ERP Data Migration: A SQL + Excel Playbook
Lessons from a SAGE CRE 300 → CMiC ERP migration at Keeley Companies — how to extract, validate, and reconcile large datasets with SQL, VS Code, and advanced Excel when you don't own the ETL pipeline.
The problem
An ERP migration that's already in flight is a different animal than a greenfield one. You're reconciling against someone else's ETL decisions, finding discrepancies after the fact, and the "source of truth" keeps moving. This is the playbook I wish I'd had on day one.
What you'll build
- A repeatable extract-validate-compare workflow in SQL + Excel
- Row-count and hash-based reconciliation between old and new systems
- A discrepancy report your ETL team can actually act on
Prerequisites
- Read access to both source and target databases
- SQL comfort — joins, window functions, basic aggregates
- Advanced Excel: XLOOKUP, INDEX/MATCH, IFERROR, SUMIFS, TRIM
1. Scoping what to reconcile
Coming soon — the tables that matter vs. the ones that don't; how to read a migration spec skeptically.
2. Extracting comparable snapshots
Coming soon — SQL query patterns I used to pull matching slices from both systems.
3. The Excel validation layer
Coming soon — why Excel is still the fastest audit tool, and the formulas that do the heavy lifting.
4. Writing discrepancy reports that get fixed
Coming soon — the format the ETL team at Keeley actually responded to.
Wrap-up
Coming soon — what I'd automate next time.