[Webinar] Discover key trends and best practices in Kubernetes observability with DevOps expert, Viktor Farcic.Register now
Our recognition in the inaugural Gartner MQ for DEM

Apdex Score

Apdex (Application Performance Index) is an open standard to measure the user satisfaction regarding a web application. It is a metric that provides a single score ranging between 0-1 (0 = no users satisfied, 1 = all users satisfied), giving business application owners an insight into the measure of their customer happiness and satisfaction levels.

Apdex Score, as a whole, is critical to measure the service levels and customer satisfaction which in turn measures the business growth. Moreover, these values are easy to decipher; unlike the traditional values of average response time and throughput, that does not accurately interpret a particular transaction that is performing slow and affect user satisfaction.

The Apdex Score is calculated using the following universal formula:

              Satisfied Count + Tolerating Count

 Apdex =                                              2                

            Total Samples

Easy to calculate and interpret, the data collected over a period of time are converted into a simple index based on the application responsiveness. Application responsiveness is categorized into three zones based on the Apdex score:

  • Satisfied: This represents the time value (T seconds) below which users are not impeded by application response time. Depicts the user is fully productive.
  • Tolerating: This represents response time greater than T (precisely, T to 4T), where the user notices performance lagging but continues the process, which depicts the response is tolerated by the user.
  • Frustrated: This represents response time greater than 4T which is unacceptable, and users may abandon the process, which depicts the user is frustrated.

where, the value T can be defined by the application owners.

The score of 1 show all the users are satisfied with the application performance, whereas a score of 0 show no users are satisfied. Score of 0.5 shows all the users are tolerating the application performance. As the application responsiveness vary, the score ranges from 0 to 1.

Example

Let us say 500 ms is configured as the Apdex threshold. In this case,

  • If the response time is less than 500 ms, it will be counted as Satisfied.
  • If the response time is between 500 ms and (500*4) = 2000 ms, it will be counted as Tolerating.
  • If the response time is greater than (500*4) = 2000 ms, it will be counted as Frustrated.

This calculation is then followed by getting to know the number of visitors for the corresponding response time. For a 5 minute poll interval, we get 288 data points per day. Say,

  • If the response time is less than 500 ms for 40 times, then the satisfied count will be 40.
  • If the response time is between 500 ms and 2000 ms for 200 times, then the tolerating count will be 200.
  • If the response time is greater than 2000 ms for 48 times, then the frustrated count will be 48.

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

= (40 + (200/2))/288 = (40 + 100) / 288 = 140/288 = 0.486 ~ 0.49

Therefore, the Apdex value obtained shows that it is time to speed up your website and improve its usability.

Back to Top