A scenario matrix helps choose tests; it does not establish that every integration is possible or working. For the selected handoff, verify system capabilities, access rights, permissible data transfers and error recovery.
GRI and GRI AI in plain language →
From problem to decision
1. Choose a customer mission
Choose a concrete situation and intended customer outcome. Gather observations and interviews, including abandoned journeys. Mark the start, channel transitions, fulfilment and return. Use de-identified descriptions.
Mission → context → intended outcome → channel path → friction → observation source.
Step deliverable: A map of one customer mission.
2. Connect the journey to system work
For each step, describe the customer action, staff action, system, context passed and accountable role. Define a shared request or order identifier. Do not enter contact details or real identifiers in this worksheet. Transfer only necessary data on an agreed basis.
Step → customer action → employee → system → context passed → owner → receipt confirmation.
Step deliverable: A map of staff actions, systems and context handoffs.
3. Run an end-to-end test
Follow a test path through first contact, handoff, continuation, payment when relevant, fulfilment and return. Check context recovery, duplicate prevention and consistent status. Include a repeated event, an error and a retry. This is a test protocol, not an integration with your systems.
Scenario → expected behaviour → actual behaviour → evidence → gap → accountable role.
Step deliverable: An end-to-end test record with actual results.
4. Fix one gap and verify the effect
Choose one gap and an observable customer outcome: fewer repeated explanations, faster continuation or more completed tasks. Record the baseline, change, duration and limitations. Repeat the test and record a decision.
Gap → change → customer outcome → business metric → before / after → limitations → next action.
Step deliverable: A tested change and a next-step decision.
How to judge whether it helped
You should leave with a document that supports a decision and a next action. Completing steps does not guarantee a commercial result. Measure the chosen change separately over an agreed period. If evidence is missing, state that and plan how to collect it.
- The test customer continues with context preserved and consistent status.
- Repeated events, errors, retries and duplicate prevention have been tested.
How this differs from S01–S13
This is a user route for a specific problem. The GRI AI research stages S01–S13 and the 13-week programme remain in the library as separate methodological work. This route is not a validated GRI diagnostic scale.