An energy group had paid support on a named user count frozen years earlier. A reconciliation against the contractual user definition exposed a large standing overpayment.
The client, a national energy group, ran Oracle E Business Suite across finance, procurement, human resources, and supply chain modules. Its support bill had grown year over year and was anchored to a named user count established several years earlier and never revisited. Support renewed annually at that count plus the standard uplift, and no one inside the group had reconciled the licensed users against the people actually using the system.
Oracle applications are licensed on user definitions that are precise and frequently misunderstood. A Named User Plus, an Application User, and an Employee metric are not interchangeable, and the same person can be counted differently depending on the module and the licensing basis. Over years of organisational change, the group's user records had drifted. Leavers remained provisioned. Service accounts were counted as users. People who held access to a module they no longer used were still licensed for it. Duplicate identities across modules were counted multiple times.
The standing count had become a ceiling that the group paid against regardless of reality, and the annual uplift compounded the overpayment every year. Because applications support is a recurring cost rather than a one time audit event, the overpayment was quiet and persistent, the kind of exposure that never triggers an alarm because nothing is going wrong, the group is simply paying for entitlement it does not use.
The group also faced an upcoming renewal, which created both the risk of locking in the inflated count for another term and the opportunity to reset it if the reconciliation could be completed and evidenced in time.
The engagement was a reconciliation rather than an audit defence, but the discipline was the same. The Measure phase reconstructed the true user population module by module against the precise contractual user definition that applied to each. We pulled the application user records, the active session history, and the human resources system of record, then reconciled identities across modules to eliminate the double counting that inflates application user counts.
We classified every licensed user into one of four states. Active users genuinely using a module under the correct metric, who would remain licensed. Inactive provisioned users, including leavers and dormant accounts, who could be deprovisioned. Misclassified users counted under a more expensive metric than their actual usage required. And non human accounts, service and integration identities incorrectly counted as named users.
The reconciliation found roughly 3,400 users that were either inactive, duplicated across modules, or misclassified. Each reclassification was documented against the contractual definition so that the reduced count was defensible rather than merely asserted. The work distinguished sharply between users the group could legitimately remove from the count and users it had to retain, because an overaggressive reduction creates its own audit risk at the next review.
In the Negotiate phase we took the evidenced reconciliation to the renewal. Because the reduced count was supported line by line against the user definition, Oracle accepted the reset rather than disputing it. The negotiation recovered value in two forms. First, a credit against the standing overpayment that had accrued. Second, a permanently lower support base going forward, which compounds in the client's favour every year the same way the uplift had previously compounded against it.
The Convert phase built the governance to keep the count clean. The renewed agreement set a defined reconciliation cadence, tied user provisioning to the human resources joiners and leavers process, and documented the user definition applying to each module so the next renewal opens from an accurate position rather than a frozen historical number.
An applications overpayment is the quietest exposure there is. Nothing is going wrong. The group is simply paying every year for entitlement it does not use.
The reconciliation recovered $8.2M in overpaid support and cut the annual support base by 31 percent going forward. Roughly 3,400 users were reclassified or removed from the licensed count. The table shows the composition of the reclassification.
| Exposure category | Oracle claim | Settled |
|---|---|---|
| Inactive and leaver accounts | 1,620 | removed |
| Duplicate cross module identities | 910 | consolidated |
| Misclassified metric users | 640 | reclassified |
| Non human service accounts | 230 | removed |
The renewed support agreement reset the named user base to the reconciled count and applied the uplift to that lower figure, so the compounding now works in the group's favour. A defined reconciliation cadence requires the count to be reviewed before each renewal, and provisioning is tied to the human resources joiners and leavers process so dormant accounts no longer accumulate.
The contract also documented the specific user definition applying to each module, ending the ambiguity that allowed the count to drift in the first place. The group now renews against an accurate, evidenced, and maintained user position rather than a number frozen years earlier.
For the buyer side perspective on this product line, our Oracle Applications licensing team and our Oracle Database licensing practice work the same playbook on every engagement. Compare outcomes across the full case study library.
If your Oracle applications support has renewed on the same user count for years, a reconciliation against the contractual definition often recovers standing overpayment and resets the base permanently.
Request a Consultation