Teamwork with Business Users Feeds Analytics ROI at Postal Service

by   |   September 24, 2013 5:09 pm   |   0 Comments

Above, a visualization of fraud risk data used by the Postal Service Office of Inspector General. The map gives a contract fraud investigator a view of where to start searching for potential fraud. It is based on an analytics model that considers 34 risk indicators to generate a score. The tool can look at contractors in areas such as supplies, services, equipment and transportation. Image courtesy of Bryan Jones.

Above, a visualization of fraud risk data used by the Postal Service Office of Inspector General. The map gives an investigator a view of where to start searching for potential fraud. It is based on an analytics model that considers 34 risk indicators to generate a score. The tool can look at contractors in areas such as supplies, services, equipment and transportation. Image courtesy of Bryan Jones.

WASHINGTON – The most important element of success when starting up an advanced analytics operation in a federal agency is not the amount of funding you get, it’s finding the users with the right background who want to attack a juicy problem—and then using existing resources to prove what analytics can do. That was the lesson shared by Bryan Jones, an analytics leader at the U.S. Postal Service Office of Inspector General (OIG).

Related Stories

Texas uses data visualization to combat Medicaid fraud.
Read the story »

Risk managers visualize fresh correlations by analyzing location data.
Read the story »

Government data chiefs find varied measures for analytics value.
Read the story »

IRS implements analytics for compliance, fraud detection and workforce management.
Read the story »

In a lively presentation at the Predictive Analytics World Government conference on September 18, Jones, who is director of the Countermeasures and Performance Evaluation (CAPE) unit at the OIG, talked about his experience starting up an analytics operation for the OIG in an on-stage discussion with Federal News Radio Executive Editor Jason Miller.

Jones and his team have built analytic models, risk models and reporting tools that help inspectors zero in on potential fraud cases in the U.S. Postal Service. When officials from other government agencies looking to start an analytics group see the team’s work, the first question they usually ask is what it cost, he said. But the more important questions are:

• What is the problem that you’re trying to solve?

• What resources do you have on hand that you can use to solve it?

• What IT infrastructure is available to you?

• What staff is available and what are their backgrounds?

• Do you have at least a handful of users willing to think differently?

Jones urged fellow analytics proponents to look beyond lack of funding or even political resistance, and start where they are with what they have. “Let’s not look at these as obstacles, or excuses for not even starting something,” he said. “Evaluation of what you have in these various areas is key.”

Jones’ first project – helping to detect fraudulent workmen’s compensation claims—was a good target because it represented a large amount of potential savings to the organization.  Not only that, but the fraud inspectors were interested in working with Jones’ team. The project went into production two years ago.

“Our investigators were hungry for some sort of a tool or help in guiding them to investigate that area,” he says.  It’s important to have users who are open to new possibilities. “You need to be able to select the three or four individuals that can see things a little bit differently and connect with them,” he said. “They are the ones that are going to pave the way for the rest of the organization.”

Analytics Viewed as a Collaboration with Business Experts
Another key to acceptance is to adopt a humble attitude, noted Jones. Rather than assuming the data analytics group knows how to solve the problem, or even that the business group needs to use any of what the analytics group is offering, let the business experts guide you, he said. “We don’t try to convince people of anything,” he said. “We just put the data in front of them and ask, ‘What do you see here?’” Then they’ll start to ask questions, which helps break down any resistance they may have, and leads to useful applications, he said.

It’s important to bring some of that business expertise onto your team, if possible, he said. To help develop a program to detect workmen’s comp fraud, CAPE brought in a registered nurse from the OIG’s fraud division. She was able to spot trends in the data because she was familiar with workmen’s comp fraud schemes.

Today, the analytics team has been so successful that investigators are asking CAPE for help and the OIG is investing more money in the project, said Jones. Specifically, additional resources were reallocated to CAPE from other areas of the agency over the last four years, leading to an expansion of the original staff of seven people to 23 today.

And instead of managers asking about cost, the conversation now focuses on returns on investment. For example, the team’s models have returned four times the dollars that they cost to build, said Jones. “We had to prove it out, but then resources started to come our way. Now people are coming to us and asking, ‘How many FTEs (full-time equivalents) do you want?’ That’s the kind of conversations we’re having today.”

Tam Harbert is a freelance writer based in Washington, D.C. She can be contacted through her website.







Tags: , ,

Post a Comment

Your email is never published nor shared. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>