Consumers use numerous ways to return to Web pages, mostly via the Web browser's back arrow (Cockburn, McKenzie, & JasonSmith, 2002). Browsers also have a history list for use both within and between sessions. Users can also find and click on a link to previously visited pages, or retype the page's URL. Another option is to add a bookmark (Netscape Navigator and Mozilla Firefox), or favorite (Microsoft Internet Explorer), to their browser. Once the user has added the bookmark, the user can go directly to the page's address from the Web browser menu bar.
The Internet often resembles other offline communication media but its added flexibility complicates placing bookmarks in a historical context. For example, the layout of a Web page tends to resemble a magazine page. That it is user driven makes the Internet somewhat analogous to the phone system. Since it is software-based, this tends to make the Internet more flexible than other communication media. In communication terms, what does a bookmark resemble?
Historically, media have evolved towards more end user control. Radios and televisions let users pre-program their favorite stations and a VCR can return to one's favorite show. Digital television (Lekakos & Giaglis, 2004), digital radio and digital VCRs such as Tivo (http://www.tivo.com) add even more pre- and post-programming flexibility. Telephones often have speed dial capabilities, or directories of commonly dialed numbers. In addition to analogies with radio, TV and telephones, bookmarks resemble magazine and newspaper subscriptions. In addition to giving the user more control, subscriptions give the publisher guaranteed sales and a front-loaded revenue stream. With a universe of sites to choose from, bookmarking reflects loyalty and benefits a website (Nielsen, 1997).
The Benefits of Bookmarking
Users create bookmarks for several purposes (Abrams, Baecker, & Chignell, 1998), such as reducing the cognitive and typing work required to return to a site, remembering sites, and faster access to previously visited sites. The user interface for creating and maintaining bookmarks though, is clumsy. Bookmark collections quickly grow disorganized and cumbersome for most users (Abrams et al., 1998; Cockburn & McKenzie, 2001; Jones, Bruce, & Dumais, 2001; Pitkow, 1996). Thus academics and industry continue to research better Web browser navigation systems (Cockburn et al., 2002; Cockburn, Greenberg, Jones, McKenzie, & Moyle, 2003; Jones, Dumais, & Bruce, 2002; Kaasten & Greenberg, 2001; Tauscher & Greenberg, 1997).
Regardless of the user interface, bookmarks benefit the site. Taking the concept of loyalty online, e-loyalty or site loyalty resembles the offline concept of encouraging repeat store visits (Gommans et al., 2001; Holland & Baker, 2001). The results of a study using the power law of practice suggest a correlation between repeat website visits and future purchases (Johnson, Bellman, & Lohse, 2003). Bookmarks are a valuable tool for developing relationships with loyal visitors that return to a website (Nielsen, 1997).
Even after three decades, the importance of relationship marketing (RM) continues to gain attention as marketing turns away from a view based on individual transactions and moves towards an assessment of the lifetime of value that a customer can bring to a firm. This marketing method built on relationships, networks and interactions has a dual focus: getting customers and keeping customers (Berry, 1983; Grönroos, 1994; Newell, 2000; Wang, Head, & Archer, 2000). As acquiring customers is more expensive than keeping customers, relationship marketing focuses on the latter; good customers are more profitable and easier for relationship building. There is a strong correlation between customer retention rates and profit (Newell, 2000; Reichheld, 2001; Reichheld & Shefter, 2000; Szymanski & Henard, 2001).
The Web's interactive nature helps facilitates relationship marketing and customer support to a greater degree than traditional media (Hoffman, Novak, & Chatterjee, 1996; Newell, 2000). This interactivity shifts customers away from being passive receivers of information and towards being active searchers of information (Hoffman & Novak, 1996; Kotler, Jain, & Maesincee, 2002). Thanks to this shift, the balance of power is migrating away from businesses and “to the consumers, who can now define what they want in the way of customized products and services, prices, distribution channels, and even advertising and sales promotion” (Kotler et al., 2002, p. 8). Firms must manage customer relationships as the Internet can turn customers into partners.
Given the increased competition for more active customers, the importance of relationship marketing, and that corporate websites are commonplace (Porter, 2001), businesses want customers to bookmark and return to their website. A feature on some sites is a button, text or link suggesting “bookmark this page.” Microsoft and AOL for example, charge companies for entries in the default bookmark file that comes with a newly downloaded browser. All of this suggests the value of having an entry in consumers' bookmark files.
In perhaps the first survey of bookmarks, Pitkow (1996) asked users in his semi-annual “WWW Users Survey” about their bookmarking behavior. Over nine of out ten of the 6619 respondents had bookmarks, and eight out of ten respondents noted that bookmarks were a strategy for finding information. One out of three, however, noted maintaining bookmarks as a major usability problem.
A survey of 322 Web users, also in 1996, found similar results (Abrams et al., 1998). Almost 19 out of 20 respondents had bookmarks, ranging from under ten to over 300 bookmarks. An analysis of the bookmark files from client logs suggested a disorganized, growing mess. Users added a bookmark about every five days but rarely deleted bookmarks. One in three users never organized their bookmarks, leaving them in the chronological order of when added and those with less than 35 bookmarks generally had no folders for organizing their bookmarks.
Finally, the users complained that titles on their existing bookmarks failed to describe the Web page's content (Abrams et al., 1998). Titles, part of a Web page's html code, play an important role in communication and remembering the page. Titles are often the main reference to pages and used in navigation menus such as bookmark lists and history lists (Nielsen, 2000, p. 123).
An online survey this century, investigating bookmark use and website use, found several significant relationships and underscores the importance of bookmarks (Thakor, Borsuk, & Kalamas, 2004). The use and organization of bookmarks show a negative relationship with using search engines. Bookmarks use shows a positive relationship with Web experience. Finally, the number of bookmarks shows a positive relationship with Web experience, Web usage and orientation towards online shopping.
Using a different methodology — analyzing the logs from individual's computers —Cockburn and McKenzie (2001) studied 17 users over a four-month period in late 1999 and early 2000. As was found in earlier research (Abrams et al., 1998; Pitkow, 1996), over nine out of ten users had bookmarks. They also found varied and disorganized bookmark usage. In their sample, the number of bookmarks ranged from one user with no bookmarks to two users vying for top spot with 587 and 565 bookmarks each. The average number of bookmarks was 184, stored in 18 folders. One user had 130 and another had 90 bookmarks in the top level — none of these bookmarks were in a folder.
The users added bookmarks much faster than they deleted or updated their bookmarks (Cockburn & McKenzie, 2001). Over the four months, users added an average of 28 bookmarks and deleted four. Of the existing bookmarks, one out of four pointed to a page that no longer existed, one out of 20 was a duplicate and one out of 20 bookmarks had no title. Users did, however, periodically re-organize their bookmarks by filing top-level bookmarks into folders.
Although the top three pages accounted for almost one fourth of all page visits, just six out of the 17 participants followed a bookmark to reach a top three page (Cockburn & McKenzie, 2001). The leading technique, practiced by 14 participants, was setting a popular page as the default home page. Five participants used another technique, the personal toolbar, to reach one of their top three visited pages.
This tendency to revisit Web pages rather than visiting new pages is the recurrence rate (Tauscher & Greenberg, 1997). Previous research using the same methodology on two separate data sets found a 60% recurrence rate (Cockburn & McKenzie, 2001). Applying the same technique Cockburn and McKenzie (2001) found an 80 % recurrence rate, which suggests evolving Web use. Users spent less time looking for new pages and more time looking at previously visited pages. This evolving Web use suggests that online customer loyalty and bookmarks become increasingly important (Newell, 2000)
Although from 92% (Pitkow, 1996) to 94% (Abrams et al., 1998; Cockburn & McKenzie, 2001) of users have bookmarks, using bookmarks may be a low priority. Of all possible browser actions — including reloading a page and opening a new window — opening a URL accounted for about half the actions. The other popular actions were clicking the back button about one third of the time and opening the home page about one time in twenty (Tauscher & Greenberg, 1997, p. 111-112). For opening a URL, following a link accounted for 83% of the actions followed by typing in the URL (7%) and using bookmarks (5%).
Two final studies used the third methodology, dynamic observation, filming and observing users talking aloud while performing Web tasks (Jones et al., 2001; Jones et al., 2002). Bookmarks were one of ten ways for revisiting Web pages, along with: e-mailing the URL to themselves or others, printing or saving the page, pasting the URL into a document or personal website, searching, typing in the URL and using the browser's history function. Their second study found marked differences in the ways that managers, researchers and information specialists kept Web pages (Jones et al., 2002). For example, managers always e-mailed URLs to others while information specialists never e-mailed URLs to others.
These quantitative and qualitative studies highlight academic and industry interest in bookmarking. Over nine out of ten users have bookmarks and use their bookmarks about one time in twenty to open a URL. Web browsers and Web browsing continue to evolve, with methods to revisit Web pages important. Microsoft for example highlights “Favorites” as a feature on their latest Internet Explorer, version 6 (http://www.microsoft.com/windows/ie/evaluation/overview/default.asp). Previous research, however, approaches bookmarking from the user's perspective rather than the site owner's perspective. A user making the effort to bookmark a site reflects intent to revisit the site.