Last month, I outlined the need to move towards a more application-centric approach to IT monitoring, but what I didn’t specifically discuss is how the enduser fits into the equation. You might be wondering, “Why would you bring up the enduser here? After all, we’re focused on the data center; let the desktop guys worry about the enduser.” That, however, is the exact line of thinking we need to get away from. The whole point of application-centric monitoring is breaking down the siloes that currently exist within IT so we can improve the quality of endusers’ experience, which largely rests on application performance and availability.
With this in mind, we recently conducted an Enduser Application Performance Surveyto shed light on the application-centricity of the new IT landscape, enduser expectations in it and how IT is currently living up to the challenge of making the shift. We surveyed more than 400 business endusers in the U.S., and the key findings are quite telling. Let’s take a closer look.
Applications Are Critical
The first collection of results probably aren’t that particularly surprising to you if you use applications of any type. Almost all (93%) of endusers noted that application performance and availability affects their ability to perform their job, and almost two-thirds (62%) classified it as absolutely critical. Two-thirds (67%) of them also said that application performance and availability has become more important over the past five years. Almost certainly this is a reflection of our increased reliance on mobility.
But this group also identified other factors that have contributed to an increased criticality of applications in their jobs:
- The amount of time using applications on the job has increased.
- Their workload has increased.
- They are expected to work faster.
Expectations of Users
Consistent with the importance of apps to job performance, the majority of users also expressed an expectation of rapid response to the resolution of application issues. No longer is there a middle ground where the application is less than optimal in performance, but tolerated. Gone are the days when users just naturally accepted the need to reboot their PC a couple times a day (or more). Today, in essence, a non-performing application is a broken application.
Not only are there high expectations for application performance, but also for IT’s response to application issues. Over three-quarters (76%) of users said they were not willing to wait more than an hour for an application performance issue to resolve itself before calling the IT department. However, on the other side of that situation, two-thirds of them (67%) expected the problem to be resolved within an hour of reporting them, with over a third (35%) expecting resolution in 30 minutes or less.
The reluctance of users to wait to report issues is a good thing. In many cases, early recognition of a developing problem can be solved quicker, and easier, and affect less total users overall, than if users ignore a problem (and IT is unaware of it) for an hour or more. However, IT ought not be unaware of such issues for an hour, either.
But this new world is a key challenge for IT, particularly in consideration of the growing complexity of the entire application stack. Applications run on a myriad of devices, with multiple types of connectivity. The connections originate from both inside and outside the corporate network, and the back-end services and data being used may also reside inside or outside the corporate network.
Realities of IT
Unfortunately, for many organizations, the realities of IT are a long way from the expectations of application users. Over half of users contacted their IT department in the past year due to an app performance or availability issue, and a third of them reported having done so six or more times in that year. Over a third (36%) of those who reported issues also said they’d waited a full business day or more for the issue to be resolved, and almost a quarter (22%) waited several business days.
But perhaps the most significant revelation reported is that half of the respondents who encountered issues (and that was 70% of the total population), said that half of the time they contacted the IT department about an app performance issue, the IT department was unaware there was a problem. Application performance would seem to be a chronic issue in need of immediate help; and IT organizations should not be getting “performance alerts” from endusers.
This last point sits squarely on the lack of proactive monitoring of the full application stack. No longer is it sufficient to monitor silos within the infrastructure. Just because the network is up, and the Internet connection is not saturated, is no guarantee data will flow unimpeded, nor does the fact that servers are online and CPU/RAM/Disk are performing to spec.
The combination of dozens of individual monitoring points now must all come together to ensure that the full end-to-end spectrum of involvement in delivering an application to an enduser is fully functional. Disruption in any one of those dozens of points can cause an undesirable experience with an application. More significantly, the IT department should be aware of such disruptions before enduser call to complain about their impacts.
Have you implemented monitoring in your environment? Is it siloed monitoring or can you trace an issue at one end of the application stack to a root cause at the other end?