In 2009, Chris Anderson, an associate professor at Cornell University, published a study about Expedia and hotels that was a sensation in the industry.
When four hotels in the JHM Hotels US chain appeared on the first page of destination results of Expedia, the giant online travel agency (OTA), they saw a rise in reservations through their own websites.
The hotels had between a 7.5% and 26% uplift in direct bookings as part of a so-called "billboard effect" -- as if merely appearing on Expedia acted as a billboard for the properties in general.
In April 2011, Anderson published a larger study for Cornell's Center for Hospitality Research that found the same effect.
This more substantial study looked at 1,720 reservations for InterContinental Hotel brands for the months of June, July, and August over three years (2008, 2009, and 2010).
In this case, for every commissionable hotel reservation that came through Expedia, there were between three and nine reservations at an IHG-related website that were influenced by the hotel having been listed on Expedia.
In the study, a large majority of customers that booked directly on a hotel brand website visited Expedia specifically beforehand.
When the study appeared, Expedia trumpeted the results. After all, the billboard effect suggested that its commission rates were better value than they might at first appear.
Listing a hotel on Expedia was doing double duty as a kind of search engine marketing tool for hotels' own websites.
"Yes, it's still relevant"
Anderson is not aware of any other billboard studies done since his. He said in an interview:
"Comscore and Compete quite often do attribution studies for individual firms, where they cookie customers and track ad effectiveness etc... So, somewhat similar, and broadly endorsing of my work.
I think what we are seeing is a lot of consolidation in many areas of online travel with the big players getting bigger - for example Travelocity is now basically an affiliate of Expedia, as Expedia powers its display.
So you have Expedia and Booking.com growing in brand awareness. These larger players (in the eyes of consumers) are the place to shop to find the best prices. As a result the billboard effect intensifies."
"Not so fast"
While generally lauded for their good-faith thoroughness, the studies' conclusions weren't universally endorsed.
A key point: there was roughly as large of a billboard effect when hotels were presented on airline websites, rental car websites, and other sites as when they were displayed on Expedia.
That's according to a 2011 review of the Comscore data used in the second billboard effect study by Cindy Estis Green, CEO at hospitality analytics consultancy Kalibri Labs. (The data was shared with her.)
In an interview, Green noted:
"A billboard effect appears to be enjoyed by other channels, too. Consumers may visit and value those sites more than OTAs. And those channels don't charge commissions averaging 15% to 25%."
Green elaborated on her interpretation in a free e-book that she co-authored with Mark V. LoManno, Distribution Channel Analysis: A Guide for Hotels, which was published by the HSMAI Foundation, the research arm of trade group Hospitality Sales and Marketing Association International.
It would have taken a broader, differently constructed study to sufficiently test all the variables involved in the complex issue of identifying and appropriately crediting each of the many touch points that lead to brand.com bookings.
A diminishing effect?
In 2014, the key question is if the billboard effect is still relevant. That question matters because hoteliers feel their spending on commissions has been outpacing their total revenue growth.
Moe Ibrahim, CEO at Journeyful, a US start-up with an online hotel booking platform for hotels aiming to cut middlemen out, has this to say:
"Perhaps the billboard effect theory held some weight in the early days of OTAs where the consumer might have preferred to book direct, but the reality is today, consumers do not want to book direct.
The consumer believes booking direct is expensive. William Shatner told them so. And sometimes, it's true, because hotels are behind the technology curve and use the OTAs discounting tools wrongly."
Max Starkov, head of HeBS Digital, a hotel marketing services provider, says he has never been a big believer in "the so-called billboard effect".
Implying that a single touch point -- namely, an OTA site -- plays the most decisive role in influencing the hotel booking decision is simply ludicrous.
What about all the other super powerful touch points: TripAdvisor, social media, search engines?
The rise of metasearch is another factor, adds Flo Lugli, of the consultancy Navesink Advisory Group and formerly a marketing executive at Wyndham Hotels. She notes:
"With TripAdvisor adopting metasearch and Kayak growing rapidly, more and more consumers are starting their research with metasearch rather than an OTA.
So if the studies were redone today, I don't know if it would point to the same conclusions."
Robert Cole, who runs the RockCheetah consultancy, says the effect isn't substantial enough to have prompted hotel owners to figure it into their channel management strategy.
"I haven’t run into any hotels that have really started to calculate out the margin the OTA is getting and factor a share of their direct bookings as some sort of media buy."
Attribution gets harder in the mobile era
Anderson at Cornell wonders aloud about how mobile might change the billboard effect.
"As phone/tablet usage grows, we may see a different story.
Let's say consumers switch to having higher conversion ratios on mobile devices. Then they probably won't shop in an app and then call the hotel directly.
They will instead shop and purchase via the app. So the billboard effect may be reduced."
"Chances are users won't download hotel apps because they'll be too specific, unless. of course. the customer is chain loyal and it's a chain app.
So, the trend of increased mobile bookings will benefit the OTA and metasearch apps."
The mobile revolution also favors the intermediaries over suppliers, says Cheryl Rosner, past president of Hotels.com and current CEO of Stayful, an intermediary platform for bidding and booking boutique hotels in 10 US markets.
"This is a complex issue. But for mobile web, intermediaries have put a lot more thought into the user experience of the full booking process via "tap tap slide", instead of by keyboard and mouse, than hotels have.
That design edge may reduce the temptation among users to click off to a direct site. So we may see the billboard effect diminishing."
Expedia says billboard effect is growing
Adam Anderson, director of industry relations at Expedia, responded by e-mail to the above opinions:
"Attribution is difficult, but it’s very logical to see how exposure across multiple channels and places online can drive hotel-direct bookings.
Expedia, in particular, has grown its presence significantly since the billboard effect study was published, making the exposure that its partners receive even broader.
Expedia partners are featured in more than 150 brands in 70 countries, exposing partners through more than 140 mobile sites worldwide.
Further, the marketing efforts from these brands also exposes partners through other channels, including meta.
It’s reasonable to conclude that many shoppers will use these channels for inspiration, research and shopping and then book direct.
As an aside, Ms. Estes-Green’s point about exposure from airline websites, rental car websites, etc., being a cheaper alternative than an OTA is misleading.
Many or most of these sites are powered by OTAs through their affiliate programs, like Expedia's -- which has 10,000 partner sites."
Defending the original research
The travel industry is notorious for publishing studies that are poorly done, failing to look at statistically representative sample sizes and rarely accounting for the various factors at play.
In that light, the Cornell studies are remarkable for having been carefully constructed. As Anderson notes:
"During the three-month period, the properties were cycled on and off the Expedia.com listings. That is, we displayed a particular hotel on Expedia and then did not display it there. Each hotel’s cycle lasted 7 to 11 days.
When properties were displayed on Expedia.com, they were displayed on the first results page. When a property cycled off, it could not be found anywhere on Expedia.com, even if someone specifically searched for it.
At the completion of the study each property had spent 40 days displayed on Expedia and 40 days not displayed."
While no single study is perfect or definitive, the second Cornell study remains hard to punch holes through.
For instance, what about rate discounts?
Was it possible that the results were skewed, because when the hotels were on Expedia they were offering promotional rates, and that it was the promotional rates that was actually driving the bookings?
Anderson says no:
"As part of the study I also looked at average daily rates (ADRs).
The ADRs during the Expedia on phase - controlling for LOS, DOW etc... were slightly higher (as demand was higher).
What about seasonality? Might the results have been distorted by being, say, 40 days of local low season followed by 40 days during peak travel?
Again, Anderson says no.
I directly controlled for seasonality as I oscillated the weeks of ON and OFF, i.e. this week on, next week off, third week on, etc...
I did this for 12 weeks, I also ensured that I had same number of weekends, holidays etc... in each of the ON and OFF cycles.
A need for more independent research
Perhaps the biggest mystery left in the wake of Anderson's work is the problem of how to attribute the cost of acquisition of a hotel customer. Jan Freitag, senior VP of global development for STR, notes:
"Users are searching across devices, and we don't accurately map their behavior. No one, not even Google, knows how to track this and attribute properly yet. That's somewhat amazing."
NB: Billboard image via Shutterstock.