A 26% spend cut, no audit notice required.

The brief

The retailer is a European multi-format group operating supermarkets, convenience stores, and a non-food vertical across fifteen markets. SAP ECC 6.0 had been the operational backbone for twelve years, with seven engines layered on top covering warehouse management, transportation management, retail merchandising, and a custom-built engine for store-level inventory. Total in-scope named users were twenty-eight thousand two hundred at engagement opening, with a contracted run-rate of approximately eleven point four million dollars across all licence lines.

The CIO commissioned the engagement before any audit notice. The brief was twofold: identify recoverable licence value inside the existing entitlement, and prepare a defendable measurement position for the next audit cycle, expected within twelve to eighteen months based on the group's prior pattern.

The exposure was significant in absolute terms but more significant in trajectory terms. Without optimization, the next audit was likely to anchor on the existing provisioned user count and engine measurements, with a probable claim exceeding three million dollars based on the group's internal projection.

The opening baseline

The provisioned named-user count was twenty-eight thousand two hundred. Active authentication in the prior twelve months covered nineteen thousand one hundred. Approximately nine thousand one hundred accounts were either dormant, provisioned but never activated, or in a status that contractually excluded them from the named-user count.

The user-classification mix had been set at the original go-live and had not been re-audited in five years. The Professional band was significantly over-populated relative to actual transaction-history evidence. Approximately three thousand four hundred users classified as Professional showed Limited Professional or Employee Self-Service patterns over the measurement window.

Three engines — the transportation management engine, the warehouse management engine, and the custom retail merchandising engine — were measured against metrics that no longer reflected the operational footprint. The transportation management engine had been measured against a contract definition that included intra-warehouse moves; the actual contracted definition was shipments.

The four tactics

We worked four tactics in parallel against the baseline. Each one was contained inside the existing contract and required no commercial concession from SAP — only evidence, documentation, and a defendable measurement protocol.

Dormant account purge

We ran an authentication-log analysis across a rolling thirteen-month window. Eight thousand four hundred accounts had no authentication in any month. We documented each disposition with a last-authentication timestamp and either deactivated, reclassified, or scoped each account out of the next measurement.

Named-user reclassification

We rebuilt the user classification against transaction-history evidence over a twelve-month window. The Professional band was reduced by three thousand four hundred users, with the surplus reclassified into Limited Professional and Employee Self-Service. The reclassification was supported by tcode-level evidence and a configuration history showing role design that produced over-classification.

Engine measurement reconciliation

We obtained the raw measurement outputs for the three disputed engines and rebuilt the readings against the contracted measurement definitions. The transportation management engine was over-measured by a factor of three; the warehouse management engine by a factor of one point eight; and the retail merchandising engine by a factor of one point four. The corrected readings were within the contracted entitlement.

Forward measurement protocol

We documented a forward measurement protocol covering the user-classification methodology, the dormant-account purge schedule, and the engine measurement definitions. The protocol was reviewed with SAP's regional account team and accepted as the documented basis for the next audit cycle.

The settled position

The annual run-rate dropped from eleven point four million dollars to eight point four million, a twenty-six per cent reduction. The retailer reclaimed approximately three million dollars of annual spend with no licence purchase, no contract amendment, and no commercial concession requested from SAP. The forward measurement protocol is now documented and reviewed quarterly.

The named-user mix is now defendable against any audit cycle within the protocol horizon. The engine measurements are documented at the line level and supported by configuration evidence. The dormant-account discipline is built into the joiner-mover-leaver process and runs continuously rather than annually.

Total elapsed time from engagement opening to documented protocol was twenty weeks. The annual saving became effective in the next measurement cycle, and the forward audit risk dropped from the projected three-million-plus claim to a defendable position inside the contracted entitlement.

The lessons

First, optimization without an audit on the calendar produces the largest defendable saving. The retailer reclaimed three million dollars annually because the work was done before SAP anchored a position.

Second, named-user reclassification is the single largest line in most optimization engagements. The classification mix drifts over time, and the drift is always toward over-classification when the original role design is broad.

Third, engine measurements rest on contract definitions that are often forgotten by the operational teams running the systems. The corrected reading sits inside the contract, not outside it. Fourth, the dormant-account purge is a perpetual exercise, not an annual one. Fifth, a documented forward measurement protocol reduces the next audit cycle from a discovery exercise to a reconciliation exercise.