early results to answer our questions. Note that these results are relative to this specific solution. 

Are travelers even interested in ABS to book hotels? Will they use it to book? Will they understand what they're looking at? Survey results indicated that 97% of respondents indicated some level of interest with 37% very interested and 20% extremely interested.

In practice, when presented with this interface, shoppers did indeed understand and choose ABS without any direct prompting. In fact, 17.4% of total internet booking engine (IBE) bookings resulted from the ABS path, about the same number as those who chose the “most popular” option.

Would customers use ABS to book the default run of house room without any additional attributes to get the lowest price, thereby reducing hotel revenues? In the survey, “63% of travelers who pay $251 and more and 48% of those who pay $151 to $250 for their daily guest room rates are willing to pay more for preferred room features.”

In practice, when examining the actual bookings from the GauVendi system, if we index the average daily rate for the non-IBE channels (which includes OTA and group bookings) at 100, the IBE rates indexed as follows:

• Non-IBE - 100
• Best available rate - 110
• Most popular - 118
• Highest value RT/RP - 131
• ABS - 121

Some final thoughts

While this is hardly a conclusive study (we would have preferred to be able to create a control group without ABS to run alongside the standard user interface but none of the hotels they asked were willing to give up the ABS path for a test), we believe it does point to some generally positive directions for ABS. 

When an ABS option is put in front of shoppers, a significant portion of them can understand it, use it and will buy through it. Further, shoppers are not using ABS to try to find a cheaper run of house rate - they are attaching attributes that add up to an average daily rate significantly higher than the lowest price room type/rate plan combination or the OTA rate category.

Additional observations

• Though the initial results are promising, there is ample room for improvement. One big area would be in presenting attributes that are priced individually so the properties can test how much shoppers are willing to pay for each in different contexts. A balcony will have a very different value in the summer vs. the winter; a fireplace might have a very different value in the winter for a couple vs. in the summer for a family; etc. 
• The system optimizes room assignments periodically through the day to free up un-purchased attributes; doing this more often (for example, before each shop) and applying increasingly intelligent dynamic room assignments would likely increase ADRs for ABS bookings as more attributes are made available for sale. 
• Note that this implementation highlights that ABS is a solution for direct channels, which drives more bookings to the hotel booking engine or call center.
• ABS is also difficult for automated rate shoppers to decode. While capturing the standard lowest price or recommended room type / rate plan combinations is relatively widespread, attempting to capture all of the various combinations of attributes available to a shopper would quickly become a daunting task. This makes it easier to avoid detection when offering personalized rates for customers.
• Continued use of ABS will generate lots of new customer-specific preference data that hotels can use to personalize recommendations in new ways. For example, more frequent shoppers who express an interest in ABS might see attribute bundles personalized for them instead of the traditional lowest price and recommended categories.
• All the attribute shopping, pricing, and purchasing data will be a bonanza for revenue managers and marketers. Besides testing price elasticity of individual attributes vs. pre-set bundles, marketers can explore differences in descriptions for each attribute in different contexts. Personalization takes on a whole new dimension with individual attributes in the mix.

It's still very early days for ABS, but there are some encouraging proof points. Let's hope this new shopping method can help travelers get more of what they want and keep them coming (directly) back for more!