Lesson 5 of 5
Testing and transformation risk
8 min
Ask experienced SAP program leaders what goes wrong in transformations, and the answers sound similar: invoices that do not print, interfaces that drop orders, reports that show different numbers, users who cannot do their job on day one, balances that do not tie out. Almost every one of these is something a test could have found. Testing is not a phase at the end; it is the program's main instrument for turning unknown risk into known, managed risk.
Why transformation risk is high
- Hidden change: screens look familiar, but data structures, objects and components have changed underneath, so old assumptions fail quietly.
- Breadth: a transformation touches every process, interface, report and role at once, not one area at a time.
- Data: years of data are migrated or converted, and data problems surface only when processes use them.
- Custom code: custom programs that read old tables or rely on old behavior may run without error and give wrong results.
- One window: cutover happens over a short period, often a weekend, with limited room for a second attempt.
The test types that manage it
| Test type | What it proves | Typical timing |
|---|---|---|
| Unit and string testing | Single configuration items, developments and short process chains work | Realize, during build |
| Integration testing | End-to-end processes work across modules and interfaces | Realize, in cycles |
| Data migration or conversion validation | Data arrived complete and correct and is usable | Each mock or rehearsal |
| Regression testing | What worked before still works after the change | After conversion runs and for later upgrades |
| Performance testing | Critical processes and batch jobs meet time targets at volume | Late Realize or Deploy |
| User acceptance testing | The business can do its work and accepts the solution | Deploy |
| Cutover rehearsal | The cutover can be done in the window with the evidence needed | Deploy, before go-live |
Focus testing on risk
No program can test everything equally. Testing is planned around risk: how likely is a failure, and how bad would it be? Order-to-cash for fleet customers, the month-end close and payment runs at Contoso are high impact, and they are touched by many changes, so they get the deepest testing. A rarely used report with a manual fallback gets less. The matrix below is how Priya and the process owners agree where effort goes.
Evidence and sign-off
Testing reduces risk only if its results can be trusted and acted on. That means test cases linked to requirements and risks, recorded results with evidence, defects tracked to closure with severity, and exit criteria agreed before each cycle. At the end, the business signs off on that evidence: Maria for order-to-cash, Dan for finance. SAP's own tools, such as SAP Cloud ALM for test management, can hold the plan, the results and the defects in one place; what matters most is that the evidence is complete and owned.
- Identify risksProcess, data, code, integration and cutover risks with owners
- Plan testsDepth and type by risk; entry and exit criteria per cycle
- ExecuteResults and evidence recorded; defects raised and tracked
- MeasureCoverage, pass rates, open defects by severity, trends
- DecideBusiness owners sign off on the evidence; sponsor decides go
What would you do?
Six weeks before Contoso's go-live, the program is behind schedule. A senior manager proposes cutting the second integration test cycle and the final cutover rehearsal to save time. What should Priya recommend?
- Transformation risk is high: hidden change, breadth, data, custom code and a single cutover window.
- Test types: unit and string, integration, data validation, regression, performance, acceptance, cutover rehearsal.
- Focus testing by risk: business impact and likelihood of failure.
- Go-live decisions rest on evidence and on sign-off by business owners.
Check your understanding · not graded
Lesson check
1. Why do familiar screens increase transformation risk?
2. Which test type proves the cutover can be done in the planned window?
3. In risk-based testing, which area gets the deepest testing?
4. Who signs off order-to-cash readiness at Contoso?