Lesson 6 of 6
Recovery to a defensible go-live
12 min
The end point of recovery is not "go-live at any cost". It is a defensible go-live decision: one that the sponsor, the process owners, the auditors and the people who run the business on Monday can all understand and support, because it was taken on evidence against criteria agreed in advance. This lesson brings the course together into that decision.
What makes a decision defensible
The evidence pack
Prepare an evidence pack for go/no-go that a newcomer could read and reach the same conclusion. It should take minutes to find any piece of evidence, because every item links to its source.
| Area | Evidence | Owner |
|---|---|---|
| Testing | SIT and UAT exit status; open defects by priority with workarounds; regression results for late fixes | Priya |
| Business acceptance | Signed acceptance per process from each process owner | Dan, order-to-cash owner and others |
| Data | Mock and dress rehearsal reconciliation sign-offs per object | Data lead with business owners |
| Cutover | Dress rehearsal timings, float, fallback test result, runbook frozen | Luis |
| Readiness | Training completion, authorizations tested, support desk and hypercare roster ready | Change and support leads |
| Risk | Residual risk register with acceptance by named owners | Grace |
The go/no-go evidence pack and who owns each part.
Residual risk, accepted by the right person
No go-live is risk free. The question is whether each remaining risk is known, bounded, has a workaround, and has been accepted by someone with the authority to accept it. A finance risk is accepted by the finance process owner or above, not by the test lead. Record each acceptance with its conditions. If a risk cannot find an owner willing to accept it, it is a no-go item.
- Are all go-live criteria met with evidence?Yes → go · No → next question
- Is each gap bounded, with a workaround and an owned, dated fix?Yes → next question · No → no-go for that area; consider scope deferral or new date
- Has each residual risk been accepted by someone with authority for that area?Yes → go with conditions, recorded in the decision log · No → no-go until accepted or resolved
- Is the fallback tested and the hypercare plan credible?Yes → confirm decision · No → fix before proceeding
The recovery timeline
- Week 0Warning signs acted onGap between status and evidence made visible
- Weeks 1–3Independent assessmentConfidence rating and findings
- Week 4Re-baseline approvedScope, plan, gates and contingency reset by the sponsor
- Weeks 5–14Controlled deliveryFlow restored; SIT and UAT to exit criteria; mocks
- Week 15Dress rehearsalFinal plan proven with contingency
- Week 17Go/no-go on evidenceDefensible decision, conditions recorded
After go-live: learn and keep the habits
Recovery habits are worth keeping after go-live: measured status, daily triage in hypercare, gates with evidence, and stable metric definitions. Hold a lessons-learned review within a few weeks, focused on causes rather than blame, and feed its findings into the next release and the organization's standard approach. The best outcome of a rescue is a program office that would spot the next drift months earlier.
What would you do?
At the recovered go/no-go, everything is met except one item: the intercompany billing output between company codes 1710 and 1810 has a known defect. A manual workaround exists and has been tested, the fix is scheduled for week two of hypercare, but the intercompany accountant has not yet been asked to accept the workaround. What should Grace do?
- A defensible go-live rests on pre-agreed criteria, linked evidence, explicit residual risk, named deciders, fallback and hypercare.
- Prepare an evidence pack a newcomer could follow to the same conclusion.
- Every residual risk needs acceptance by someone with authority for that area.
- Keep recovery habits after go-live and learn from causes, not blame.
Check your understanding · not graded
Lesson check
1. What makes a go-live decision defensible?
2. Who should accept a residual risk in finance?
3. A residual risk has no owner willing to accept it. What does that mean?
4. What should a lessons-learned review after a rescue focus on?