Compliance Assessment
A pre-audit examination of named users, engine measurements, and indirect-access pathways. We surface the exposure before SAP does, and we quantify the remediation cost.
Read the brief →A European industrial conglomerate rebuilt the data-volume calculation for its SAP Business Warehouse engine, demonstrated double-counted volumes across replicated landscapes, and settled an $11.4M claim at $3.6M with a forward measurement protocol baked into the contract.
Every result on this site is anonymised at the client's request. Specific figures are real and verifiable through a confidentiality-protected reference call arranged on request.
The conglomerate's Business Warehouse landscape spanned three regional instances, all licensed under SAP's tiered data-volume engine metric. The original 2018 contract specified 40 TB of licensed BW data volume with a tiered overage rate. The self-declaration submitted ahead of the May 2025 renewal showed 47 TB of cumulative storage across the three instances. SAP's commercial team converted that into an overage claim of $11.4M based on the contracted tier price and a four-year unlicensed retrospective.
The claim arrived eight weeks before the renewal date, with explicit language linking acceptance to the renewal terms. The procurement leadership was prepared to settle in the seven-million-dollar range based on what they assumed was a defensible volume position. We were engaged to test that assumption before any settlement number was written down.
The trigger for re-examination was a single observation. The 47 TB total included full replicas of two production datasets that flowed through BW for reporting only, then were replicated into a downstream analytics platform. Under the BW engine licence definitions in the executed contract, replicated reporting data sourced from a measured productive instance should not be double-counted. The contract language was unambiguous on this point. The internal team had not raised the question.
SAP's opening position rested on the raw self-declared volume figure of 47 TB across all three instances, with no deduplication and no source-replication exclusion applied. The supporting workpaper SAP provided was a single-page summary derived from the most recent USMM-style measurement run on the BW estate, with the headline number presented as a deterministic obligation.
We obtained the full BW data dictionary, the source-system manifest, and the replication topology documentation. The deconstruction of the 47 TB figure took eleven engagement-weeks of analysis. The breakdown that emerged was: 19 TB of primary productive data sourced from ECC, 14 TB of replicated reporting data sourced from the primary, 8 TB of historical archive volumes that had been migrated from a decommissioned BW instance, and 6 TB of indexing and aggregation overhead written by BW itself.
Three of those four categories were either excluded by the contract definitions or eligible for reduction under documented SAP measurement practice. The remaining productive primary volume was 19 TB — well below the contracted 40 TB tier ceiling and triggering no overage at all under a clean reading of the contract.
The defence strategy ran on four parallel workstreams, each producing an evidentiary submission to the SAP licence compliance team. The first workstream documented the replication topology with system manifests and timestamped data-flow diagrams, demonstrating that the 14 TB of reporting replicas were sourced from the productive BW instance already counted in the primary volume. SAP's own measurement note from 2021 was cited as the procedural reference for excluding such replicas.
The second workstream addressed the 8 TB migrated from the decommissioned BW instance. The decommissioning had been formally documented in 2023 with a SAP-acknowledged retirement note. Under the contract's decommissioned-system clause, the archive volumes migrated from that retired instance were carry-over data, not new productive volume, and were excluded from the engine metric on the same basis as the original retirement.
The third workstream tackled the 6 TB of BW-internal indexing and aggregation overhead. We obtained the BW basis logs that distinguish productive data writes from system-generated overhead writes. The contractual definition of "BW data volume" referenced productive data, not the sum of all bytes on disk. The internal overhead was demonstrably system-generated and was excluded from the measurement.
The fourth workstream was the contractual rebuttal. We assembled a written counter-position, twenty-three pages with supporting exhibits, and submitted it through the existing SAP commercial contact under privilege. The counter-position offered to settle at the productive volume figure of 19 TB — below the contracted tier ceiling — with a goodwill payment of $3.6M to close the historical period and a renewed tier of 50 TB to provide forward headroom.
Settlement closed at the proposed $3.6M figure, payable against the existing contract floor and treated as a goodwill close of the historical period rather than an overage payment. The forward tier was renewed at 50 TB with a clarifying amendment to the engine metric definition that explicitly excluded replicated reporting data, decommissioned-system carry-over, and BW-internal indexing overhead from the volume count.
The amendment language was important. It was not a discount on the original claim; it was a redefinition of the measurement methodology that applied to all future declarations. The forward measurement protocol attached to the contract specified the deduplication procedure, the source-system attribution rules, and the documentation standard for any future overage discussion. The next self-declaration cycle would be measured against a defended baseline rather than a contested one.
Total elapsed time from engagement to executed settlement was twenty-two weeks. The reduction against opening was 68 percent. The renewal closed on the original schedule.
Further detail on the methodology behind the defence is set out in the Engine Metrics Decoded white paper. For the broader topic context, see the cluster pillar on Engine metrics, the cluster pillar and the deep dive on The BW engine data-volume trap. The topic landing at Engine Metrics provides the broader category context.
A pre-audit examination of named users, engine measurements, and indirect-access pathways. We surface the exposure before SAP does, and we quantify the remediation cost.
Read the brief →When SAP escalates a measurement claim, the right counter-position is contractual, not technical alone. We negotiate the settlement, the contract language, and the forward measurement protocol.
Read the brief →The cluster pillar covering measurement methodology, contractual definitions, and the standard defence patterns for the engine in question.
A multinational CPG group rebuilt its FI and SD engine measurements and removed $7.2M from a self-declaration claim.
Engine measurement claim contested on contract definitions and decommissioned-system rules.
It is the opening position of a negotiation. Speak with a specialist before responding. The first conversation is at no cost and under privilege.
Contact Us →Every Wednesday. Field reports from active matters, decoded SAP communications, and what to look for in the next audit cycle. Work email only.