Find AI Prompts
Home›Operations & Supply Chain›Order Fulfillment Analysis
Operations & Supply ChainLogistics & Warehousing

AI Prompts for Order Fulfillment Analysis

Order fulfillment is where the supply chain meets the customer, and the metrics that matter are the customer's: did the order arrive complete, on time, undamaged, with the right paperwork? The perfect order rate combines these, and its power is in the failure analysis — every imperfect order has a cause somewhere upstream, and the causes cluster.

These prompts analyze the failures, break down the lead time so you can see where orders wait, and handle the communication when a delay is unavoidable. The delay notification prompt is included here because how you tell customers about a problem is part of fulfillment performance.

Before you use these

Have these ready to replace the highlighted [variables]:

The prompts

1. Analyze perfect-order failures

Best forFinding which failure types and causes drive the perfect order rate.
Inputs needed
  • Order-level performance data
  • Failure reasons if coded
How to use itGive the model the four components separately (on time, in full, damage-free, documentation) so it can show which drives the rate and whether failures overlap.
Expected outputPerfect order rate with component breakdown, failure Pareto by cause, segment comparison, and the improvement priorities.
Act as a fulfillment analyst reviewing perfect-order performance for [business / period].

Data: [order, customer/segment, promised date, delivered date, ordered qty, shipped qty, damage flag, documentation/invoice error flag, failure reason code if available, channel, ship-from location]
Definitions: [on-time tolerance; in-full tolerance]

1. Compute the perfect order rate and each component rate: on time, in full, damage-free, documentation-correct. Show overlap — orders failing more than one component.
2. Failure Pareto: by reason code where available; otherwise by component and segment. Identify the top causes contributing 80% of failures.
3. Segment analysis: perfect order rate by customer segment, channel and ship-from location. Identify where performance is worst relative to the commitment.
4. Pattern analysis: day-of-week, order size, product family, or lead-time-requested patterns in failures.
5. Cost of failure: expedite, re-delivery, credits, claims — where data allows; otherwise the categories to capture.
6. Priorities: the three causes to fix first with the expected effect on the perfect order rate, and the owner (planning, warehouse, carrier, order management, master data).

Present the component and Pareto tables, then a short narrative. Where reason codes are missing or unreliable, say so and propose a coding scheme.

2. Break down order-to-delivery lead time

Best forSeeing where orders spend their time and which stage to compress.
Inputs needed
  • Timestamps by stage
  • Service commitments
How to use itAsk for the distribution at each stage, not just the average — the tail is where late orders come from.
Expected outputStage-by-stage lead-time table with average, median, 90th percentile, share of total and wait vs work, plus the compression opportunities.
You are analyzing order-to-delivery lead time for [business].

Timestamps per order: [order received, order entered, credit/hold released, allocated, released to warehouse, picked, packed, shipped, delivered]
Commitment: [promised lead time by segment or service]
Context: [cut-off times, shift patterns, carrier pickup schedule, credit hold policy]

1. For each stage: average, median, 90th percentile duration and share of total lead time. Separate value-adding time from waiting (e.g. released-to-picked is mostly queue).
2. Identify the stages with the largest 90th-percentile tail and the orders in that tail — what do they have in common (holds, backorders, large orders, specific locations, day of week)?
3. Cut-off and calendar effects: how much lead time comes from missing a daily cut-off or pickup, and the effect of moving the cut-off.
4. Compare lead time to commitment by segment: share of orders meeting it and the stage where late orders lose the most time.
5. Compression opportunities: for each candidate stage, the mechanism (process, policy, schedule, system), expected reduction, and the trade-off (e.g. later cut-off vs carrier pickup).
6. Recommend the two changes with the largest effect on the on-time share, with assumptions.

Present as a stage table plus findings. If timestamps are missing for a stage, say which and what it hides.

3. Draft delivery delay notifications

Best forA set of customer notifications for different delay situations that are honest, specific and reduce inbound queries.
Inputs needed
  • Delay scenarios
  • Customer segment and channel
  • What you can offer
How to use itGive the model the real scenarios you face and what you can actually promise. Notifications that over-promise generate a second, worse conversation.
Expected outputNotification set by scenario with subject, body, variables, tone guidance, and the follow-up cadence.
Act as a customer operations lead drafting delivery delay notifications for [business].

Scenarios: [e.g. carrier delay with new date known; carrier delay with date unknown; partial shipment/backorder; stock allocation issue before ship; weather or disruption event; address or documentation issue requiring customer action]
Channels: [email, SMS, portal]
Customer segments: [B2B accounts, consumers, distributors — with tone expectations]
What we can offer: [new date, partial ship, expedite at our cost, cancellation, discount — with rules]

For each scenario, draft:
1. Subject line that states the situation plainly.
2. Body: what happened (brief, factual, no blame-shifting), what it means for their order (specific: items, quantities, new date or the date we will confirm by), what we are doing, what they can choose to do, and how to reach us. Use [variables] for order number, items, dates, names.
3. Length appropriate to the channel; an SMS version where relevant.
4. Tone notes per segment.
5. Follow-up rule: when the next update is sent if the situation is unresolved, and the trigger for escalation to a personal call.

Constraints: never promise a date we cannot confirm; do not use 'unforeseen circumstances' as an explanation; put the customer's decision options before our apology. Include a short internal checklist: what must be verified before each notification is sent.

Related prompts

Logical next step

After this, most operations teams move on to Service Level Analysis.

Get the free Operations & Supply Chain AI Starter Kit → Nine of these prompts as a diagnose → analyze → plan workflow with an intake worksheet, delivered by email. See what's inside

All Operations & Supply Chain prompts · Search the full library