Assurance practice · 12 Mar 2026
Sampling for usage exceptions without losing the plot
When the revenue assurance audit app returns thousands of usage exceptions, clearing every row is theatre. Sampling with intent is the craft.
Most desks start with “we will look at the top of the queue.” That rewards whatever sort order the vendor chose, not what threatens billed revenue. Begin instead by tagging each exception family with an estimated revenue band — even a rough band beats none.
Three bands, one rule
High-band items get full investigation until root cause is known. Mid-band items get a fixed sample size agreed with finance before the close begins. Low-band items get a smoke check for systemic patterns, then a residual-risk note.
Write the sample sizes down. If you renegotiate them every Friday, you are not sampling — you are bargaining under fatigue.
What the app cannot decide
A revenue assurance audit app can cluster similar signatures. It cannot tell you whether your organisation accepts residual risk on mid-band items during a migration week. That is a management call. Your job is to make the call visible.
A workable starter
- List exception families by estimated monthly revenue impact
- Agree sample counts with finance before the cycle starts
- Log what you skipped and why
- Revisit bands after each major catalogue change
We teach this sequence inside the Sampling under pressure module of Revenue Leak Detection.