Common Mistakes: Start Using Web Site Analysis Tools by Nicolas Buerki Issue===== What are the pitfalls to avoid when starting to measure Web site performance using Web Site Analysis Tools? Response ======== Web site analysis tools can contribute to design decisions to improve visitors' online experience but also to inform site owners, business owners about the performance of their Web sites. However, in many cases, eBusiness Managers, Webmasters or Web operations managers, start deploying Web site analysis tools just as a 'nice to have' tool. Instead of measuring performance and comparing it to the business objectives, they mostly communicate the raw analysis data to senior management, without any further explications or any recommendations for site improvements. To exploit the value that Web site analysis tools can provide, the following major pitfalls should be avoided: 1. Business objectives not set for individual sections of the site: At most organizations, business owners responsible for a sub part of the corporate Web site have not defined their specific business objectives of their sub parts. For example, a sub part of a Web site could be customer support, which can be further broken down into self-service tools (e.g. frequently asked questions, download of drivers, etc.), support contact information, warranties, user manuals, etc. Measuring the performance of a site or a sub part of the site is only valuable, if the measured performance can be compared against the targeted objectives. If business owners do not set business objectives, the analysis cannot determine the site's performance. eBusiness managers or Web operation managers that are in charge of deploying Web site analysis tools need to help business owners defining the detailed business objectives of their Web site's sub parts (read also, "Who Should Set Business Objectives", Steve Telleen). 2. Site owners or site section owners are not trained to understand the analysis reports: At most organizations, site owners or business owners for a specific sub site can not leverage the reports of Web site analysis tools as many of them don't understand the benefits of such a measurement's approach. eBusiness managers need to explain as a first step the benefits and the importance of this performance measurement as integral part of their Web performance measurement program. The second step includes explaining of what is measured and how the measured data can help business owners to further improve their sub-parts of the site (e.g. changing navigation or cross-linking sections, providing and updating section with 'most used links', etc.). If business owners are not sensibilized and/or not trained how to leverage value out of the Web site analysis report, they most likely will not even start reading the reports and all efforts in measuring Web site performance using Web site analysis tools is wasted. 3. Data points of Web site analysis tools are not related to each another: Web site analysis tools measure data points for a defined time period such as number of viewed pages, number of unique visitors, number of visits, etc. However, these data points do not provide any value. For example a high number of viewed pages may indicate that site visitors are lost within the navigation and browse a lot of pages to find what they are looking for versus a lower number of viewed pages, which may indicate that site visitors find directly their information in few clicks due to effective site navigation. It is crucial that Web analysts, relate the individual data points to derive value that can be translated into site improvements and that the value can be communicated to business owners to inform them about their specific site performance. For example to measure the effectiveness of online support, Web analysts should measure and relate the following key performance indicators (KPI): a. Stickiness = total amount of time spent viewing all pages in the support section divided by the total number of unique site visitors in the support section and b. Focus = average number of pages visited in the support section divided by total number of pages in the support section. For online support the stickiness and the focus should be low, which indicates that the navigation to the specific support is effective (e.g. site visitors need few clicks and minimal time to get to a specific support page). It is crucial that the measured data points of Web site analysis tools are related to each another and that these data points are measured for a specific section on the site. If data points are not related to each another and not measured for a specific section, they do not provide any value or at worst if they are interpreted they may lead to wrong design decision. 4. Not enough skilled human resources available to analyze the reports: Web site analysis tools do only track and store the visitor's online behavior. Web site analysis tools cannot interpret the measured data. To get value out of the measured data, dedicated Web analysts need to analysis further the collected data. The main duty of the Web analysts is to relate the individual measured data to each another to obtain KPIs. These KPIs allow Web analysts to deduce the design changes to increase site visitors' experience. In addition, Web analysts need to be able to relate the correlated analysis data with Web site design (read also, "When Not to Use Web site Analysis Tools", Nicolas Bürki. 5. Web site analysis tools are considered as the only way to measure Web site effectiveness: Many organizations rely mainly on Web site analysis tools to measure Web site effectiveness and do not perform any other site performance measurement approach to improve the visitor online experience. To continuously improve a Web site, the Web performance measurement program should include as well, usability lab testing, focus groups analysis, etc. Discover all our research articles About the Author Nicolas Bürki is the founder of www.effinfo.com, a Web research portal providing pragmatic articles to make your Web sites, Intranet and portals more effective. Leading organizations worldwide use effinfo's research and opinions to make Web-related decisions; our documents are regularly among the top 10 most read research documents on ITForbes.com and PcPro.co.uk . We are also consistently quoted on a wide range of IT portals. Nicolas Bürki |
|
