Lesson 5 of 5
Planning plan-to-produce regression after conversion
10 min
Listen to this lesson
Regression after a conversion answers one question: does the business still run the same way, except where we chose or had to change it? For plan-to-produce, the riskiest answers are in MRP, because the planning engine and its data changed, and in costing, because production order values flow into inventory and the books. Those two areas lead the regression plan.
Priorities
| Priority | Area | Why | Example test |
|---|---|---|---|
| 1 | MRP results | New engine, MRP areas, sourcing and production versions | Same frozen October data: compare proposals in EQ1 and CMQ material by material |
| 1 | Production order costing | Values reach inventory and the P&L | Planned and actual costs, variances and settlement for a reference order |
| 2 | Order execution | Production versions select BOM and routing | Convert, release, confirm, receive for CMC-ALT-100 in 1710 and 1720 |
| 2 | Changed processes | Creation indicators, MRP lists, MRP areas | Requisition hand-off and release; spare stock in 171S not planned |
| 3 | Custom code and reports | Adapted or replaced objects | Morning exception work in apps finds the same shortages |
| 3 | Interfaces and output | Order prints, shop floor feeds | Order print and confirmation feed for ASSY-10 |
Contoso's plan-to-produce regression priorities after conversion.
The strongest MRP regression test is a comparison on identical data. A mock conversion copies production, so the same requirements, stock and master data exist in the converted system. Freeze the planning inputs, run MRP in the source copy and MRP Live in the converted system, then compare proposals by material, quantity and date. Every difference is explained as expected (documented behavior or a decision Contoso made) or logged as a defect.
- Mock 1Baseline and first comparisonMRP and costing comparisons on converted production data; differences classified
- Between mocksFix and retestMRP areas, production versions, BAdIs and reports corrected in CMD
- Mock 2Full regressionAll priority 1 and 2 scenarios; differences must all be explained
- Dress rehearsalTimed checksKey MRP and costing comparisons inside the cutover window
- Cutover weekendGo/no-go evidenceReference order and MRP sample checked before Monday 06:00
Agree entry criteria for the regression run with the planner and Dan: the mock conversion is complete and its technical checks are signed by Sam, MRP areas and production versions are in place, adapted custom code is transported, planning inputs are frozen, and test users have their converted roles. Starting regression on a half-finished mock wastes the cycle, because differences caused by missing setup look exactly like defects.
Before-and-after validation tools
SAP's Conversion Guide describes Data Transition Validation (DTV), a tool that compares business data before and after a system conversion or upgrade, using business reports or transactions as instruments of comparison, and states that it is available as part of the standard license. The guide's procedure: define the reports with the lines of business and auditors in the preparation phase, set up a project and specifications, extract data from the source during ramp-down, and validate in the target afterward. For manufacturing, candidate reports are stock values by plant, open production orders with their balances, and WIP.
| Check | Before (EP1 copy) | After (converted) | Expected |
|---|---|---|---|
| Stock value CMC-ALT-100, plant 1710 | From stock report | Same report or app | Equal |
| Open production orders and balances | Order list with costs | Same selection | Equal |
| MRP proposals for October | Classic MRP on frozen data | MRP Live on the same data | Equal or explained |
| Production versions | Count of materials without one | Count after migration reports | Zero for manufactured materials |
Sign-off criteria for the converted process
- All priority 1 and 2 scenarios passed in mock 2 with evidence.
- MRP comparison differences all classified; none unexplained.
- Stock values and open order balances equal before and after conversion.
- Every manufactured material has a valid production version; MRP areas in place for former storage location MRP.
- Adapted custom code and reports tested; replaced reports confirmed by planners.
- No open Critical or High defect without an accepted, documented workaround.
What would you do?
Mock 2 shows MRP proposals differing for 37 of 4,200 materials. Thirty are explained by documented MRP Live behavior; seven are not yet analyzed. Go-live is in six weeks. What should Priya report?
- MRP and costing lead plan-to-produce regression after conversion.
- Compare MRP on identical frozen data; explain or log every difference.
- Use before-and-after comparisons, including DTV where it fits.
- Sign-off needs reconciled values, complete production versions and tested changed processes.
Check your understanding · not graded
Lesson check
1. Why include MRP in regression after conversion?
2. What makes an MRP comparison between ECC and SAP S/4HANA meaningful?
3. What does Data Transition Validation compare?
4. Seven MRP differences are unexplained before go-live. How are they reported?