Making Metrics Pay

The government has been making a big push on performance management in IT for a few years now, but there’s been a missing piece, says David Paschane of the Veterans Affairs Department.

And what is the missing piece? According to the senior analyst in the IT Operations and Field Division for VA’s Office of Planning and Programs, it’s a lack of applying a scientific approach to enterprise architecture and business process re-engineering.

Technology could really play a role here, say Paschane, who spoke at the recent GITEC Summit in Orlando, Fla.

“IT has the opportunity to help with analytic reviews to track performance,” he says.

To take advantage of performance data, “you want repeatable knowledge,” says Paschane. “That’s what you want to find, pull out and take advantage of.” That way you can make sustainable and culturally acceptable changes, he says.

Survey and research conducted by the Government Accountability Office back up Paschane’s viewpoint. In the past decade, GAO has surveyed federal IT employees about performance management at least four times, says Randolph Hite, director for IT architecture and systems.

“There’s a ton of information that’s being collected but very little of it’s being used,” Hite says. Why? Two reasons: value and timeliness. He says that there’s a lack of understanding on how to use performance metrics and how to use the data to advance the mission.

“I have had program managers tell me, “It’s not my job to deliver performance capability; it’s my job to deliver a capability by a given date.’ ” says Hite.

Paschane suggests use of an approach similar to the Performance Architectural Science System to get a robust analytic capability, which has four steps: control your initiatives, stabilize conditions, optimize across functions and contextualize outcomes.

The current approach to performance management in many agencies is to look at projects one at a time — what Paschane calls the “one egg versus a dozen view.” This approach can create problems with transparency because an agency or team can manipulate outcomes to drive up the performance scores for a single system or project. That’s much harder to do if the approach used requires that a system be viewed in context to other systems, projects and programs within an organization.

To create a PASS-like practice within an agency, Paschane recommends taking four steps:

• Develop and intelligent infrastructure vision that’s unique to the agency.
• Set an incremental but cohesive (agencywide) plan.
• Create solid teams with the right experience to monitor results and act on the data.
• Include enough people to avoid burnout — this is not something that can be handled by just a few people in large organizations, Paschane says.

Hite says many agencies are good at coming up with strategies and developing metrics. The rub is making use of the information gathered to get better results within agencies and for developing strategies and priorities going forward, he says. Ultimately, it comes down to holding people accountable to achieve better results and to share lessons learned, Hite says.