USPS OIG vet: Analytics can make audits help, not hurt

A wealth of data and analytics tools can help auditors better analyze agency activities -- and encourage officials to act on their recommendations.

Shutterstock image (by Laborant): clear magnifying glass, blurred binary code.

(Laborant / Shutterstock)

Big data and analytics can make an auditor's job easier and more effective. But the auditor must be prepared to face resistance from the organizations he or she scrutinizes -- and from within.

"Human beings are wired for survival," said Bryan Jones, former director of the U.S. Postal Service's Data Mining Group in the Office of Inspector General, at the Predictive Analytics World Government conference in Washington. "They're not wired for truth."

Jones, who is now director and program manager at Elder Research, reminded would-be innovators of the resistance they could face from people eager to hold onto outdated jobs.

USPS knows the technology/employment tension all too well, and Jones, who worked on USPS' Risk Assessment Data Repository, said he had encountered resistance from many job-protecting managers during his investigations.

But the impartiality of big-data analytics can make a numbers-based "truth" more readily apparent, he said, as long as agencies promote cultures that accept the numbers for the value they provide.

A single team, backed by a supportive executive, can start the pro-analytics culture shift at an agency, Jones added.

"The ultimate measure of success is when an investigator comes to us and says, 'Hey, when are you going to refresh the data in the model?'" he said. "Then I know we're successful because when you hear an investigator say something like that, you know that you've got the organizational shift, you know that you've got the culture going in the right direction, and you know the front line of the organization is buying into this."

Auditors at USPS, on the other hand, decided that an analytics-heavy, flexible audit model was worthy only after six months of disassembling and reassembling the model, Jones said.

Analytics improve audits because they can be "based on a real problem and not what I think might be going on."

Big data should also help audit reports carry more weight because "it doesn't matter how right you are" as an auditor if organizations don't implement your recommendations, Jones said.