Cycle counting: how to count stock without shutting the warehouse
In short
Cycle counting checks a rotating slice of your stock each week so inventory records stay accurate without shutting the warehouse down. Classify stock by value using ABC analysis, set a count frequency for each category, and investigate every discrepancy rather than just correcting the number.
- Cycle counting checks a small slice of your inventory on a rotating schedule, so the warehouse keeps running and records stay accurate year-round.
- ABC analysis determines which items get counted most often: high-value A items weekly, mid-range B items monthly, and low-value C items quarterly.
- Inventory accuracy should be calculated after every session using the formula: Physical Count divided by Recorded Count, multiplied by 100; the target is 97% or higher.
- Every discrepancy above your tolerance threshold needs a root-cause investigation, not just a record correction - most errors trace to receiving or data entry, not shrinkage.
- A cycle counting programme that misses more than one or two consecutive weeks loses its value; consistency over a short session beats thoroughness over an infrequent one.
Cycle counting checks a small slice of your inventory every week, without stopping the warehouse. This guide explains how to classify stock by value, set a count schedule, investigate discrepancies, and track inventory accuracy year-round.
Cycle counting is a method of auditing your inventory by checking a small, predetermined slice of stock on a regular schedule - weekly or even daily - instead of stopping everything once a year for a wall-to-wall count. Rather than counting every SKU in one exhausting session, you rotate through your entire inventory in smaller batches over the course of several weeks or months, so the warehouse keeps moving while the count happens around it. Which items get counted first depends on value: high-value, fast-moving lines go to the top of the schedule, because an error there costs more and compounds faster. Done consistently, cycle counting replaces the annual stocktake as the primary mechanism for keeping inventory records accurate.
Why the Annual Stocktake Creates the Problem It's Meant to Solve
A full physical stocktake sounds thorough. In practice, it concentrates twelve months of potential error into a single high-pressure session. Staff rush, counts get skipped, and the results reflect one day's snapshot rather than the reality of stock that has been moving all year. By the time the numbers are reconciled, the business has been operating on stale records for months.
The financial cost of inaccurate stock is significant. Research cited by inventory software provider Cin7 puts the global cost of inventory distortion - the combined damage from stockouts and overstocking - at roughly $1.73 trillion a year. Most of that damage accumulates between counts, not during them.
For businesses in categories like promotional merchandise, where goods are often ordered to a tight spec, stored in branded configurations, and dispatched in time for client events, an error found in October about stock that went missing in June is too late to do anything useful with. Cycle counting surfaces discrepancies within days rather than months.
Step 1: Run a Baseline Count to Know Where You Stand
Before you can set a cycle counting schedule, you need an honest picture of your current inventory accuracy. Run a full count - not to replace the cycle counting programme that follows, but to establish a starting point.
Use the inventory accuracy formula to measure the result: divide the physical count by the recorded count and multiply by one hundred. A score of 83% means your records are wrong on roughly one in six lines. Businesses should aim for 97% or higher to support reliable order fulfilment, according to Unleashed Software's inventory management guide.
Record your baseline. That number is what the cycle counting programme is designed to improve.
Run a full baseline inventory count
Count every SKU physically and compare the results against what your system shows. Calculate your inventory accuracy rate using the formula: Physical Count divided by Recorded Count, multiplied by 100. Write down the result. This is your starting point - not a cause for alarm, but the number you are working to improve.
Step 2: Classify Your Stock Using ABC Analysis
ABC analysis sorts your inventory into three tiers based on value and movement. The split is not arbitrary - it determines how often each group gets counted.
Category A covers the items that represent the largest share of your total inventory value, typically 70 to 80 percent of value from around 20 percent of your SKU count. These get counted most often. Category B covers mid-range items, roughly 15 to 20 percent of total value. Category C covers everything else - high volume but low value, or slow movers that account for around 5 percent of total stock value.
For businesses that sell kits or bundled products, the kits and variants feature in your inventory system should reflect how component stock is tracked, because errors in kit components flow straight into order accuracy.
Classify every SKU into category A, B, or C
Pull a report that shows your inventory by total value (unit cost multiplied by average quantity held). Sort descending. Mark the top 20 percent of SKUs by value as A items. The next tier - roughly the middle 30 percent - becomes B. Everything else is C. Review the list once a year, as categories shift as your product mix changes.
Step 3: Set Your Count Frequency by Category
With categories assigned, set a counting frequency for each. A common schedule is A items counted weekly, B items counted monthly, and C items counted quarterly. Businesses with very high A-item values or significant shrinkage risk may count A items daily.
The goal is to complete at least one full rotation through the entire inventory before your year end. At that point, cycle counting effectively replaces the annual stocktake as a control measure. Cin7 recommends running a full cycle at least once per quarter, and targeting 95 to 98 percent accuracy across all counts.
Set counting frequencies for each category
Write down: A items - weekly. B items - monthly. C items - quarterly. Calculate how many SKUs fall into each category and divide by the number of available counting sessions in each period. That gives you the number of lines to count per session, which you can adjust if it exceeds what your team can cover in the time available.
Step 4: Build a Weekly Count Schedule
A cycle counting programme that lives only in someone's head will stop running within a month. Build a structured schedule and assign it to a specific role.
Choose a fixed window each week - early morning before picking begins, or during a quiet mid-afternoon period. Schedule sessions to last no more than one to two hours. List which SKUs are due that week, along with their bin locations, so the counter can move efficiently through the warehouse without hunting for items.
Suspend stock movements on the specific lines being counted during each session. That means pausing receipts, picks, and adjustments for those SKUs while the count is in progress. Do not freeze the whole warehouse - only the specific lines being counted.
Create and publish a weekly count schedule
List the SKUs due for counting in each session, grouped by location so counting is physically efficient. Assign a named person to own the schedule. If you rotate who does the counting - which reduces the risk of errors being covered up - make sure every counter has access to the same schedule and knows which lines they are responsible for.
Step 5: Execute the Count and Freeze Transactions on Active Lines
On count day, the counter works through the list: locates each SKU, counts the physical quantity, and records the result. Barcode scanning or a mobile inventory app reduces entry errors and speeds up the session.
The counter should never see the system's recorded figure before completing the physical count. Seeing the number first introduces bias - counters unconsciously count toward what the system expects.
Count physically before looking at system records
Block or hide the system quantity until the physical count is recorded. Count every item in the bin or location - do not trust quantity labels on packaging without opening and verifying. Mark each line as counted to prevent double-counting. Once the session is complete, reveal the system quantity and note any discrepancy.
Step 6: Reconcile Variances and Find the Root Cause
A discrepancy between a physical count and a system record is not itself the problem. It is a signal. The useful question is not how many are missing, but why the system does not know what the count just confirmed.
Common root causes include receiving errors (goods booked in at the wrong quantity), picking errors (the wrong number taken for an order), shrinkage (theft or damage not recorded), and data entry mistakes. Each cause has a different fix. Treating every discrepancy as potential theft - when most are receiving or entry errors - wastes time and misses the real issue.
Investigate every variance above your tolerance threshold
Set a tolerance - for example, any discrepancy greater than 2 percent by value triggers an investigation. For discrepancies within tolerance, adjust the record and move on. For larger variances, trace the stock movement history for that SKU back to the last confirmed count. Look for unrecorded adjustments, delivery notes with incorrect quantities, or picks that do not match dispatch records.
Step 7: Update Records and Log the Reason for Every Adjustment
Once the investigation is complete, update the inventory record to match the physical count. Do not adjust the record first and investigate second - the sequence matters. Changing a record before you understand why it is wrong means you lose the audit trail.
After updating, review whether the discrepancy points to a process that needs fixing. A receiving error on the same SKU in three consecutive counts suggests the problem is at the goods-in stage, not in the count itself.
Update records systematically and log the reason for each adjustment
Every adjustment should carry a note explaining why it was made. A note that reads "cycle count correction" is not useful. A note that reads "count confirmed 47 units; system showed 52; traced to receiving shortfall on incoming PO booked in full" is. That audit trail is what lets you spot patterns over time.
Step 8: Track Your Inventory Accuracy Rate and Refine the Schedule
After each count session, calculate the session accuracy rate using the same formula as the baseline. After four to six weeks of data, you will have enough to spot which categories, locations, or SKUs produce the most errors - and whether your overall accuracy is trending up or down.
Use Zigaflow's inventory feature to track stock movements, flag discrepancies, and keep a running record of what has been counted and when. That live view means you can see, at any point, which lines are overdue for a count and which are producing repeated variances.
A target of 97% inventory accuracy, tracked week by week, is a meaningful operational metric. If accuracy falls below 95%, a one-off full count alongside the cycle programme is worth considering to reset the baseline.
Review your inventory accuracy rate weekly and adjust the schedule
Calculate the accuracy rate after each session. Track the trend. If a category consistently produces errors, increase the counting frequency for that group until accuracy improves. If sessions consistently finish early, you may be able to add more SKUs per session. The schedule should evolve as the programme matures.
Common Pitfalls That Stall Cycle Counting Programmes
The most common reason cycle counting programmes fail is not accuracy - it is consistency. Counting happens for three weeks, then a busy period pushes it off the schedule, and by the time it restarts, the data is stale enough that nobody trusts it.
The second common failure is using the programme only to correct records, without acting on the root causes those corrections reveal. If the same SKU produces a variance every month, that is a process problem, not a counting problem.
A third risk is over-counting. A schedule that demands three hours per day from a small warehouse team quickly gets deprioritised when orders come in. Start with sessions that fit within one hour and scale up once the process is embedded.
Cycle counting is not the same as a full annual stocktake, and in most small business settings it is a better tool for daily inventory accuracy. But for businesses where external audit or regulatory compliance requires a full count at year end, the cycle counting programme supports that process rather than replacing it - by the time the formal count arrives, the discrepancies will be small enough to clear quickly.
Sources
- Inventory Cycle Count: Methods, Steps & Best PracticesCin7 · accessed 2026-10-13
- Guide to Cycle Counting & Stock Taking in InventoryUnleashed Software · accessed 2026-10-13
- Cycle counting vs. annual stocktake: how to set a count scheduleTana Inventory Management · accessed 2026-10-13
- WMS Cycle Counting vs Annual Stocktake: Which Keeps Your Numbers AccurateBlue Lotus 360 · accessed 2026-10-13