Web Analytics Reveal User Behavior: TTU Libraries' Experience with Google Analytics

• Published in: Journal of web librarianship 2013-10, Vol.7 (4), p.389-400
• Main Authors: Barba, Ian, Cassidy, Ryan, De Leon, Esther, Williams, B. Justin
• Format: Article
• Language: English
• Subjects: Academic Libraries; academic library Web sites; Analytics; Assessments; Behavior Patterns; Behavioral Science Research; Computer Software; Computer System Design; Evaluation; Exact sciences and technology; Expectation; Feasibility Studies; Feedback (Response); Focus groups; Formative Evaluation; Google Analytics; Help Seeking; Information and communication sciences; Information science. Documentation; Libraries; Library and information science. General aspects; Library Automation; Library Development; Needs Assessment; Program Effectiveness; Reference Services; Sciences and techniques of general use; Spots; Studies; Surveys; Texas; Use and user studies. Information needs; Use Studies; User behavior; User behaviour; User Satisfaction (Information); Users (Information); Utilities; Web 2.0 Technologies; Web analytics; Web sites; Websites
• ISSN: 1932-2909; 1932-2917
• DOI: 10.1080/19322909.2013.828991

Proper planning and assessment surveys of projects for academic library Web sites will not always be predictive of real world use, no matter how many responses they might receive. In this case, multiple-phase development, librarian focus groups, and patron surveys performed before implementation of such a project inaccurately overrated utility and positive impact. The Web Site Support Team of Texas Tech University Libraries conducted post-assessment usage of a locally developed reference tool using Google Analytics, which revealed a significant disparity between expectations and results. Web analytics tools, while not able to predict users' needs, are adept at describing users' behavior. This user-provided evidence is invaluable for informing the decisions that academic libraries make about their Web sites. While the initial incarnation of the reference tool failed to provide the intended service, Web analytics allowed the team to refine, modify, and integrate elements of the tool into other areas of the Web site, saving the project from being completely scrapped. Other academic libraries and libraries in large organizations are encouraged to use Web analytics and click analytics tools to assess the outcomes of Web projects. Such data can reveal blind spots in predictive usage, which may originate even in the patron base, and can allow modification of projects based upon real user behavior.