AI prompts for logistics companies

Act as an operations analyst helping a logistics company improve the specific workflow below. Build recommendations from the actual operating facts I provide. Do not invent shipment volumes, savings, carrier performance, regulatory requirements, delivery promises, warehouse capacity, or system features. Where evidence is incomplete, state what must be measured or confirmed before changing the process.

Inputs:
[LOGISTICS WORKFLOW AND OBJECTIVE]
[OPERATING DATA AND EXCEPTIONS]
[NETWORK, SYSTEMS, AND PARTNERS]
[OWNERS AND CONSTRAINTS]
[SERVICE TARGETS AND TIMEFRAME]

First, map the flow from order or inbound trigger through final handoff, proof of completion, and exception resolution. Identify each handoff, cut-off time, scan or data event, carrier or warehouse dependency, decision owner, and customer communication point. Separate normal flow from exceptions such as address problems, stockouts, missed pickups, damage, delivery failures, returns, and claims.

Produce:
1. A current-state operating map and a concise statement of the service problem.
2. A diagnosis table ranking likely causes by evidence, operational impact, and confidence.
3. A prioritized improvement plan covering process, data quality, carrier or warehouse management, systems, and customer communications. For each item, state owner, dependency, expected mechanism of improvement, risk, and acceptance test.
4. A 30/60/90-day rollout plan with a contained pilot before broader deployment where appropriate.
5. A daily and weekly control tower scorecard: on-time performance, cycle time, exception rate, backlog aging, scan compliance, cost per shipment or order where relevant, and one quality or safety guardrail.
6. An exception playbook with trigger, owner, response SLA, customer message decision, recovery action, and escalation path.

Do not recommend reducing controls that protect safety, customs, regulated goods, or chain of custody without qualified review. Before answering, check that every proposed action has an accountable owner and measurable operational signal, and flag any recommendation based on missing baseline data. Ask up to 3 clarifying questions only if a required input is missing.

Create a carrier performance scorecard for [CARRIERS] using [SHIPMENT DATA] for [LANES OR REGIONS] over [TIME PERIOD]. Calculate only metrics supported by the data: tender acceptance, pickup compliance, transit performance, on-time delivery, damage or claim rate, tracking event completeness, exception aging, and cost where supplied. Segment results by lane, service level, and shipment type where possible. Flag data-quality limits, define investigation thresholds, and propose an agenda for a carrier review. Do not infer fault from a single metric. Ask up to 3 questions only if required inputs are missing.

Write a warehouse SOP for [WORKFLOW] at [FACILITY] using [CURRENT PROCESS AND SYSTEMS]. Include safety prerequisites, roles, equipment, start-of-shift checks, numbered task steps, required scans, label or inventory checks, quality-control sampling, exceptions, escalation, and end-of-shift reconciliation. Specify what proves each step is complete in [SYSTEM OF RECORD]. Do not create safety rules or equipment instructions absent from approved policy; flag them for the warehouse safety owner. Ask up to 3 questions only if required inputs are missing.

Develop a returns-process improvement plan for [PRODUCT TYPE] across [RETURN NETWORK] using [RETURN DATA AND POLICIES]. Map the path from return request to customer resolution, receipt, inspection, disposition, inventory update, and refund or replacement. Analyze avoidable return reasons, transit delays, inspection backlog, disposition recovery, and customer communication gaps. Deliver prioritized actions with owner, baseline, target, pilot, and guardrail for fraud, quality, or safety. Do not change refund, regulated-goods, or disposal rules without authorized review. Ask up to 3 questions only if required inputs are missing.