Lesson 6 of 6
Reporting for sign-off
7 min
The regression cycle ends in a decision: can this change go to production? Process owners take that decision, and they need a short, honest report that shows coverage, results and residual risk. The test lead assembles it; key users supply the evidence and their judgment.
What the report shows
| Area | In scope | Executed | Passed | Open defects | Status |
|---|---|---|---|---|---|
| Core pack | 24 | 24 | 23 | 1 medium with workaround | Met with exception |
| Order to cash (targeted) | 18 | 18 | 18 | 0 | Met |
| Procure to pay (targeted) | 11 | 10 | 10 | 0 (1 blocked: supplier test data) | Met with exception |
| Finance close and reports | 9 | 9 | 9 | 0 | Met |
| Role checks | 7 roles | 7 | 6 | 1 high: plant controller cannot see stock value | Not met |
Contoso feature pack regression summary. A single table often says more than ten slides.
Talking about residual risk
Residual risk is what remains after testing: scenarios not tested, defects open, areas only partially covered. Express it in business terms. "One procurement scenario blocked by test data" is a test fact. "Subcontracting receipts in plant 1720 were not regression tested; if they fail, receipts would need manual posting for up to a day" is a risk a process owner can weigh. Pair each risk with how it will be watched in production, such as a key user checking the first receipts after go-live.
Making the decision
- Are the exit criteria met?Yes → recommend go · No → next question
- Is each unmet item covered by an accepted workaround or a fix before production?Yes → recommend go with conditions, each with an owner and date · No → next question
- Do you control the production date?Yes → recommend no-go until fixed · No (Public Edition) → document the risk, workaround and monitoring; escalate to the process owner and SAP as needed
In the Contoso example, the plant controller role defect is high: the controllers cannot see stock values, which they need at month-end. The role fix is small and can be retested in a day. Priya recommends go with conditions: fix and retest the role before the production move, with Dan as approver. The medium core pack defect has a documented workaround in the work instruction.
After production
- Run a short smoke check in production right after the change, using a few read-only checks and agreed test postings.
- Watch the risk areas named in the report during the first days and the first month-end.
- Feed any production issue back into the regression pack.
- Record what the cycle cost and what it found, so the next plan is better sized.
Telling users what changed
Regression proves the old still works; users also need to know what is new or different. Turn the relevant release items into a short note per role: what changes on their screens, what they should do differently, and which work instructions were updated. Maria sends sales reps a half-page note and a two-minute demo at the team meeting before the production move. Changes that users discover on their own become hypercare tickets; changes they were told about become non-events.
What would you do?
Contoso's process owners are asked to sign off a release upgrade. The report says "312 of 340 tests passed (92%)". Dan asks what he needs before signing. What is the best response from Priya?
- A sign-off report shows scope and coverage, results, defects, residual risk and a recommendation.
- Report the core pack separately; percentages alone hide critical failures.
- Express residual risk in business terms and say how it will be monitored.
- Recommend go, go with conditions or no-go against exit criteria; name owners for conditions.
- After production: smoke checks, watch risk areas, feed issues back into the pack.
Check your understanding · not graded
Lesson check
1. Who takes the decision to move an upgrade to production?
2. Why report core pack results separately?
3. Which is a well-expressed residual risk?
4. What should happen right after the change reaches production?