What is Apdex?

Every business wants to ensure its customer experience (CX) stands apart from the crowd with fast, reliable digital applications and services. Apdex, short for application performance index, is a useful method for measuring user experience for just about anything with a digital interface. Apdex helps organizations to stay competitive by delivering continuously optimized, high-quality digital applications.

The application performance index, or Apdex, is an open standard used to quantify and qualify user satisfaction in relation to an application's response time. It provides a numeric Apdex score that gives insight into user satisfaction with the performance of an application.

An Apdex score is the numerical measurement of a set number of end users' perceived satisfaction rated on a scale of 0 to 1. It's important to note that Apdex scores are application-level scores with specific thresholds, as opposed to real user monitoring (RUM) metrics used to determine user experience delivered across applications within a single session.

Apdex scores are based on application response time metrics, including request per second, data in and out, average response time, peak response time, and CPU usage. These metrics are converted into a single numerical value representing the end user's perceived satisfaction with an application's performance. Apdex scores are bucketed into categories to show if users are satisfied (S), tolerating (T), or frustrated (F) with the application response time across web, mobile, or Internet of Things devices.

An Apdex score is calculated using this simple equation: the number of satisfied samples plus half of the tolerating samples divided by the total number of samples. Written in standard equation format, it looks like this:

Apdex = (Satisfied Count + Tolerating Count / 2) / Total Samples

The resulting score will be a sum in the range of 0 to 1, with 1 being the highest user satisfaction rating (all users are satisfied), and 0 being the lowest rating (no user is satisfied).

At the start of the evaluation, an Apdex threshold (Apdex T) is set. Apdex T represents the time above which a user transaction is considered tolerable. Apdex T values can also be defined separately for a server, browser performance, or specific transactions. The satisfied, tolerating, or frustrated user response rating used to calculate the Apdex score are defined in relation to Apdex T as follows.

1. Satisfied. A satisfied (S) rating results from a fully productive and satisfied user not impeded by an application response time that is less than Apdex T.

1. Tolerating. A tolerating (T) rating results when a user continues the transaction though there are performance lags in the session and a response time above Apdex T.

1. Frustrated. A frustrated (F) rating results when a user abandons the process due to poor performance with response time much greater than Apdex T.

To illustrate further, below is a hypothetical set of 100 end users of an application. Apdex T is set at 400 milliseconds (msec). Any response time less than 400 msec will be given a satisfies rating. Out of 100 users, the following results are measured.

• 60 satisfied measurements (60 X 1 = 60)

• 20 tolerating measurements (20 X 0.5 = 10)

• 20 frustrated measurements (20 X 0 = 0)

The Apdex calculation would be 60 (satisfied) plus 10 (half of the tolerating) divided by 100. This would result in an Apdex score of 0.7.

Default Apdex ratings are based on application-specific thresholds. On a range of 0 to 1, an Apdex score of 0 indicates no users were satisfied, whereas a score of 1 indicates all users were satisfied. In general, a score range of 1 to 0.94 is considered excellent, a score of 0.93 to 0.85 is good, and scores that fall in the 0.84 to 0.70 range are fair.

The Apdex score 0.7 score in our example above would be considered fair.

Apdex scores of 0.69 to 0.49 are considered poor, with any score lower than 0.49 considered unacceptable. Low Apdex scores may result in frustrated customers who could abandon the application or potentially overload a call center with service complaints.

Apdex scores give ITOps, DevSecOps, and other teams the ability to track and improve an application's performance over time by understanding how it impacts end users. It can also be used to define service-level agreements (SLOs) and help development and other teams stay in compliance with SLOs.

An Apdex score is used solely to measure perceived end-user satisfaction levels. Monitoring Apdex scores for different user segments can help define business priorities and identify critical issues that need attention to improve customer satisfaction.

The most effective way to improve an Apdex score is by defining, measuring, and setting targets for user experience levels.

To monitor an application's response times and evaluate improvement (or deterioration) of Apdex scores, organizations can define an Apdex T threshold for each monitoring check and then track an application's performance against this set benchmark.

Once an organization has the context behind their Apdex score and has defined SLOs, they can implement quick improvements and fixes to regularly update software with agile practices.

Apdex scores are a valuable tool for quickly gauging end user experience with a company's applications and services, but response time metrics don't provide a complete picture of CX. Additionally, different applications have different user segments and varied expectations depending on the intended purpose, which affects Apdex T thresholds and related user ratings.

RUM goes beyond Apdex to include the collection of metrics on each user action within a session while the application is in use. This might include tracking the user's path through a sales conversion funnel or detecting differences in an application's performance when used on different mobile devices.

With Dynatrace, you have the full flexibility to adjust Apdex thresholds within RUM based on your business's unique requirements. Learn how Dynatrace helps to capture full visibility into end-user experiences, including Apdex scores, in the on-demand Observability Clinic, Digital Performance 101—Observe, analyze, and optimize your End User Experience.