Fill Rate

Goal of the analysis:

Quantify how much customer demand is satisfied immediately from available inventory and diagnose the drivers of shortfalls. Fill Rate measures the share of orders, lines, or units that ship complete at the first shipment against the requested/committed date. Executives use Fill Rate to protect revenue and customer satisfaction, reduce cancellations and backorders, minimize premium freight, and guide inventory investments (safety stock, allocation, network positioning), supply reliability, and warehouse execution. A robust approach disaggregates fill by channel, node, SKU class, and customer tier; distinguishes demand capture at the “first ship” from later backorder recovery; and ties gaps to inventory, forecasting, and process issues.

Data required:

• Order and demand data (OMS/ERP/CRM):
  • Order header/line: order ID, customer/tier, channel (e-comm, retail, B2B), SKU, ordered qty, requested ship/delivery date, promise date, ship-complete policy, substitutions/backs, cancellations/returns.
  • Event timestamps: order entry, allocation/reservation, release, pick/pack, ship confirm, partials/backorders.
• Shipment actuals (WMS/TMS):
  • Shipped qty per line and shipment, first shipment date, short-pick/exception codes, pick accuracy, staging/packing events.
  • Node (DC/FC/store) fulfilling each line; split shipments.
• Inventory/ATP and policy:
  • On-hand/available-to-promise (ATP) snapshots at order time and at release; cycle count accuracy; replenishment status.
  • Safety stock targets, reorder points, lead times, order frequency/MOQ, substitution and allocation/rationing rules.
• Supply and planning context:
  • Open POs and expected receipts by node, supplier lead-time performance, forecast (demand plan) and accuracy (bias, MAPE), seasonality/promotions.
  • Network assignment rules (nearest node, load balance) and backorder recovery strategy.
• SLA/compliance and normalization:
  • Customer SLAs for fill (line-level vs ship-complete), allowed substitutions, early/late rules; time zones and holidays; 52/53-week.

Detailed step-by-step instruction on how to conduct the analysis:

1. Define metrics and scope.
   • Unit Fill Rate (UFR) = Units shipped in first shipment ÷ Units ordered.
   • Line Fill Rate (LFR) = % of order lines shipped complete in first shipment.
   • Order Fill Rate (OFR) = % of orders shipped complete in first shipment (no backorders).
   • Define the reference date: requested ship date (preferred) or promise date; measure “first-ship fill” (immediacy) and optionally “period fill” (within X days).
   • Set policy for substitutions (counted as filled if customer-accepted?) and exclusions (cancelled-before-ship).
2. Assemble and reconcile datasets.
   • Join orders to shipments by order/line IDs; determine first shipment per line; compute shipped qty at first ship.
   • Snapshot or reconstruct ATP at order-entry time (using inventory logs) to distinguish promise quality vs inventory drift.
   • Reconcile totals to invoiced shipments and inventory movements; align time zones.
3. Compute fill rates and ancillary KPIs.
   • Calculate UFR, LFR, OFR by day/week and aggregate by channel, customer tier, node, SKU, SKU family, and ABC class.
   • Derive Backorder Rate (lines/units not filled at first ship), Backorder Age, and Lost Sales Rate (cancelled lines ÷ ordered lines).
   • Compute First-Cycle Service Level (probability of no stockout during cycle) if using inventory-theory targets.
4. Segment and localize issues.
   • Heatmap fill rates by node × ABC class and by customer tier; identify chronic low-fill SKUs and long-tail behavior.
   • Compare e-comm vs B2B; ship-from-store vs DC; single-line vs multi-line orders; promotion vs non-promo periods.
5. Attribute root causes with a mutually exclusive tree.
   • Inventory not available: forecast error/bias, safety stock too low, delayed receipts, MOQ/lot size constraints, network mis-position.
   • Allocation/prioritization: rationing to higher-tier customers, ship-complete constraint, substitution not allowed/used.
   • Execution exceptions: short picks, inventory inaccuracy, late put-away, pick/pack capacity (missed cut-offs).
   • Policy/design: node selection rules causing splits with zero stock, order promising ignoring true lead times (ATP/CTP miscalibration).
   • Use timestamps, ATP at order time, cycle count variance, PO ETA slippage, and exception codes to assign root cause per imperfect line.
6. Quantify impact and trade-offs.
   • Estimate revenue at risk = units backordered/cancelled × ASP; premium freight triggered by backorders; chargebacks from low fill.
   • Relate Fill Rate vs inventory investment (stock turns, days of supply) to identify over/under-stocking and target MEIO opportunities.
7. Plan vs actual service analysis.
   • Compare achieved Fill Rate to planned service levels (e.g., CSL targets by class). Identify gaps where safety stock or lead-time parameters are mis-set.
   • Decompose variance: plan → forecast error → supply delay → allocation → execution errors.
8. Scenario and what-if.
   • Model Fill Rate change from: +15% safety stock on A SKUs, MEIO repositioning across nodes, improved forecast accuracy (MAPE −20%), supplier LT −2 days, enable customer-approved substitutions, add nightly cut-off, and fix pick accuracy.
   • Quantify inventory and cost impacts (working capital, carrying cost, premium freight saved).
9. Integrity checks.
   • Ensure “first shipment” is correctly identified; treat partials consistently; exclude cancel-before-ship from denominator or report separately.
   • Verify ATP reconstruction logic; confirm substitution handling aligns with policy; validate cycle count accuracy in low-fill SKUs.

Format of the output of analysis:

• Executive scorecard: UFR, LFR, OFR (overall and by channel/customer tier), Backorder Rate and Age, Lost Sales %, revenue at risk, trend vs target.
• Heatmaps: Fill Rate by node × ABC class and by SKU family; bottom quartile SKUs/customers highlighted.
• Root-cause bridge: contribution of inventory gaps, allocation/rationing, execution exceptions, and policy/design to total shortfill.
• Inventory-service panel: service vs inventory (DoS/turns) by class; CSL target vs actual; forecast error (MAPE/bias) overlay.
• Scenario deck: projected Fill Rate uplift and inventory/cost trade-offs from MEIO, safety stock changes, forecast improvement, substitutions, and process fixes.

How to interpret results:

• High Fill Rate (≥95–98%) on A SKUs with healthy turns: Right-sized inventory and solid execution; focus on long-tail optimization and cost reduction.
• Low Fill on A SKUs with high lost sales: Under-investment, poor forecast or supply unreliability—raise safety stock, improve forecast, secure supply first.
• Low Fill on C/long-tail SKUs with high inventory: Mis-positioned stock; consider MEIO, rationalization, or substitution strategies.
• OFR much lower than LFR: Partial shipments common; review ship-complete policies and pick/pack capacity at cut-offs.
• Fill gaps concentrated in certain nodes or periods: Node capacity/put-away delays or promo planning gaps; rebalance inventory and labor, pre-build for peaks.
• Near-100% Fill with rising obsolescence: Over-stocking; reduce safety stock or broaden substitution/ship-from-store to use inventory efficiently.

Steps a company can take to improve on this measure:

• Inventory strategy and positioning:
  • Implement multi-echelon inventory optimization (MEIO) with CSL targets by ABC class and customer tier; reposition stock across nodes.
  • Recalibrate safety stock (z × σ of LT demand) using current lead-time variability and forecast error; review MOQs and ordering cadence.
• Forecasting and demand shaping:
  • Improve forecast accuracy with causal models and promo flags; monitor bias; pre-allocate for promotions and key accounts.
  • Offer customer-approved substitutions or variant mapping to increase effective fill on constrained SKUs.
• Supply and replenishment reliability:
  • Shorten/ stabilize supplier lead times; introduce VMI/consignment for A SKUs; dual-source critical items; enforce PO confirmation and ASN timeliness.
  • Increase order frequency for volatile SKUs to reduce exposure to forecast error.
• Allocation and promising:
  • Align allocation to margin/strategic tiers; protect A SKUs for top customers; enable dynamic ATP/CTP with real-time inventory.
  • Refine node selection (nearest-node, ship-from-store) to boost local fill and reduce splits.
• Warehouse execution and accuracy:
  • Improve pick accuracy (scan compliance, put-to-wall, weigh-check); accelerate put-away for inbound to raise same-day availability.
  • Increase cut-off frequency (waveless release) and labor flexing at peaks to avoid short picks at first ship.
• Governance and data quality:
  • Daily fill-rate huddles on bottom-quartile SKUs; root-cause closure; cycle counting focused on high-variance items.
  • Maintain a substitution/assortment policy and track customer acceptance rates; tune rules accordingly.
• Scenario guidance:
  • If A-SKU LFR is 89% with frequent short picks, raise safety stock 15%, accelerate put-away (target ≤4h), and add a 10 p.m. cut-off; expect LFR ≥96% within 6–8 weeks.
  • If long-tail UFR is 78% with high inventory, enable cross-node fulfillment and customer-approved substitutions; target UFR ≥90% while reducing DoS by 20%.
  • If OFR trails LFR by 12 pts due to ship-complete policy, segment customers: relax ship-complete for e-comm, keep for key retailers; improve pack capacity at peaks.

Benchmark comparisons:

General benchmarks (directional):

• E‑commerce make-to-stock: UFR/LFR typically 94–98% for mature programs; top quartile ≥97–99% on A SKUs.
• Retail/B2B replenishment: Line Fill ≥96–98% on core assortment; Order Fill often lower (85–95%) where ship-complete is enforced.
• ABC targets: A ≥97–99%, B ≥95–97%, C ≥90–95% (with substitutions/assortment rationalization).

Constructing internal benchmarks:

• Track weekly UFR/LFR/OFR by node/channel/customer tier and ABC class; publish quartiles and set redlines (e.g., A-SKU LFR <97% triggers action).
• Pair Fill Rate with forecast error, safety stock attainment, inventory turns, and lost sales; prioritize SKUs/nodes with low fill and high revenue at risk.
• Set service–inventory target curves by class; revisit quarterly after promotions, network changes, or supplier lead-time shifts.
• Monitor backorder age and recovery; aim to clear ≥90% of backorders within X days (by SLA) to protect customer experience.