The Challenge: Four Companies, Three ERPs, No Common Metrics
The client was the operating partner team at a mid-market private equity firm preparing a platform company for exit. The platform had grown through four acquisitions over three years, and each portfolio company was running a different ERP: one on NetSuite, one on SAP S/4 HANA, and two on Dynamics 365 Finance.
Every operating review involved a manual reconciliation exercise. Each company defined Days Sales Outstanding (DSO) differently: one used trailing 60-day revenue in the denominator, another used trailing 30-day, a third used annual revenue divided by 365. DPO definitions had similar divergence. The result was that even when numbers were directionally comparable, they were never auditably so.
The specific problems the operating partner was trying to solve:
- No consistent KPI definitions across the four companies. Each CFO could defend their own number, but none could be compared to the others.
- Manual rollup process that took 3–4 business days before each board meeting, with frequent last-minute corrections.
- No drill-down capability: board-level numbers could not be traced to underlying AR aging buckets or AP payment terms without additional offline analysis.
- Lender covenant reporting was being done separately by each company controller, with no platform-level view to validate consistency.
The Approach: Normalize Exports, Build One Semantic Layer
Rather than consolidating the ERPs (a multi-year, multi-million dollar program), DataQubi built a governed normalization layer on Microsoft Fabric that sat above all four ERP environments.
Results
The operating partner team used the dashboard in the next two board meetings without modification. The consistent metrics also materially simplified lender covenant reporting; each monthly submission was generated from the same certified model, with a documented lineage trail the audit team could independently verify.
Why This Worked Without Consolidating ERPs
The ERP systems were left completely intact. No migrations, no integrations, no system changes at the portfolio company level. The architecture that made this possible:
- Read-only ERP exports as the data source: each company's controller ran a standard month-end export from their ERP, which landed in Fabric automatically via a scheduled Dataflow.
- Governance-first design: the KPI contract was defined and approved before any technical work began. This meant the data model was built to serve agreed definitions, not the other way around.
- Fabric Lakehouse as the normalization surface: bronze/silver/gold layering made it possible to apply company-specific transformation logic in the silver layer while exposing a common gold model to the dashboard. Changes to a single company's extraction logic could not break another company's metrics.
Structural insight: Most PE working capital rollup failures happen because teams try to consolidate metrics before consolidating definitions. Locking the KPI contract first (even a simplified one) is the highest-leverage step in any multi-entity finance data engagement.
Key Takeaway for PE Operating Partners
If your operating review meetings regularly include 15 minutes of metric reconciliation before anyone can discuss performance, the problem is almost never the ERPs. It is the absence of a shared definition layer that sits above them.
A governance-first, ERP-agnostic approach means you can have portfolio-level comparability in days, not after a multi-year consolidation program. The board and lender value the consistency of the number, not which ERP produced it.
For firms preparing for exit, this kind of normalized, auditable working capital view is also directly relevant to quality of earnings processes, giving both the firm and potential buyers a defensible financial data foundation.