Hotels are always looking for new ways to sell their inventory - often based on anything other than guest behavior.
They get surprised when these tactics don’t take off, or when they backfire spectacularly. This occurs so frequently, it’s hardly worth mentioning.
Which brings us to attribute-based selling.
As someone who sells hotel rooms for a living, I’m generally in favor of finding better ways to merchandise our inventory. Hoteliers of all stripes should continually look for opportunities to improve the guest experience, offer greater value and, if you can manage it, shift share from higher-cost channels (ahem, online travel agencies) to direct revenue.
I’m just not convinced attribute-based selling meets those criteria.
A number of reasons illustrate why:
- Current customer behavior suggests guests aren’t likely to use attribute-based selling (ABS)
- Hotels frequently have limited variety of distinct, differentiated room inventory
- Most importantly, ABS posses serious risks for rate integrity and long-term revenue capture
- OTAs are much better positioned to benefit than we want to admit
Oh, and all of these ignore the fact that truly effective solutions to this problem already exist.
But I’m getting ahead of myself. Let’s take a look in more detail.
Guest behavior
From a user experience perspective, attribute-based selling (or if you prefer, attribute-based shopping) isn’t a terribly new concept. Guests have long had the ability to filter their choices based on any number of search criteria using faceted search tools on OTAs, chain brand websites and the websites of some smaller collections, too.
However, experience shows that these tools are seldom used in practice. Most guests conduct their research based on their preferred dates, location and price, then tend to look for alternatives to balance against those. What you don’t see is a ton of filtering after the fact.
Yes, I’m aware that faceted search is a significantly limited comparison relative to what ABS enthusiasts propose. But I still don’t see why “better filters” matter when guests don’t seem to care for filters in the first place.
Will some customers use ABS? Sure. But will that use represent a game-changer? Current data suggests otherwise.
When we’re talking about an entirely new paradigm for providing “more choice,” I’d question who’s actually asking for this. Is it guests? Hotel revenue teams? Technology providers? Only one of these ultimately leads to more sales. And that doesn’t seem to be the case here.
Room type differentiation and data
Compounding the issue is that selling distinct products by attribute requires clear differentiation around the various products for sale, as well as the data to support it. While technology providers are clearly willing to provide a platform to move data around in new and exciting ways, many hotels struggle to tell a compelling story to guests around their existing inventory.
There’s a simple reason for this: How can guests choose the attributes that make your rooms truly distinct when so many rooms in the industry exhibit extreme same-ness? I’d argue that we wouldn’t be talking about additional ways to differentiate our inventory if we currently did a better job, y’know, differentiating our inventory.
Yes, hotels exist who have plenty of distinct offerings that can be split apart and repackaged in myriad forms. But it’s in no way the overwhelming majority of rooms. And, even for those who “need” these capabilities, a larger problem looms...
ABS will make component prices transparent, driving down revenues
How much do you charge for breakfast? For late checkout? For ocean views - or at least views that don’t overlook the dumpster out back? Does your guest know? Once you start using attribute-based selling, they will.
Have you ever watched guests shop for their optimal trip? We’ve all seen the stats around the number of searches and sessions guests perform prior to making a purchase. What makes you think they won’t use that to find out exactly how much you think a particular attribute is worth ... then choose the cheaper option for everything that doesn’t matter for their stay?
A guest doesn’t care enough about the balcony, ocean view or breakfast to spend the extra $5 to $50 each? Good for them. They no longer have to choose them.
Worse, when pricing these components individually, you’re teaching guests to shop around to find a better value for the amenities that do matter. Even proponents of attribute-based selling acknowledge guests will see lower prices.
Now, I'm no math major, but I'm not clear how exactly guests will pay less and hotels will simultaneously earn more. Something's got to give here; I'm betting that it's not the one you're looking for.
OTAs can’t react? I don’t think so
As for the idea that OTAs won’t adapt readily? Please. They're not likely to stand by and let significant share walk away. And hotels dependent on OTAs for any meaningful share of their revenue are likely to work directly with the OTAs to provide the needed content or risk losing placement and revenue.
Should they? No. But will they? History suggests yes.
In Simone Puorto's original article on the topic, he stated: “With virtually unlimited attributes available, an ABS-model-based industry could (finally) properly feed predictive models, instead of relying on low-quality, superficial and arbitrary data.”
No disrespect to my hotel industry colleagues intended, but most hotel companies have lagged when it comes to investing in the necessary data science to develop such predictive models.
The OTAs, on the other hand? Not so much. As Robert Cole noted on PhocusWire:
"Big data-powered predictive analytics [used by OTAs] are able to capture demand signals both internally from airline and car rental verticals, and externally from social, news, event, weather, traffic or retargeting campaigns. Extensive data inputs and substantial historical booking data allow for robust testing to identify the highest-quality signals that predict the best probability of profitable outcomes.
"Given less sophisticated big-data repositories and lower-quality competitive conversion metrics, hotel brand websites are challenged to be opportunistic."
Sounds like the OTAs have done the ground work - again. Have you? And, even if ABS merchandises your inventory more successfully, have you done everything necessary to drive traffic to your channels in the first place? The OTAs certainly have.
Better merchandising options already exist
Of course, ABS also ignores two huge - but often underutilized - ways hotels can benefit from improved merchandising that already exist.
The first is upselling. Hotels using upselling typically see an average reservation value increase of 30% or more. And upselling works because you're offering available inventory to those guests who've already chosen you for their stay.
Subscribe to our newsletter below
Close the sale first, then price individual upgrades or package options separately based on availability and interest among target customers.
Second, start looking at personalization in practice. Think about what's most important to your guests and tailor offers to their prior behaviors.
For instance, why do so few hotels offer repeat stays for the room types guests previously enjoyed? Seriously. We've seen consistent benefit by merchandising previously selected rooms and dates to past guests. And it doesn't require an entirely new CRS or PMS to make it happen.
Conclusion
Again, I want to see hotels sell more rooms. And I'd be happy to be wrong about ABS' limited ability to drive increased revenues. There's undoubtedly some market for it. But the reality is it's likely not huge.
It doesn't seem that guests really need ABS. It appears unrealistic to assume that you'll be able to both improve prices for guests and also drive more revenue. And it's certainly untrue that OTAs will stand by and let some new tech bypass them without responding in kind.
If you're looking for ways to improve your revenue, remember the best way to drive more revenues is to respond to your guests' needs.
The question is, does ABS actually work to deliver on that? It sure doesn't seem like it.