Bloggo back to the blog
Risked Based Testing & Web Analytics-->
Web analytics involves collection, measurement and analysis of metrics related to end user activity on a web site. Modern web analytics systems like Omniture, Tealeaf, and Google Webtrends etc. offer powerful tools and infrastructure for measuring and analyzing website traffic patterns & usage. These web analytics systems provide data on various traffic and end user statistics including number of visitors, page views, average session duration, popular pages, common user workflows / click paths etc. Historically, these metrics were primarily utilized for market and business research. Web analytics systems have grown fairly sophisticated over the last decade and the current generation systems can capture more granular and finer statistics for accurately “visualizing” end user behavior patterns & interaction with a web site. These detailed metrics are extremely useful for proactive web application optimization. Web analytics systems are increasingly being utilized by business as well as technical stakeholders for managing, measuring, analyzing and enhancing end user experience.
The power of web analytics can be used by testing teams for understanding end user behavior and patterns. What business workflows & their underlying test scenarios are most widely used and how they are traversed becomes transparent. The test cases constituting the paths are now measurable and the effort quantifiable. Automatically we are building in a risk based knowledge about what is important (by usage) to the end user. We know what is at risk if it is not tested, or tested to a higher level of validation. These statistics can help the test team design robust, accurate and optimized regression suites that will ensure a risked based approach to testing. So in the unlikely event that your project has decreased testing time, you can state facts about risk to the business based in web statistics for tests not being executed.
The number of test cases from a web application gets exploded due some of the following:
• Numerous Application (Entry – Exit) Paths are possible
• Large number of end users, different behavior & from unknown origins
• Explosion in browser/OS combination
• Mobile browsers
Web statistics will tell you where the risks are and when crunch is on and time to test reduced, you can with confidence state the risks to the business.