AI Prompts for Service Level Analysis

Service level is a term that means at least four different things — probability of no stockout, share of demand filled, share of orders on time, share of lines shipped complete — and organizations argue about performance because different functions are measuring different ones. Before analyzing service, agree on the definition; after that, the useful question is the cost of each increment and whether every customer and product needs the same one.

These prompts nail down measurement, quantify the service-cost curve, and design a differentiated policy. They complement the safety stock page (which does the inventory arithmetic for a given service target) and the order fulfillment page (which analyzes the failures).

Before you use these

Have these ready to replace the highlighted [variables]:

• Current service metrics as defined and measured by each function
• Demand and supply data to compute fill rates
• Customer and product segmentation with revenue and margin
• Inventory and expedite cost data

The prompts

• 1. Design service level measurement
• 2. Analyze the service–cost trade-off
• 3. Set a differentiated service policy

1. Design service level measurement

Act as a supply chain performance analyst designing service level measurement for [business].

Current metrics: [name, who reports it, how it is calculated as far as known, the figure it currently shows]
Data available: [orders, order lines, shipments, promised dates, requested dates, stock positions, backorders]
Customer commitments: [contractual or promised service levels by segment]

1. Define precisely and give the formula for: cycle service level (stockout probability per replenishment cycle), unit fill rate, line fill rate, order fill rate, on-time delivery (against requested vs against promised date), on-time-in-full, backorder age. State what each is useful for and its biases (e.g. on-time against promised date hides poor promising; unit fill rate hides small-line failures).
2. Reconcile the current metrics: which definition each corresponds to, why they disagree, and which are measuring the same thing differently.
3. Recommend a headline set (3–4 metrics) for the business and the operational set behind it, with data source and measurement frequency.
4. Specify the measurement rules: tolerance windows, treatment of customer-requested changes, partial shipments, cancelled lines, substitutions.
5. Propose the reporting view: by segment, product class, location, and trend.

Do not propose a metric the data cannot support. Flag any current metric that should be retired and why.

2. Analyze the service–cost trade-off

You are analyzing the service–cost trade-off for [business] by segment.

Data: [by segment: current fill rate or CSL, inventory value, demand variability, lead time, carrying cost rate, expedite spend, estimated stockout cost (lost margin, backorder handling, penalty), revenue and margin]

1. Build the service–inventory curve per segment: inventory required at [e.g. 90/95/97.5/99/99.5%] service, using the safety stock relationship and the variability given. Show the method.
2. Marginal cost: inventory carrying cost of each step up in service. Highlight where the curve steepens.
3. Stockout cost curve: expected stockout cost at each service level (probability × cost per stockout event or per unit short).
4. Total cost = carrying cost + stockout cost + expedite cost at each level; identify the minimum-cost service level per segment.
5. Compare the minimum-cost level to the current target and commitment; quantify the cost of over- or under-serving each segment.
6. Sensitivity: how the optimum shifts if stockout cost is 50% higher or lower — this is the least certain input.

Present the curves as tables with the optimum marked. State clearly that the stockout cost assumption drives the result and how to improve the estimate.

3. Set a differentiated service policy

Act as a supply chain director setting a differentiated service policy for [business].

Segments: [customers by value/strategic tier; products by margin, ABC class, substitutability, lifecycle]
Service–cost findings: [optimum service level by segment, cost of over-serving]
Commitments: [contractual service levels, market expectations]
Constraints: [systems ability to differentiate, sales acceptance]

1. Build a policy matrix: customer tier × product class → target service level, inventory policy (stocked, stocked with lower target, make/buy-to-order), promised lead time. Justify each cell by value and cost.
2. Allocation rules under shortage: the order of priority, how it is applied (fair share, tier first, contractual first), and the exceptions process.
3. Promising rules: how lead times are quoted by cell so promises reflect the policy rather than optimism.
4. Commercial implications: where the policy changes what a customer currently receives, the conversation needed and the options (pay for higher service, accept longer lead time, minimum order rules).
5. Governance: who can approve exceptions, how often the matrix is reviewed, and the metrics that show it is working (service by cell, inventory by cell, exception volume).
6. Implementation: system changes, sales training, communication sequence.

Present the matrix and rules. Flag any cell that conflicts with a contractual commitment.

Related prompts

• Safety Stock Calculation
• Order Fulfillment Analysis
• Inventory Optimization
• Supplier Scorecards and SLAs
• ABC and ABC-XYZ Inventory Analysis
• Operational KPI Reviews

Logical next step

After this, most operations teams move on to Safety Stock Calculation.

All Operations & Supply Chain prompts · Search the full library