Getting a Handle on the Impact of Search Marketing

April 26, 2011 in Journal by Hotel Compete

A few weeks ago we published a piece about Online Travel Agents (OTAs), and their search marketing habits.  As you can see, it prompted some discussion in these pages.  In that article we suggested that while search marketing strategies and their impact receive a lot of commentary, they remain under-researched by the hotel industry.  We undertook to publish research based on new industry data, which we will do, starting with this post.

There are numerous measurement steps that we must take to understand this topic.  First, since there is a finite amount of space available on search engine pages for advertisers, we must identify who is currently occupying that scarce real estate.  And Hotel Compete has a methodology for measuring just that. 

Hotel Compete collects for each individual hotel the results of web searches on each of the major search engines.  We base our analysis on the four search terms most frequently used by hotel bookers:  Hotel name; Hotels in City/State; Hotels in Neighborhood and Brand in City/State.  For each of these searches, we analyze the results that are returned on page one, using automated text recognition software to identify when the search results are a match for the hotel.  

In this analysis, we applied the methodology described above to every hotel in the top 25 US markets.  Our objective in each case was to identify the quantity, source and page rank of the search results, as well as the source of the URLs in the search results (ie to understand whose links are appearing in each hotel’s results).  The sample included 3,785 hotels, which would generate an unwieldy number of individual URLs, so for this analysis we stratified the results to six URL types:   

• Brand: Any URL that directs to brand web page
• Hotel: Any hotel-specific URL (ie brand or individual hotel) that takes the user to the hotel’s booking or information page
• OTA 1: Major OTA (eg Expedia, Orbitz, Priceline, Travelocity)
• OTA 2: Secondary OTA (eg Hotelscombined, IHS Advantage, Booking.com, etc)
• OTA 3: Non-booking sites (eg Kayak.com, TripAdvisor, TravelGuides)
• Other: All others – primarily includes local results (eg yellowpages.com, yelp.com, etc)

The following tables present results for Branded and then Non-Branded hotels for the week ending April 15 2011.  “Branded” in this research indicates hotels belonging to a single brand with 15 or more properties nationally.  In each case, we have separated paid from organic search positions, measuring both the number of returns and their Average Position, where 1 is the highest position on either the paid or the organic portions of the display.  The average for paid search is based on position within ~6 spots and ~11-13 spots for organic results.   

Analysis Results - Week Ending April 15, 2011

The most obvious insight from this analysis is the dominance of brand URLs in search results for branded properties.  Despite only accounting for 38% of the hotels in the sample, the absolute number of paid search results is a huge multiple of the equivalent number for non-branded properties.  The average page rank is also telling: the brand score of 1.7 suggests that brand sites own positions 1 and 2 in paid search, but are less prominent in organic results. 

In other words, the prime paid search positions on major search engines are occupied far more frequently for branded hotels than independents, and it’s usually the brand site that’s occupying them. 

Interestingly, however, the non-branded hotels direct a higher proportion of users straight to a hotel-specific booking site.  The overwhelming majority of hits for branded hotels included a non hotel-specific URL, confronting the booker either with multiple options, or the need to complete another search within the brand site.  Either way this leads branded hotels to compete against one another in the same city.  Whether or not this practice continues to be successful in a world where searches are becoming increasingly targeted remains to be seen. 

Of course, this analysis does not tell the whole story.  Bookers do not make decisions based on aggregated market data: they make decisions based on where – specifically – they want to stay.  The tools that today’s bookers use to search for hotels make it easy to pinpoint hotels based on their location, price and other key variables.  To understand the impact of search on performance we must characterize it not just in terms of the whole market (as we have here), but in terms of the local competitive set.  That is a whole new can of worms, and one that we will open in our next blog post.