I have been following blogs by organizations that perform expense report auditing. It’s a habit in my line of business, being the “software guy” who helps Fortune 500’s analyze data to determine fraud, policy misuse and errors. Routinely I see these expense auditors uniformly promote auditing 100% of expense reports as the way to begin reviewing claims before “tapering off” to approximately 50% of claims once compliance and spending are under control. They contend that when employees know claims are being reviewed, behavior will change and the requirement to review all expense claims is eliminated.
I agree. 100% review of expense claims is the only way to begin auditing claims, and it is possible. The question many doubters ask is “is it sustainable?” Let me break this down for you: automated transaction monitoring of expenses is the only way to manage an expense reporting process on an ongoing basis. Use of automated monitoring and analysis systems allows organizations to continue 100% review for a fraction of the cost of the expense management solution or manual audit currently in place. It is sustainable and very necessary.
Why bother with ongoing monitoring? Here are a number of reasons:
- Automated systems reduce the requirement for human review to less than 5% of total expense transactions.
- Less human review means fewer false positives (effectively every expense line that is reviewed during a manual audit that is compliant is, essentially, a "false positive" since time is spent reviewing it)
- More time to address the root causes of non-compliant transactions. Further, focusing on expense lines (i.e., transaction detail) allows for a greater focus on behavior that occurs across employees, across the expense claims of a single employee, and over time.
Human review remains an important component of the process. We're not trying to get rid of audit services, in fact, we think our technology is a nice complement to third party audit services who want to analyze transactions better and faster. When leveraging automated analysis, the level of effort required for identifying non-compliant claims goes down, and the time available to address underlying behaviors and make adjustments to policy goes up when leveraging automated systems to perform the initial analysis and review. After all, we only flag exceptions, it is up to our human users to leverage the data we find to investigate behaviors, shape policy, and figure out how to decrease the bottom line using the analysis we provide.