This blog is part of a series in which we share some of the results from our State of Performance in Modern Applications Survey. Want to binge read the results? Download the full report here.
43% are held accountable for application performance by measuring end-user complaints, while 31% measure specific KPIs
In a previous post, we discussed how and how often organizations are measuring application performance. Now let’s talk about measuring accountability. When we asked, “How is your job held accountable for application performance?” 43% responded that they measure end-user complaints, while only 31% measure specific KPIs. Now our question is, is that good enough?
Answer: No. If you’re waiting on end-user complaints to know when things go wrong, you’ve waited too long. Most users will simply move on. It takes time to complain, time that most people don’t have, at least not for you. Those that do take the time to complain are really frustrated by the time they do so. You’re on the back foot by then. Measuring specific KPIs, on the other hand, allows you to be more proactive. These might include response time, error rates, request rates, etc. Recognizing the need to be more proactive here, the cloud native landscape, in particular, is seeing a lot of growth in observability offerings, spanning monitoring, logging, tracing, and even chaos engineering.
Application, Container Platform, and DevOps teams take the lead in being held accountable for application performance.
When we break this question out by function, Application, Container Platform, and DevOps teams take the lead in being held accountable in some form. With just 13%, 21%, and 23%, respectively, having no measured accountability, see below.
Interestingly, over 70% of infrastructure/Operations teams are held accountable for application performance. Historically these teams are managing shared environments, which makes it very difficult to connect the performance of individual applications to how resources are being managed. That an overwhelming majority are nonetheless held accountable suggests that the industry is acknowledging the direct impact that infrastructure has on application performance.
Application code and resource contention are the leading causes of performance degradation.
In fact, when we asked about the most common causes of application performance degradation, bugs in application code (24%) and resource contention (21%) were the top two causes. Fortunately, these challenges can be addressed by Application Performance Monitoring (APM) tools and Application Resource Management (ARM), respectively.
For example, take a look at Carhartt. They found that combining APM and ARM has powerful implications for the customer experience, which was proven out on their busiest day of the year. Watch Carhartt Black Friday Success.