ClientDefense Tech
My roleHelped build the measurement and controls behind a decline in estimated improper payments.
The situation
When I began working on the Navy’s improper-payments program, there was a more fundamental problem than reducing improper payments.
The problem
We couldn’t reliably measure them. Before you can control a problem, you need to know where it happens, how often it happens, and why.
What I found
The program needed its own machinery: a way to estimate the problem, find the points of leverage, and see whether controls were actually working.
The decision
Working with finance and payment experts, we built much of that machinery from scratch: statistical estimation models, risk assessments, sampling and review methods, pre-payment controls, training, local review activity, and reporting for leadership.
The result
From 2015 to 2019, internal program reporting showed roughly $250 million less in estimated improper payments. In a separate two-year comparison, underlying error rates fell by about 45%.
The more interesting result was that an enormous, fuzzy problem became something people could see, investigate, and control.
These are results across the program. The $250 million is a decline in estimated improper payments, not recovered cash.
An enormous problem becomes much more tractable once you can measure it, find the causes, and put controls where they matter.
Facing something similar?
We start with a free 30-minute conversation about the business and what’s on the table.