A device’s anticipated performance may look different on paper than it does within the real-world. This can lead to consumer disappointments, frustrations and negative customer reviews as your product fails to deliver upon its promise. To best prepare your product for true to life scenarios, Intertek works to develop a custom test plan that places emphasis on the aspects of your product that will determine its success.
The key to a successful launch is to ensure that your application can handle the load of all of your new users. The most important aspect of performance testing is defining a metric to identify how performance is evaluated. While performance may be overlooked during the development stages; it’s something that should be a part of it. Developing key performance indicators (KPI’s) is key to any successful performance evaluation.
There can be several KPI’s that are tracked to help create a clear picture; however availability, response time and utilization make up a large portion of users perception.
Our daily tests can help identify availability of a service over the course of hours, days and weeks. Availability can be a key indicator of deeper issues around the overall performance of your application. If users cannot use your product due to a non-responsive application or an exceptionally slow response means they may not return. In the current landscape 23% of users launch an application only once. This means that your application needs to function at 100% every time a user launches it or risk losing that customer forever.
Testing response times with real users prior to launch is an important factor in how your application is perceived. It’s essential to test an increased load to identify any potential bottlenecks. In the application space our experience is that anything over 4 seconds will inhibit continued use of your application. Can you track down these areas and identify them before going to the public? We can help provide user numbers that will imitate a real-world launch while performing standard tasks which can help you identify any potential weak spots within your application or server architecture.
This can come down to basic bandwidth availability. How much can your servers handle at any given time. Evaluating an application based on non-peak times can often mislead teams into thinking their product is ready; however, scaling up to 1000’s or even 10’s of thousands is often needed to help identify how much the product can handle. Simply understanding what the max utilization is can help avoid costly downtime later. Intertek has tools to send as many real or simulated users into your product to help identify max utilization before going public.
Since there are no standards along the performance space it’s important to have a partner who can help define a successful process to evaluation. Simply sending users is not the only option to consider. Defining the process and executing based on expected audience is a key factor in a successful performance evaluation.