By Roland Schegga & Thomas Steiner

In the quest to improve their websites, hotel management ponders innumerable questions. What information do visitors seek on hotel websites and what paths do they follow in order to access this information? Which are the most visited pages? What sites send potential customers to the website and what is the entry page of these visitors? Similarly, which search engines send visitors to the hotel website and what search terms do visitors use on those search engines?

Unlike traditional mass media such as radio and television, the Internet is interactive and digital. These characteristics facilitate measuring consumer responses and hospitality management can draw on data from actual visitors to a website to improve the site design. While research has investigated server log files on one web site, this paper combines log files from 15 Swiss hotels to investigate common trends on how visitors navigate and use hotel web sites.


In total, the 15 hotels' log files yielded 345’440 web site visits and 2’107’397 page views. The hotels received between two and 268 visits per day with an average of 67 and a median of 23 visits. Given the uncertainty on when visitors leave a site, log files track the time on site until the penultimate request. Median visit length per hotel ranged from 60 seconds to 172 seconds, and the average visitor across all hotels stayed almost two minutes. During their visit, the web surfers requested an average of 1.6 to 11.7 web pages, depending on the hotel, and an overall average of 4.7 pages.

Every tenth visitors opens a page and exits without viewing another page. In over half of these one-page visits (56%), the visitors land and leave from the site's home page. Four hotels provided data with the visitor's domain name, such as rather than the visitor's IP address such as In these cases, one can infer the visitor's country of origin; the domain .ch represents Switzerland. For these four hotels, approximately one third of their traffic came from computers in Switzerland.

The homepage is the most requested page, roughly one third of all visits in this sample. The home page is also the most popular "front door"; two out of three visitors enter the site through the homepage. After the homepage, information-related pages are the visitors' main target. Visitors seem to seek general hotel information (news on the hotel, F&B facilities and menus or location of the establishment) or specific room information (room infrastructure or rates). The sites' enhanced distribution and customer relationships functions account for just one out of each hundred requests however, which suggests that visitors are lookers not online bookers.

The referrer field sheds light on how visitors arrive at the site. Over half the cases show no referrer, which suggests familiarity. No referrer means that the visitor typed the domain name directly into their browser or clicked on a bookmark/favourite, a link in their browser's history file, or a link in an email program such as Eudora or Microsoft Outlook. No referrers imply that visitors knew the hotel’s URL, previously visited the site, or followed an email link.

External traffic generators to these 15 hotels were tourism-related websites – local or regional destinations and online intermediaries – and search engines. The small proportion of search engine referrals and high proportion of no referrers suggest that the analysed websites garnered few visitors outside traditional marketing channels. Most visitors knew the site or followed distribution partners' links.

The analysis of search engine referrals shows that Google is the most popular search engine (61%), followed by MSN (10%) and Yahoo (8%). The referral field also includes the keywords that visitors used in a search engine query. The most popular search terms related to combinations of hotel, the city, the hotel's name, the region and activities/events.

Managerial Implications

The results of analysing 15 hotels log files provide practical implications for hospitality managers and promising insights into the behaviour of online visitors. Hotel management must appreciate that a web site is as important as traditional communication channels such as the telephone. Half of the visitors in this hotel panel left the web sites after less than two minutes, even though the high proportion of no referrers and keyword analysis suggest that most visitors already knew of the hotel. Possible reasons for visitors leaving so quickly include disliking the products on offer, disliking the online presentation, or difficulty finding the sought after information. Better web site design would give hotels an immediate competitive advantage via improved retention rates.

As every tenth visitor requests just one page, hoteliers should reflect on how to keep visitors on site. Our results support the idea that visitors to the 15 web sites in this hotel panel sought basic information (detailed information on the hotel, the rooms and the rates). The hotels in this study, however, tended to bury important information for their visitors on low navigation levels, i.e. out of an easy access.

As the home page is the favourite page in nearly one third of all visits, essential information should be available on this top navigation level. As importantly as moving basic information higher, these hotels should move transaction-related information higher than the two clicks found in this study. Popular third-party sites such as and have their transaction-related information on the site's home page.

Acquiring new online visitors would boost site traffic and possible improve the bottom line of these hotels. The small proportion of search engine referrals suggests that the analysed web sites created few new visitors outside the traditional marketing channels. Most visitors knew the site or followed links from distribution partners (e.g. DMOs). Hoteliers should use more online and offline methods to drive traffic to their websites. Offline methods include greater use of their website address in collateral media. Online promotions include optimising their web sites for the major search engines by complying with ranking criteria such as meta tags, content, link popularity or click popularity.

Download the complete version of the paper (PDF 213KB)

Roland Schegg and Thomas Steiner work at the University of Applied Sciences Valais (HEVs), Switzerland.