A stock take counts every item in the warehouse at one time, usually with operations paused, typically once a year for the accounts. A cycle count checks a small, rotating set of SKUs or locations every day or week, so the whole inventory is covered without stopping work. Most warehouses need both: cycle counts for day-to-day accuracy, and a stock take for the year-end audit.
Inventory accuracy is invisible when it is good and expensive when it is not. Phantom stock shows available but cannot be found. Lost sales hide in “system says zero”. Write-offs appear at year-end. The cure is not counting more, but counting the right things at the right frequency and fixing why the numbers drift. This guide covers both methods, how to design a cycle count programme, and how to measure whether it works.
Stock Take vs Cycle Count at a Glance
| Full stock take | Cycle count | |
|---|---|---|
| Scope | Every SKU and location | A rotating subset |
| Frequency | Annual or half-yearly | Daily or weekly |
| Impact on operations | Receiving and dispatch usually frozen | Minimal: counted alongside normal work |
| Main purpose | Financial statements and audit | Ongoing accuracy and root-cause correction |
| Error detection | Once a year, after the damage is done | Continuous, close to when the error happened |
| Labour | Large team for one to several days | Small, regular effort |
The Full Stock Take
A full physical count supports the value of inventory in the financial statements, and auditors often observe it. Under the Malaysian accounting standard for inventories (MFRS 102), inventory must be stated at the lower of cost and net realisable value, so an accurate quantity is the foundation. A good stock take needs:
- Cut-off: stop receiving and dispatch, or record exactly which documents fall before and after the count.
- Preparation: tidy locations, consolidate part pallets, label every location, and resolve open transactions.
- Count teams in pairs: one counts, one records, with team members swapped between zones.
- Blind counts: counters see the location and SKU but not the system quantity, to avoid confirmation bias.
- Recounts: any variance above tolerance is recounted by a different team before adjustment.
- Reconciliation and approval: adjustments are reviewed and approved before they are posted.
Designing a Cycle Count Programme
Step 1: Classify SKUs (ABC)
| Class | Typical share of SKUs | Typical share of value or movement | Count frequency |
|---|---|---|---|
| A | ~10–20% | ~70–80% | Monthly (or more) |
| B | ~20–30% | ~15–20% | Quarterly |
| C | ~50–70% | ~5–10% | Once or twice a year |
Rank by annual consumption value (units × cost) for finance-led programmes, or by pick frequency for operations-led programmes. High-risk items, such as small high-value goods, theft-prone products and items with expiry dates, can be moved up a class.
Step 2: Add Trigger Counts
- When a location shows zero or negative stock.
- When a picker reports a short pick or an empty location.
- After a damaged receipt or a returns batch.
- When stock falls to its reorder point, when there is least to count.
Step 3: Schedule Daily
Divide the count plan into daily lists so every A, B and C item is covered in its cycle. Count early or late in the shift, when locations are least active, and count blind.
Step 4: Set Tolerances
Define what counts as a match, for example an exact match for A items and ±2% for low-value C items, and require a recount above tolerance.
Measuring Accuracy
Inventory record accuracy (IRA) = locations (or SKUs) counted with no variance outside tolerance ÷ total counted × 100
Track it weekly, by class and by zone. Well-run warehouses target 98% or higher at location level, and A items should be close to 100%. Also track the value of net and absolute adjustments. A small net figure can hide large offsetting errors.
Fixing the Causes, Not Just the Count
An adjustment fixes today’s number. Root-cause analysis stops tomorrow’s error. Common causes:
| Cause | Fix |
|---|---|
| Receiving without checking quantity against the delivery note | Scan-to-receive and blind receiving |
| Putaway to the wrong location | Scan the location on putaway |
| Picking the wrong item or quantity | Scan-to-pick and a pack check |
| Unrecorded damage or samples | Damage bins with a daily posting routine |
| Unit-of-measure confusion (carton vs piece) | Clear UOM set-up and labelling |
| Returns not booked in | A dedicated returns location and process |
Most of these are prevented by barcode scanning at every movement, a core feature of a warehouse management system. Our
warehouse best practices guide covers the wider operating discipline.
Can Cycle Counting Replace the Annual Stock Take?
For some businesses, yes, partly. If a documented cycle count programme covers all inventory through the year and accuracy is consistently high, auditors may accept it in place of a full count, or reduce the scope of the full count. This is a decision to agree with your auditors in advance. It should not be assumed.
Counting When Stock Is with a 3PL
When your stock is in a third-party warehouse, accuracy is a shared responsibility. Agree these points in writing:
- the cycle count frequency by class, and the annual stock take date;
- the accuracy target and how it is reported;
- who investigates and approves adjustments, and within what time;
- who bears the cost of shrinkage above an agreed tolerance;
- your right to attend or audit counts.
Our guide to choosing a 3PL warehouse covers the wider service-level agreement.
Inventory Accuracy with De Hubs
De Hubs’ warehousing service includes inventory management and stock reporting. Our warehouse management system tracks stock in real time, so you always know what is in the warehouse across our sites in Klang, Kuching, Miri and Kota Kinabalu. Tell us your count requirements and reporting needs, and we will build them into your service.
Frequently Asked Questions
What is the difference between a stock take and a cycle count?
A stock take counts every item in the warehouse at one time, usually with operations paused. A cycle count checks a small, rotating set of items or locations every day or week, so the whole inventory is covered over a period without stopping work.
How often should I do a stock take?
Most businesses do a full stock take at least once a year, at financial year-end, for their accounts and audit. Warehouses with a strong cycle-count programme and good accuracy can sometimes reduce reliance on full counts, if their auditors agree.
How do I calculate inventory accuracy?
Inventory record accuracy is the number of counted locations or SKUs whose system quantity matches the physical count (within tolerance), divided by the total counted, × 100. Well-run warehouses target 98% or higher.
What is ABC cycle counting?
ABC cycle counting ranks SKUs by value or movement. A items, the few that matter most, are counted most often, for example monthly. B items are counted quarterly, and C items once or twice a year.
Does a 3PL do cycle counts for me?
A good 3PL runs regular counts as part of inventory management and gives you stock reports. Agree the count frequency, accuracy target and variance process in the contract.