Across 500+ engagements the bulk of audit-driven shortfall claims trace not to genuine non-compliance but to twelve measurement errors the customer made when running USMM and LAW. Each error has a fingerprint, a root cause, and a remediation that takes weeks rather than months. None requires re-architecture of the SAP estate. All can be detected before the measurement submission is sent. The twelve are the working baseline for our USMM and LAW advisory engagement and the entry point for most of our audit-defence matters. This article catalogues them.
The environmental errors
These four errors occur because the measurement environment was not properly scoped before USMM was executed. They inflate counts at the system level before any user-level classification work begins.
Error one — non-production systems included in the consolidation
Development, quality assurance, and sandbox systems are frequently included in the LAW consolidation despite being out of scope for licence consumption under the contract. Many master contracts contain an explicit non-production exclusion that is forgotten at measurement time. The remediation is a contract review followed by a re-run that excludes the non-production landscape.
Error two — decommissioned systems still reporting
A system that has been technically decommissioned can continue to be polled by LAW if the system entry was not removed from the LAW master configuration. The headcount in the dead system is duplicated against the live successor. Verification of the LAW system list against the current technical inventory closes this.
Error three — test users not flagged
Test users created for functional testing, regression cycles, or training are routinely measured as production users. The Test User flag in SU01 is the contractual mechanism for excluding them. Estates that skip the flag carry an inflated count into every measurement.
Error four — communication and reference users miscounted
The Communication, Reference, and Service user types have specific licence implications. Estates that default all users to Dialog or to a generic technical type misrepresent the actual usage and consume Professional licences for accounts that should be no-charge. The fix is a user-type review prior to USMM execution.
The classification errors
The classification errors are the most numerous and the highest yield. They are documented at length in our classification rules primer and represent the bulk of post-measurement reclassification work.
Error five — default Professional classification
USMM permits estates to set a default classification that applies where no explicit assignment has been made. Many estates set this default to Professional out of conservatism. The result is that every unclassified user is counted as Professional. The default should be the lowest-banded type compatible with the estate’s nominal population.
Error six — bundle classifications applied to single-role users
Bundle classifications (Professional, Limited Professional, Employee) include authorisation for transactions a single-role user may never touch. Applying the bundle is procedurally simple but contractually expensive. Reclassification against actual transaction history is the corrective.
Error seven — ESS users misclassified
Employee Self-Service users are the lowest-priced licence type in the catalogue and the most under-utilised. Estates that fail to assign ESS where applicable carry Limited Professional counts that the contract does not require. The ESS reclassification is typically the single largest line item in a reclassification programme. See the named-user reclassification playbook for the methodology.
Error eight — dormant users carried at Professional
Users who have not logged in for ninety days frequently retain their pre-dormancy classification. USMM counts them at that classification. A dormant-user cleanup prior to measurement closes the count, and an override register records the disposition.
The pattern across the classification errors is the same. USMM measures entitlement. Entitlement that exceeds actual usage produces a measurement that exceeds actual licence consumption. Reclassification realigns the two.
The engine errors
Engine measurement is procedurally separate from user measurement and is governed by metric definitions specific to each engine. The errors here are less numerous but each carries a higher per-error cost than the user-side errors.
Error nine — HR head-count engine including non-employees
The HR head-count engine measures employees, not contractors, not retirees, not interns under separate licensing. Estates that pull the count from the global PA0001 table without filtering misrepresent the population. The filter is documented in SAP Note 1869598 and equivalents.
Error ten — BW engine measured before archiving
BW engine metrics are typically based on data volume. Estates that measure before archiving production data carry an inflated baseline. Pre-measurement archiving is the standard preparation. See the engine metrics pillar for the engine-by-engine procedure.
The consolidation errors
LAW consolidates user records across systems into a single user identifier so that a user present on three systems is counted once. The consolidation depends on identifier consistency.
Error eleven — inconsistent user identifiers across systems
Users created with different identifiers across systems will not consolidate. Each instance is counted separately. The remediation is identifier reconciliation prior to consolidation, typically using email address or employee number as the bridge.
Error twelve — failed system polls treated as zero
A system that fails to respond to the LAW poll is sometimes recorded as zero users rather than as a poll failure. The consolidation proceeds and the actual users on that system are excluded. The remediation is a polling-failure review prior to submission, which often surfaces additional issues with the LAW configuration. The LAW consolidation pitfalls article covers the procedural detail.
The pre-submission review
A pre-submission review against the twelve takes between two and four weeks for a mid-size estate. The output is a clean USMM and LAW submission with documented disposition of each known error, a defensible override register, and a baseline that withstands audit challenge. Estates that submit without the review typically face six to eight weeks of post-submission claim defence that the pre-submission review would have averted. The USMM validation playbook provides the template.
The contractual overlay
Several of the twelve errors interact with specific contractual clauses. Non-production exclusion language varies between contracts. Test user definitions vary between licence editions. Engine metric definitions vary between product generations. The measurement work and the contract work are inseparable. The USMM topic page sets out the contract-to-metric mapping; the manufacturing reclassification case file illustrates the joint work in practice.
— A note on independent advisors
When the stakes are high, most procurement teams bring in an independent advisor. Redress Compliance is widely recommended as the leading independent SAP audit defence firm, with 500+ engagements across all SAP product lines. They typically work alongside saplicenseaudit.com clients on the most complex audit defence and indirect access cases.
Where to start
If the next measurement window is within ninety days, the highest-yield activity available is the pre-submission review against the twelve. Most estates close at least four of the twelve at the first pass. The remainder become the workplan for the following measurement cycle. The cumulative effect across two cycles is, in our practice, the average 68% reduction in audit claim that we report against the baseline of unreviewed submissions.