Service Recovery
Part of Customer experience prioritisation
Comparing customer harm with implementation effort
Compare the customer consequence of a service problem with the work involved in each possible fix, without hiding serious failures in a score.
Compare harm and effort as two separate judgements. Describe what the customer loses when the problem occurs, then estimate the work each plausible response requires. Do not let a small, easy change automatically outrank a difficult fix to a serious failure.
Define harm in the customer’s task
State what the person was trying to achieve and the consequence of failure. “Could not see a confirmation” is an observation; “the request failed” needs a check of the underlying result. Distinguish delay, extra work, an incorrect result and an inability to proceed.
Ask whether the customer can recover through another route, what that route demands and whether the failure may recur. Record whose experience is represented. A handful of sessions may show a serious problem for those participants without establishing its prevalence among all customers.
Describe the consequence in words before assigning a rating. For example, needing to call to check a request and being unable to obtain the result at all are different hypothetical consequences. Neither is a finding about a real service.
Estimate effort for each response
Implementation effort belongs to an option, not to the problem itself. A clearer message, a manual interim route and a change to the underlying processing system may address different parts of the same problem.
For each option, ask what must change, which teams and systems are involved, what evidence is needed before release, what ongoing work it creates and who can approve it. Include the effort of checking the customer outcome. Give a range or state the uncertainty where a dependency is not understood.
| Customer consequence | Possible response | Decision question |
|---|---|---|
| Consequential and well understood | Immediate remedy or targeted change | Can the team act safely now and verify the result? |
| Consequential but cause uncertain | Help affected customers and investigate | What evidence distinguishes the possible causes? |
| Limited and well understood | Small change or planned work | Does the expected benefit justify the effort now? |
| Unclear consequence and uncertain fix | Bounded check | What finding would change the decision? |
These are prompts, not automatic priority rules. A potentially serious case can need immediate attention while its wider frequency is unknown. Common minor friction can also justify work across many customer tasks.
Key Metrics for Decision-Making in Service Recovery
- Customer Impact SeverityBased on real user sessions; distinguish between observed and hypothetical harm
- Implementation Effort RangeIncludes cross-team coordination, testing, compliance checks (e.g., ATO data handling), and release approvals
- Interim Solution Required?Yes if long-term fix is uncertain or high effort; must be documented and reviewed
Choose and review a response
If a major system change is the likely long-term fix, consider a safe interim way to help customers while its cause and feasibility are checked. State what the interim action resolves. A workaround does not show that the underlying failure has disappeared.
Record the preferred action, alternatives considered, evidence limits, owner and review point. Before delivery, define the customer outcome to check and any adverse effect that would prompt reconsideration. Revisit both harm and effort when new evidence changes the apparent scope of the problem or the work required.
