5.5 Cycle counting & reconciliation

Inventory data accuracy is the foundation of reliable factory operations. If the ERP system indicates sufficient material exists but the physical shelf is empty, production will stop. Conversely, overstating physical inventory leads to unnecessary material purchases. A continuous Cycle Counting regime provides a more reliable and less disruptive approach to inventory validation than a traditional “Annual Physical Inventory.”

The cycle count architecture (“ABC” logic)

Counting efforts must be aligned with the financial and operational risk profile of the inventory using the “ABC” stratification logic.

Frequency rules

• Class A (High Value / Critical): Expected to be counted frequently, typically Monthly (12x/year). These items represent the most value and carry the highest risk if misaligned.
• Class B (Medium Value): Expected to be counted moderately, typically Quarterly (4x/year).
• Class C (Low Value / Bulk): Expected to be counted less frequently, typically Annually (1x/year), or employ a “Count on Zero” trigger (a task generated specifically when the system indicates < 5 units remaining).

Sample size calculation

It must be ensured the warehouse has the capacity to execute the scheduled cycle counts without falling behind.

• The Calculation: [(Total A SKUs × 12) + (Total B SKUs × 4) + (Total C SKUs × 1)] / 250 Working Days = Daily Count Target.
• The Execution: A routine must be established where cycle counts are performed early in the shift, ideally before significant order picking occurs, to ensure a clean data state.

Operational execution

Cycle counting should ideally occur while the warehouse remains operational, requiring careful systemic controls.

1. Blind Counts (Recommended): The counting mechanism (RF scanner or paper) should not display the expected ERP quantity to the operator. This prevents confirmation bias and encourages an honest physical count.
2. Bin Freeze: During an active cycle count task, the specific bin location should be digitally locked in the WMS to prevent picking or put-away until the count is complete and posted.
3. Zero-Crossing Check: If an operator finds a bin empty during regular picking, but the ERP indicates positive stock, this should trigger an immediate “Emergency Count” flag to resolve the discrepancy.

Discrepancy protocol & root cause analysis

When a cycle count identifies a variance, the cause must be investigated rather than simply adjusting the system balance.

Step 1: the blind recount

• The Trigger: A difference between the physical count and the system record is detected.
• The Protocol: The count task must be assigned to a different operator.
• The Resolution: If the second count matches the system, the first count was likely an error; no adjustment is needed. If the second count confirms the initial physical count, investigate further.

Step 2: transaction audit

• The Trigger: The recount confirms a real variance exists.
• The Protocol: The Materials Lead should review recent transactions (Receipts, Picks, Adjustments). It should be looked for:
  • Pending receipts (material present but not systematically posted).
  • Open pick jobs (material allocated but not yet moved).
  • Unit of Measure (UOM) errors (e.g. counting sealed reels vs. individual components).

Step 3: adjustment & categorization (RCA)

When posting an inventory adjustment, a reason code must be assigned to help identify systemic issues.

WMS Code	ERP Category	Definition	Typical Process Owner
RC-RX	Receiving Error	Incorrect quantity received, or supplier label mismatch.	Receiving Team
RC-PK	Picking Error	Material pulled from the wrong bin, or incorrect quantity picked.	Warehouse Operations
RC-BM	BOM / Process	Standard backflush does not align with actual physics (Engineering structure incorrect).	Manufacturing Engineering
RC-SC	Unreported Scrap	Components damaged during assembly but not formally recorded as scrap.	SMT Production Leadership
RC-VN	Vendor Variance	A sealed reel contained the incorrect physical quantity.	Sourcing / Quality

Pro-Tip: If RC-BM (BOM Error) constitutes a large portion of variances, it suggests an issue with the Engineering Bill of Materials rather than warehouse execution. Collaborate with Engineering to correct the consumption logic.

The control outputs

Structured reporting must be used to monitor the health of the cycle count program.

1. The cycle count calendar

A dynamic schedule driven by the ERP/WMS engine, pushing tasks to scanners automatically. Completion percentage must be tracked to ensure the annual count plan remains on track.

2. Variance KPI dashboard

• IRA (Inventory Record Accuracy): Target > 98.0%.
• Absolute Variance Value: Financial discrepancies must be tracked as a percentage of total inventory value.
• Recount Rate: The percentage of counts requiring a recount. High rates may indicate training gaps or poor bin discipline.

3. The corrective action log (CAPA)

A record of chronic, recurring errors must be maintained to drive systemic improvements.

• Example: “Consistent shortages identified on specific 0402 passives. Action: Re-calibrate physical reel counters in the kitting area.”

Final Checkout: Cycle counting & reconciliation

Control Point	Engineering / Process Requirement	Target Goal
Verification Gate	Utilize cycle counts to validate critical stock.	Especially prior to large MRP buys.
Authority	Implement financial approval limits for write-offs	Escalate large adjustments to management.
Discipline	WMS hides expected quantities during counting.	Minimizes confirmation bias.
Timing	Execute cycle counts consistently.	Daily execution prevents backlog.
Feedback Loop	BOM errors are communicated to Engineering.	Generates corrective PLM tickets.