5.2 Inventory Control Operating Model (KPIs, Ownership, Audit Cadence)

Inventory is money frozen in time and risk sitting on a shelf. Without a rigid control model, entropy guarantees that physical stock and digital records will drift apart, leading to phantom inventory, line-down events, and massive financial write-offs at year-end. You cannot wish for accuracy; you must engineer it through a closed-loop system of measurement, detection, and forced correction. This chapter defines the management architecture that monitors the warehouse execution described in Part 6.

The Control Loop: KPIs & Definitions

Do not measure everything—measure only what drives behavior and signals failure.

Core KPI Architecture

1. Inventory Record Accuracy (IRA): The probability that the system quantity matches the physical quantity within tolerance.
   • Target: ≥ 99% for A-Items; ≥ 98% aggregate.
   • Calculation: (Total Correct Counts / Total Counts Performed) × 100.
2. Cycle Count Variance (CCV): The financial impact of errors.
   • Absolute Variance: (Sum of absolute $ discrepancies). Measures operational chaos.
   • Net Variance: (Sum of +/- $ discrepancies). Measures financial loss.
   • Threshold: Absolute Variance must be < 0.5% of total inventory value counted.
3. Dock-to-Stock Time (DTS): The velocity from the receiving dock to a purchasable bin location.
   • Target: < 24 hours (Standard); < 4 hours (Priority/Shortage).
   • Risk: If DTS > 24h -> Planning creates duplicate orders (phantom shortages).
4. Pick Error Rate: The frequency of incorrect pulls detected by downstream consumers (Production/Kitting).
   • Target: < 500 PPM (Parts Per Million lines picked).

Pro-Tip: Do not confuse "Net Variance" with operational excellence. A +$10k error and a -$10k error cancel out financially (Net $0), but they indicate a broken physical process. Track Absolute Variance to judge warehouse discipline.

Ownership & Segregation of Duties

The fox cannot guard the henhouse. Operational execution must be separated from audit and financial reconciliation.

Warehouse Manager (Execution Owner):

• Owns: Physical velocity, bin discipline, safety, and execution of counts.
• Accountable for: Dock-to-Stock time, housekeeping, and labor efficiency.

Supply Chain / Materials Lead (System Architect):

• Owns: The counting schedule (Cycle Count logic), discrepancy investigation, and root cause analysis.
• Accountable for: IRA (Accuracy) and preventing recurrence of errors.

Finance (The Gatekeeper):

• Owns: Validation of large write-offs and annual audit compliance.
• Accountable for: Ensuring inventory value on the Balance Sheet is real.

Audit Cadence & Decision Logic

Abandon the "Annual Physical Inventory" shutdown. It is costly, disruptive, and statistically inferior to continuous Cycle Counting.

The Cycle Count Routine

Use the ABC classification (defined in 5.1) to drive frequency.

• If Item is Class A (High Value/Turnover) -> Count Monthly (12x/year).
• If Item is Class B (Mid Value) -> Count Quarterly (4x/year).
• If Item is Class C (Low Value) -> Count Annually (1x/year) or on 0-balance trigger.

Discrepancy & Escalation Logic

When a count does not match the system, strictly follow this logic path:

1. First Fail: Counter reports discrepancy -> System flags location as "Counting."
2. Blind Recount: Assign a different person to count the bin without seeing the system quantity.
   • If Recount matches System -> Close (First count was human error).
   • If Recount matches First Fail -> Variance Confirmed.
3. Reconciliation Trigger:
   • If Variance Value < $50 -> Adjust immediately (Auto-approval).
   • If Variance Value $50 – $1,000 -> Requires Materials Lead review & sign-off.
   • If Variance Value > $1,000 -> Requires Ops Director + Finance sign-off + RCA Report.

Root Cause Analysis (RCA) Mandate

Do not just adjust the balance. You must identify why it drifted. Assign a code to every adjustment:

• RC-01: Receiving Error (Wrong Qty/Label).
• RC-02: Pick/Put-away Error (Wrong Bin).
• RC-03: BOM/Backflush Error (Engineering Data Wrong).
• RC-04: Supplier Packaging (Pack Size Variance).

Pro-Tip: If you see frequent "BOM Errors" (RC-03), the warehouse is innocent. The Engineering team must fix the consumption ratios in the ERP.

Dashboard Specification (The Weekly Pulse)

Management does not need raw data; they need a single-page signal of health. Construct a "Materials Control Tower" dashboard with the following quadrants.

Quadrant	Metric Focus	Signal / Alert
Top Left	Accuracy (Quality)	IRA Trend (12 weeks): Bar chart. Red line at 98%. Net Variance ($): Monthly cumulative loss/gain.
Top Right	Velocity (Flow)	Dock-to-Stock Average: Hours. Warning if > 24h. Aged WIP: Number of jobs sitting in "Picked" status > 48h.
Bottom Left	Exceptions (Risk)	Negative Balances: Count of SKUs with < 0 qty (System failure). Uncounted A-Items: % of A-class not counted in last 30 days.
Bottom Right	Capacity (Space)	Bin Utilization: % of bins occupied. Quarantine Age: Avg days inventory has sat in MRB/Hold.

Final Checklist

Control Point	Setting / Rule	Non-Negotiable
Primary KPI	Inventory Record Accuracy (IRA)	Target ≥ 99% for A-Items.
Cycle Count Freq	A=Monthly, B=Quarterly, C=Yearly	Do not skip A-items. They drive 80% of the risk.
Authority	Segregation of Duties	Warehouse staff cannot authorize their own write-offs > $50.
Escalation	Financial Threshold	Variance > **$1,000** triggers mandatory detailed investigation.
Detection	Negative Balance Rule	System must block picking from negative locations; force a count immediately.
Drift	Dock-to-Stock	Material sitting > 24 hours on dock must be flagged as "At Risk."