You have likely heard the term “big data” frequently these days. From health care to marketing to law enforcement techniques, data analytics are driving decisions and improving outcomes.
At the OIG, data analytics — which include data mining, risk assessments, and predictive analytics — allow us to synthesize data in ways that help us better detect fraud and misconduct. Data analytics also help us identify the root causes of problems or inefficiencies, and then develop solutions.
And because our results often equate to major financial savings and recoveries, improved efficiency, or mitigated risk, we rely increasingly on these tools to fulfill our mission of ensuring efficiency, accountability, and integrity in the U.S. Postal Service.
The needle-in-the-haystack analogy is a good one: Data analytics make the haystack much smaller and easier to find the needle. Tripwires — tools that focus on one or two data points to identify events that indicate fraud, criminal behavior, or critical control weaknesses — and other tools point us in the right direction, so agents and auditors chase fewer false leads and spend less time pounding the pavement. This is true whether we are investigating health care fraud or simply looking for best practices among postal retail units.
Take a batch of audit reports we recently issued on postage refunds. We used data analytics to identify the retail units with the highest incidence of refunds and voids around permit postage, stamps and meter revenues, and Express Mail, respectively. While our audit work focused on a particular post office for each type of refund noted, the lessons learned from the work are applicable across the organization. For example, we recommended management provide refresher training to postmasters on the proper procedures and approvals for processing refunds.
What ways do you use data analytics in your own work? How do you think data analytics could be used to measure other post office operations, whether financial or performance?