Why scheduling matters more than counting
Most cycle count programmes fail for a scheduling reason rather than a counting reason. Counting itself is straightforward: someone with the right sheet goes to a location and records what is there. The hard part is deciding which items to count, how often, and whether the resulting workload can actually be delivered by the people available.
A schedule that demands more counts than the team can perform does not fail loudly. It fails quietly. Counts are skipped, then rescheduled, then quietly dropped. Within a few months the programme exists on paper while key items have not been counted in a year.
A good schedule has three properties: it targets the items where error matters most, it repeats at a frequency justified by that item's risk, and its total workload fits within realistic daily capacity.
Count frequency should follow the cost of being wrong, and the plan must fit the counting hours you actually have.
Setting intervals from ABC class and risk
The usual starting point is ABC classification, which groups items by consumption value rather than unit count. A relatively small number of items typically account for most of the value moved, and those items justify the most frequent counting.
| Class | Typical share of items | Common count frequency | Counts per year |
|---|---|---|---|
| A | Around 20% | Monthly | 12 |
| B | Around 30% | Quarterly | 4 |
| C | Around 50% | Annually | 1 |
Those frequencies are conventional, not mandatory. ABC class captures value, but it does not capture all risk. Two items in the same class can deserve very different counting attention if one is a controlled substance, a serialised component or a part prone to damage. For how the classes themselves are derived, see the ABC Inventory Analysis Guide.
A practical approach is to start from the ABC baseline and then apply adjustments for specific risk factors:
Raise the frequency
High theft exposure, high unit value, serialised or regulated stock, parts with a history of repeated discrepancies, and items in dispersed or congested locations.
Lower the frequency
Slow-moving, low-value, bulky items that are difficult and time-consuming to count, provided they have no history of unexplained variance.
Do not use frequency as a punishment
If an area is error-prone because of broken process rather than genuine risk, more counting will not fix it. Counting measures the problem; it does not solve it.
Turning intervals into a daily workload
Count frequencies are only useful once converted into work. The arithmetic is simple and should be done before committing to a schedule. For each item, divide the number of counts per year by the number of counting working days in the year.
Items per day = (Annual counts for item) ÷ (Counting working days per year)
Sum that figure across every item in scope and you have the total number of item counts the plan demands each day. Compare it with the number the team can realistically complete, allowing for travel between locations, opening and resealing, locating stock behind other stock, and recording findings properly.
Two details commonly distort this estimate:
Counting days are not calendar days
If counting only happens on weekdays, the divisor is roughly 250, not 365. Using calendar days understates the daily workload by around 30%.
An "item count" is not a fixed unit of work
Counting ten units of a well-organised pallet position takes far less time than counting ten units scattered across a bulk area. Workload should be estimated in time, not just in lines.
Planning within real capacity
This is where most plans break. The required daily count volume is compared against theoretical staffing rather than the hours genuinely available after receiving, put-away, picking and dispatch have taken their share.
A defensible capacity figure is the number of counting hours per day that survive a normal week, including the days when someone is absent, a delivery arrives late or a stock issue demands immediate attention. Planning against a perfect day guarantees slippage.
Budget counting hours, not headcount
Two counters available for one hour is two hours of capacity, not two days of work. Converting to hours makes the comparison honest.
Include the non-counting parts
Locating, moving stock, recording and reconciling are all part of the work. A count that is performed but not properly recorded has not reduced risk.
Reserve contingency
Leave capacity unallocated for investigations. A variance found by a count still needs someone to investigate and correct it.
When required workload exceeds available capacity, there are only three honest responses: reduce the scope of the programme, accept a lower frequency for some items, or add counting resource. Quietly pretending the schedule will fit is the fourth option, and it is the one that fails.
The DIXANI Cycle Count Planner automates exactly this step: it takes an inventory list, count intervals and available working days, and produces a dated schedule within your
Timing counts and controlling stock movements
When a count happens matters as much as how often. Counting an item while stock is actively moving produces a variance that reflects the movement, not an error. The schedule should reduce that interference rather than create it.
Count when movement is naturally low
Early shifts, quiet periods or outside peak picking hours reduce the chance of counting stock that is halfway out of the door.
Freeze or flag the location
A short freeze, or at minimum a clear flag that the location is being counted, prevents put-away and picking from changing the quantity mid-count.
Count the location, not just the item
An item in the wrong location appears as a shortage in one place and a surplus in another. Counting by location surfaces placement errors that item-only counting hides.
Keep blind counts
Counters who can see the expected quantity tend to confirm it. Withholding the expected figure makes the count a measurement rather than a confirmation.
Worked scheduling example
Starting figures
A warehouse has 1,200 active items split roughly 20/30/50 across ABC classes, and counting happens on weekdays only. Using the conventional frequencies, the annual count volume is:
Class A
240 items × 12 counts = 2,880 counts per year
Class B
360 items × 4 counts = 1,440 counts per year
Class C
600 items × 1 count = 600 counts per year
Total
4,920 counts per year
With roughly 250 weekdays available, the plan requires about 20 item counts per day. If each count takes an average of six minutes including travel and recording, that is around two hours of counting work every single working day.
4,920 counts ÷ 250 days = 19.7 counts/day, at 6 minutes each ≈ 2 hours/day
That is the number to test against reality. If only one hour per day is genuinely available, the plan as written is not achievable, and no amount of intention will change that. The options are to reduce Class C scope, lengthen intervals for lower-risk items, or find additional counting time.
Notice also what happens if the plan silently assumes 365 calendar days: the target drops to about 13.5 counts per day, or roughly 1.4 hours. That looks comfortably achievable, which is precisely why the error is dangerous. The plan appears to fit right up until it starts slipping.
A schedule is only useful if the count itself is performed properly. The Warehouse Cycle Count Checklist covers the counting, recount and adjustment steps, while the Inventory Accuracy Guide shows how to measure whether the programme is actually improving accuracy. Buffer levels feed the same decision from the other direction; see the Safety Stock Guide.
Common mistakes
Scheduling against perfect conditions
Capacity should reflect a normal week including absences and interruptions, not the best possible day.
Using calendar days for a weekday operation
It understates the daily requirement by around 30% and hides the fact that the plan does not fit.
Counting everything at the same frequency
A uniform interval ignores value and risk, spending effort on low-consequence items while critical ones wait.
Never revising the schedule
Item mix, locations and staffing change. A schedule built once and never adjusted drifts out of fit within a year.
Not acting on variances
A count that finds a discrepancy and does not trigger investigation and correction has consumed time without reducing risk.
Treating skipped counts as harmless
Each skipped count pushes risk forward. Persistent slippage is a signal that the plan, not the team, is wrong.
Questions and answers
How often should items be cycle counted?
Frequency should follow the cost of being wrong. A common starting point is monthly for Class A, quarterly for Class B and annually for Class C, then adjusted upward for risk factors such as high value, theft exposure or a history of repeated discrepancies.
Should cycle counting replace an annual physical inventory?
Where a council-approved perpetual inventory programme exists, structured cycle counting can often replace the full annual count. The requirements differ by jurisdiction and by the standards your auditors apply, so the replacement should be agreed with them in advance.
How many counts per day can one person complete?
That depends on travel, accessibility and how carefully findings are recorded. Estimating in hours rather than lines is far more reliable: measure how long a representative set of counts actually takes, then multiply.
What should happen when a count finds a variance?
The count should trigger investigation before any adjustment. Most variances have an explainable cause, such as a mis-placed item, an unrecorded transaction or a misidentified part, and correcting the cause is more valuable than adjusting the number.
Why does my schedule keep slipping?
Almost always because the required workload exceeds genuine capacity. The plan was built against theoretical availability rather than the hours that survive a normal week. Recalculating the daily requirement in hours and comparing it honestly is the first step.
Can I count during normal working hours?
Yes, and it is often necessary. The requirement is to control movement during the count, not to stop operations. A short location freeze or a clear in-progress flag is usually enough.
Use free DIXANI resources
Build the schedule online, then run the count and measure the result using these practical resources.