Choose a problem-solving approach

Choose a problem-solving approach

First protect the operation when a current abnormality needs immediate attention. Then decide what kind of learning is required. Restoring a known condition differs from investigating a recurring gap, reaching a better future condition or exploring a new service. The four examples help choose proportionate effort, ownership and cadence. A large project template can slow a simple response, while a quick fix can leave a systemic cause untouched. Revisit the choice as evidence develops. Methods such as A3, DMAIC or experiments should support the problem, not become rituals imposed on every situation regardless of need.

Follow the method

1. Restore
2. Close standard gap
3. Reach new condition
4. Explore new value

Route four work problems by the learning they require

Fictional service-desk case: Harbor Support processes customer change requests. Every Monday the routing service stops sending acknowledgments. A documented restart restores routing. The duty lead has marked four weekly incidents complete after restarting, but the same failure has returned. A separate improvement request asks for faster handling of routine changes.

Follow the method

1. Active 90-minute routing failure
2. Fourth Monday recurrence
3. Reduce 1 hour to 30 minutes
4. Unformed new service

A restart closes the outage, not the recurring problem

Use a known standard, observed gap and learning need to choose an initial problem-solving route, while separating immediate restoration from recurrence investigation.

Fictional service-desk case: Harbor Support processes customer change requests. Every Monday the routing service stops sending acknowledgments. A documented restart restores routing. The duty lead has marked four weekly incidents complete after restarting, but the same failure has returned. A separate improvement request asks for faster handling of routine changes.

Role: Duty lead deciding the immediate response and process owner assigning subsequent learning.

Normal condition

The current service standard is acknowledgment within one working hour. The restoration procedure returns the service to that condition.

The gap

Four Monday incidents breach the standard; the reason for recurrence has not been established.

• Restore the active service through its authorized procedure before conducting a prolonged study.
• Classification selects the learning need; it does not prove a cause or prescribe a particular form for every issue.

1. Identify the immediate gap

   The duty lead compares the 90-minute interruption with the one-hour standard and initiates authorized restoration, recording affected requests and verification.

   Why: An urgent known-condition gap needs response now. Opening a lengthy analysis template must not postpone the immediate service response.

   Evidence: Restoration record shows what resumed and which requests still need acknowledgment.

2. Keep the recurrence question open

   The process owner opens a separate learning record for four repeated Monday failures. The problem statement names frequency, condition and impact without asserting a software cause.

   Why: Repeated restoration demonstrates that the immediate action did not prevent recurrence. A gap-from-standard investigation is needed even when the latest queue clears.

   Evidence: Incident history and an explicit untested-cause status remain attached to the problem.

3. Distinguish improvement beyond the standard

   The 30-minute request is classified as a target-condition challenge where the existing one-hour standard is currently met. The owner asks what current steps consume time and proposes bounded experiments.

   Why: Calling all improvement a defect hunt can make a new performance ambition look like employee noncompliance.

   Evidence: Current standard, proposed target and the gap to that target are separately recorded.

4. Recognize an open-ended question

   The new service category is routed to customer-value exploration with stated unknowns and a small learning activity. It is not assigned an invented current operating standard.

   Why: Where the service and value proposition are not yet known, a routine restoration checklist cannot supply the missing knowledge.

   Evidence: The record names the customer question and what the first conversation or prototype should establish.

5. Allow the routes to connect

   The lead links today’s restoration to the recurring-gap owner, and the owner can later use experiments to test a cause or a target. The classification is reviewed as evidence changes.

   Why: These are different learning purposes, not sealed departments or competing branded forms. One incident can require both immediate response and deeper work.

   Evidence: The handoff names the next owner, evidence and review point instead of merely assigning a category.

The standard is missing

A fifth complaint concerns a request type whose promised response time is disputed. Sales says one day; operations has no approved definition or time basis.

Clarify the customer commitment and operating definition with the responsible owner before labeling the team below standard. Address the live customer need while gathering facts.

A gap cannot be measured against an invented or ambiguous reference. Clarification may reveal a performance gap, a target discussion or a service-design problem.

Retain the promise, request type, time basis and who can resolve the standard; revise classification only after this is clear.

Run a practice session

Materials

• Four worked-case cards
• Practice cards with the final ambiguity concealed until discussion
• Blank routing/handoff record

1. Frame restoration versus learning · 4 minutes

   What closes an outage? What prevents its return?

2. Demonstrate four routes · 7 minutes

   Which case lacks a current service standard?

3. Classify changed cases in pairs · 9 minutes

   What fact supports your category?

4. Reveal ambiguity and debrief · 5 minutes

   What evidence would make you change your answer?

Debrief

• Coach a category-only answer by asking for the current reference and next evidence.
• Accept different initial routes if the learner states defensible assumptions and customer response.

Write your assumptions before naming a route, then compare your reasoning rather than only the category label.

Transfer into the work

Owner: Frontline lead and the owner of the affected process

Record: Brief problem-selection and handoff record linked to the live response

Review: At the next daily review and when new evidence changes the problem definition

Evidence: Clear current condition, standard/target, observed gap, tested versus unknown cause and assigned next learning step

Escalate an unclear standard or missing ownership; retain containment/restoration records without declaring recurrence solved.