As part of its cross-channel brand strategy, KIABI websites use a range of services to promote a quality shopping experience. These services were interconnected with many hosted systems—such as secure payment solutions and loyalty point programs—and it was important that page display time remained rapid and errors be identified and repaired as soon as possible.
In parallel, AppDynamics worked on an ROI study with KIABI. An internal audit was conducted with technical and business teams including marketing, commercial operations and customer relations to collect metrics to develop plans for increased profits. It was clear from this study that the AppDynamics tool could reduce the time spent on troubleshooting by 50% within 6 to 9 months after the implementation of the solution."At the end of this POV, which was more than convincing, we decided to implement AppDynamics solutions on the French site www.kiabi.com in the month of February," said Théry.
Java APM solutions were deployed for monitoring application servers, and Browser Real User Monitoring (RUM) was put in place to monitor the browser behavior on the user-side of the site. [C1]The AppDynamics Application Intelligence Platform utilizes a dynamic approach to performance management, learning and adjusting to trends. After a phase of collecting metrics, these adaptive tools set dynamic thresholds that make it easier to spot outliers. For example, a warning system could be programmed to alert of heap saturation in the Java Virtual Machine (JVM), slower web page response times, or degradation of call times for certain services such as the payments or package delivery tracking KIABI programmed alerts in servers and application layers, taking care that the metrics monitored had a direct impact on improving the business.
"Six months after the implementation of these solutions, we are already seeing significant gains at the operational level. At each start of production of a new version of the site, the Build and Run teams are much more proactive when it comes to the behavior of applications. The time spent on diagnosing problems has decreased considerably, and the indicators provided by AppDynamics have already improved the display time of the main pages of the site," said Théry.
According to the ROI study, the goal to cut diagnosis time in half should be reached in the Fall of 2014.