We believe the old way of solving application performance issues isn’t working and simply visualizing complex health and performance data in a new way is not enough. Design is more than just making it look pretty.
In order to “lead by design” we started an intense research project where we observed real people in their workplaces to identify, durable problems. This practice of conducting first-hand user research is known academically as “cultural anthropology” or “ethnographic research”. We talked to dozens of customers, not only ours but also those of our competitors. The goal was to encourage users to tell stories and observe the user working on real issues in their own environment.
We observed not only how a single user might use a tool to find and resolve a performance issue but also observed the sticky notes of tribal knowledge and expertise each user knows about the environment. We also observed how teams interact with each other and with the tool. From those interviews, we learned volumes of information that just can’t be gleaned by filling in a survey or answering questions on a phone call. In all we interviewed over 80 users across 17 customers capturing 500 pages of interview notes.
Now comes the fun part. This research was analyzed and synthesized into nine different distinct personas; each was given a name. We then brought all this information back to our designers and engineers, where we have ongoing discussions about our various users and what they’re really trying to accomplish. It’s no longer about focusing just on APM-centric symptoms and solutions but a focus on the macro-level problems our users face each day. It really is amazing to hear the hallway conversations about “Pete”—who is one of our focus personas—by name. But that’s not all.
We also built an APM NextGen demo as a “concept car” to show the world what the correct APM solution for one of our nine profiled users might look like. Once again going back to the users to validate the outcome of our research. The results of this work has truly transformed the way we think about application performance management and how we understand the problems to be solved. It’s also changed the way in which we interact with and have constructive dialog with our customers.
This is happening right now and nets a very different outcome from the past. You see this in our newest release of CA APM 10. In fact, this is the strategy we’re employing to bring about a fundamental change in the way those personas add value to their shops. One example is our patent-pending APM Team Center Perspectives.
Perspectives provide role-based views that allow users to “pivot” and show the environment from their perspective. This is much like one would pivot data in a spreadsheet to only show the relationships that makes sense. Using custom attributes such as location, supporting team, or any other organizational construct, users can quickly reorient and filter complex views into simple perspectives that represent the language and vocabulary of YOUR persona and/or business.
For example: Pete, a level-1 IT Operator, may wish to look at the map oriented to location. Ryan, a middleware-tier admin, might wish to see that same data oriented by architectural component. And Andrew, a developer, might want to see that same data broken down by detail. Notice that this is the same map and the same data but filtered to reduce the noise and focus on what’s really important to the user and task. You can see how this helps in accelerating triage and diagnostics of application performance issues.
Leading by design allows us to not just develop a new UI but one that that addresses the issues faced by APM users. After all, in the Application Economy, companies who excel and differentiate will win by moving more quickly to design, build, deploy and fix software better than anyone else, by knowing where the real issues are, and then deploying their experts on those most-important tasks. APM—an APM very different from the APM of today—should be the catalyst to create such behavior.