01 · Definition and boundary
Definition and boundary
A reaction plan states the actions to take when a defined process abnormality occurs. It connects a trigger to containment, communication, investigation, recovery authorization, and evidence of effectiveness. Its purpose is to make response reliable under pressure. It is different from a general instruction to “inform the supervisor” and different from a long-term root-cause project. The immediate response protects the customer and preserves evidence; subsequent investigation determines what must change.
Use it for important process controls, pilot guardrails, and foreseeable failure modes. Prerequisites include a specific trigger, someone with authority to act, an approved fallback where appropriate, and clear criteria for resuming normal work. A trigger may be a critical event, a requirement violation, or a valid statistical signal. These trigger types should be named separately because they do not imply the same response or evidential conclusion.
02 · Design triggers and roles
Design triggers and roles
Fictional Northbank Dispatch uses carton-to-lane scan confirmation. The team defines a critical trigger as a confirmed wrong-carrier load or loss of traceability. That event causes immediate containment regardless of the average performance rate. A separate operating trigger is a confirmation queue exceeding the locally approved waiting limit for a collection wave. The waiting limit is an operational guardrail, not a universal safety threshold or a statistical control limit.
The dispatcher may stop loading the affected lane. The shift lead coordinates containment and customer-facing operational decisions. The systems owner investigates the record and configuration. The dispatch process owner authorizes restoration after evidence is reviewed. A transport coordinator contacts the carrier when a loaded vehicle may be affected. Roles include deputies so the plan remains usable at night. The plan specifies which channels are available during a systems outage.
03 · Contain first, then investigate
Contain first, then investigate
Containment isolates affected cartons and prevents further loading from the suspect lane. The team identifies the last verified good assignment and uses dispatch records to establish the possibly affected interval. It preserves labels, timestamps, configuration information, and relevant operator observations. Reprinting every label immediately may destroy useful evidence and can create duplicate identifiers, so replacement follows the controlled procedure.
Investigation checks whether the failure came from assignment timing, stale labels, scanning behavior, system response, or data completeness. The first plausible explanation is treated as a hypothesis. The plan distinguishes correcting the specific carton from preventing recurrence in the process. It also states when the issue must escalate—for example, when the affected population cannot be bounded or traceability cannot be reconciled. Continuing dispatch under an approved fallback requires a separate authorization and defined conditions; it is not an informal workaround created during the incident.
04 · Worked incident and decisions
Worked incident and decisions
During a collection wave, Northbank detects one carton whose printed lane differs from the current carrier record. The dispatcher stops loading lane B and isolates its thirty-two staged cartons. Review shows that six were assigned before a carrier update, while twenty-six were assigned afterward. The team checks all thirty-two because the boundary is initially uncertain. Five of the six earlier cartons still have valid assignments; one needs a controlled relabel. No cartons have left the warehouse.
The systems owner finds that the change signal failed after a configuration update. The immediate correction restores the signal in the tested configuration. The team also checks other lanes touched by the update rather than assuming lane B is the whole affected population. This broader check is driven by the shared mechanism. The incident is recorded as one detected assignment mismatch and a contained exposure of thirty-two cartons, not thirty-two confirmed errors.
05 · Authorize recovery with evidence
Authorize recovery with evidence
Before reopening lane B, the process owner requires reconciliation of all contained carton identifiers, a successful reassignment test, and confirmation that the approved configuration is active. The shift lead verifies the lane is physically clear of unresolved cartons. The owner signs the restoration record with the time and scope. A planned observation of the next collection wave checks that the correction works under normal workload.
These criteria are specific to this incident and design. They are not a universal rule that one test or one successful wave always proves recovery. If the cause remains uncertain, the team may retain enhanced checks or continue under fallback while investigation proceeds. The plan records who accepted that residual risk and when it will be reviewed. Recovery evidence includes both technical function and operational use; a software test alone does not show that staff can recognize and handle a reassigned carton.
06 · Weak versus strong response and advanced adaptation
Weak versus strong response and advanced adaptation
Weak practice says “stop and notify management” without defining affected work, evidence, or restoration authority. Another weakness is blaming the operator and restarting after a reminder while a configuration failure remains. Strong practice contains exposure, preserves evidence, distinguishes symptom correction from cause investigation, and makes recovery a documented decision. It avoids an unconditional restart deadline that pressures people to ignore unresolved traceability.
Advanced teams create different response levels for a local abnormality, a cross-lane system issue, and a customer exposure. They rehearse the transitions and test the plan when the normal communication channel is unavailable. Incident learning updates standard work, controls, and risk analysis. The plan remains concise at the point of use, with deeper investigation guidance linked separately. A usable reaction plan reduces improvisation while leaving room for responsible judgment when an event falls outside the scenarios anticipated by the team.
07 · Operating conditions and evidence
Operating conditions and evidence
The response record should distinguish discovery time, event time, and containment time. Northbank may discover a stale assignment at loading even though the relevant software change occurred earlier. Those times affect the scope of the investigation. The team records which carton population was checked and what evidence established the last known good boundary, rather than declaring that “all stock was checked” without a defined denominator.
Communication follows the exposure. If no carton left the warehouse, an internal operational response may be sufficient. If a vehicle departed, the transport coordinator follows the approved customer and carrier notification route with verified identifiers. Messages should state known facts and the actions underway, not an unconfirmed explanation. After recovery, the owner reviews whether containment was timely and whether the plan itself was usable. Delayed access to the manifest or uncertainty about restart authority becomes an improvement action for the response process.
08 · Decision record and next step
Decision record and next step
The plan distinguishes restoration from closure. Northbank can restore operation after the assignment signal is verified and cartons are reconciled, while a broader investigation into configuration-change controls remains open. Calling the incident closed at restart would lose that prevention work. The owner assigns a due date for the wider correction and reviews whether a similar failure could affect another lane.
A reaction plan also protects decision quality under pressure. During a carrier cutoff, staff may feel compelled to load first and reconcile later. The approved plan states when that sequence is prohibited and which fallback is permitted. Deputies and escalation routes reduce delay without transferring authority to an untrained person. The post-incident review assesses the response against actual evidence: exposure bounded, customer consequence checked, recovery demonstrated, and follow-up owned. This is a practical test of the plan’s usefulness, not a search for someone to blame.
| Response stage | Northbank action | Evidence required |
|---|---|---|
| Containment | Stop affected lane and isolate suspect cartons | Defined affected population and retained identifiers |
| Investigation | Check shared configuration and assignment timing | Mechanism supported by records and reproducible tests |
| Recovery | Owner authorizes normal operation | Reconciliation, verified configuration, and operating checks |
Connections
Related methods and phase use
Placed under Improve because of the outcome it supports. Use the method in other phases when its purpose and assumptions fit the work; the phase list is a guide to selection, not a compulsory sequence.
Sources
Professional foundations and further reading
The explanation and fictional teaching examples are written for LeanTrainingLab. Linked primary sources provide professional foundations; example values and local conventions are stated in the guide. The accepted toolset draws on Six Sigma+Lean Toolset, second edition (Lunau, editor; Springer, 2013). No book text or examples are reproduced here.