What is cycle counting in inventory?

Cycle counting is counting a small part of the warehouse's stock on a schedule, such as a few bins or SKUs each day, instead of counting everything at once. Every location gets counted over a rolling cycle. Each count is compared with the system quantity, and differences are recounted, explained and corrected, so the stock number every channel sells from stays trustworthy.

Cycle count vs physical inventory (stock take, wall-to-wall): what is the difference?

A physical inventory, or wall-to-wall stock take, counts everything at once, usually once or twice a year, and the warehouse often stops shipping while it runs. A cycle count covers a few bins at a time, continuously, while orders keep moving. Cycle counts find and fix errors through the year; a stock take confirms total stock at a date.

Which cycle counting methods are there (ABC, random, location-based)?

ABC counting counts high-value or fast-moving items most often. Location-based counting walks the warehouse aisle by aisle so every bin is covered each cycle. Random sampling counts a random set of bins to measure accuracy without bias. Opportunity-based counting checks a bin when it is nearly empty or a pick finds a shortage. Most D2C warehouses combine ABC with opportunity counts.

How often should you cycle count?

Count your top-selling or highest-value SKUs (A items) about monthly, B items quarterly and C items twice a year, then adjust to your error rates. Divide each class by the working days in its cycle to get a daily count. In our example, 1,000 SKUs need about 15 counts a day. Count more often before sale events.

How do you calculate inventory accuracy from a cycle count?

Divide the number of bin-SKU counts that matched the system by the number of counts made. If 950 of 1,000 counts match, location accuracy is 95%. Also track the gross variance value, the total value of every overage and shortage without netting, because a net figure close to zero can hide large errors in both directions.

What do you do when a count doesn't match the system?

Don't adjust straight away. Have a second person recount the bin. If the difference holds, check recent picks, putaways, returns and receipts for that bin to find the cause, such as a mis-pick or stock put in the wrong bin. Then post the adjustment with a reason code, and fix the process that caused it.

Cycle count vs perpetual inventory: how do they relate?

Perpetual inventory means the system updates stock with every receipt, pick, return and adjustment, so it always shows a current number. Cycle counting is how you check that number against the shelf. The two work together: perpetual inventory gives every channel a current figure, and cycle counts catch where it has drifted from what is physically there.

Can cycle counting replace the annual physical stock count, and what do auditors need?

Sometimes, but that is your auditors' call, so ask them before dropping the year-end count. Make their decision easy: count every location at least once a year, keep dated count sheets or scans, record every variance, recount and adjustment with a reason code, and show your accuracy trend. A well-run cycle count programme usually makes any remaining year-end count faster.

What is a cycle count in inventory?

A cycle count is a scheduled count of a small part of a warehouse's inventory, such as a set of bins or SKUs each day, so that every location gets counted on a rolling cycle without shutting the warehouse. Each count is compared with the system's stock, and any difference is investigated and corrected.

This page is about counting warehouse inventory. It isn’t about battery cycle counts on phones or laptops.

How cycle counting works in an Indian D2C or omnichannel warehouse

A brand selling on its website, two or three marketplaces and quick commerce publishes one stock number to every channel. If the shelf and the system disagree, a channel sells stock that isn’t there, or hides stock that is. Cycle counting is the routine that keeps the two in line:

Plan. Decide which bins or SKUs to count today, using one of the methods below.
Freeze or note movement. Count a bin when no pick or putaway is open on it, or record the open tasks so they can be allowed for.
Count blind. The counter scans the bin and records what is there, without seeing the system quantity first.
Compare. The system flags any difference between the count and its stock.
Recount. A second person recounts any bin that doesn’t match before anything is changed.
Find the cause and adjust. Check recent picks, putaways, returns and GRNs for the bin, fix the cause, then post the adjustment with a reason code. Large adjustments need a supervisor’s approval.

Cycle counting methods

Method	How it picks what to count	Good for
ABC	Counts fast-moving or high-value “A” items more often than “B” and “C” items	Most D2C warehouses
Location-based	Walks the warehouse aisle by aisle so every bin is covered once per cycle	Finding misplaced stock
Random sample	Counts a random set of bins each day	Checking accuracy without bias
Opportunity-based	Counts a bin when it is nearly empty or a pick finds a shortage	Low-effort, frequent checks

Example: count frequency and the accuracy formula. A warehouse holds 1,000 SKUs, split as below (illustrative):

Class	SKUs	Count every	Counts per working day (about 22 a month)
A (top sellers)	100	Month	100 ÷ 22 ≈ 5
B	300	Quarter	300 ÷ 66 ≈ 5
C	600	Six months	600 ÷ 132 ≈ 5

So the team counts about 15 SKUs a day. Over a quarter, it records 1,000 bin-SKU counts, and 950 of them match the system. Location accuracy = 950 ÷ 1,000 = 95%. The 50 mismatches are the work list: each is recounted, traced to a cause and corrected.

Cycle count vs physical inventory (wall-to-wall stock take)

	Cycle count	Physical inventory
Scope	A few bins or SKUs at a time	Everything at once
Frequency	Daily or weekly, rolling	Usually once or twice a year
Operations	Warehouse keeps shipping	Warehouse often stops
Purpose	Find and fix errors continuously	Confirm the full stock value at a date

Common mistakes

Showing the counter the system quantity. People find what they expect to find; count blind.
Adjusting without a recount. A single miscount becomes a real stock error on every channel.
Fixing the number but not the cause. If the same bin is wrong every month, the putaway or pick process at that bin is broken.
Counting while picks are open on the bin. The count and the system are measuring different moments.
Counting only slow movers because they are easy. Errors concentrate where stock moves fastest.

Location (bin) accuracy = bin-SKU counts that match ÷ bin-SKU counts made. Example: 950 ÷ 1,000 = 95%.
Gross variance value = sum of the value of every unit over and under, with no netting. Example: 40 units short at ₹300 and 30 units over at ₹200 gives ₹12,000 + ₹6,000 = ₹18,000 gross, even though the net is ₹6,000.
Count completion = counts done ÷ counts planned for the period.

Cycle counting in the Indian context

Many channels, one stock number. Indian brands often sell the same stock on their own site, Amazon, Flipkart, Myntra or Meesho and quick-commerce apps. A wrong bin quantity reaches all of them through the stock feed. Read why system and physical stock drift in inventory sync across channels.
Sale events. Count A items before a big marketplace sale, when a stock error turns into cancellations fastest.
Returns and RTO. Returned parcels that sit unprocessed are a common cause of “missing” stock; count the returns area too.
Annual stock take still matters. Cycle counting doesn’t replace the full count your auditors may ask for; it makes that count faster and less surprising.

How Fretron CommerceOS handles inventory accuracy

Bin-level accuracy is live in the Fretron CommerceOS warehouse module. Cycle counts in Fretron CommerceOS are blind counts taken on the phone and posted from the office, not updated with each scan. See the warehouse management system page for what the module covers.

Questions and answers

What is cycle counting in inventory?: Cycle counting is counting a small part of the warehouse's stock on a schedule, such as a few bins or SKUs each day, instead of counting everything at once. Every location gets counted over a rolling cycle. Each count is compared with the system quantity, and differences are recounted, explained and corrected, so the stock number every channel sells from stays trustworthy.
Cycle count vs physical inventory (stock take, wall-to-wall): what is the difference?: A physical inventory, or wall-to-wall stock take, counts everything at once, usually once or twice a year, and the warehouse often stops shipping while it runs. A cycle count covers a few bins at a time, continuously, while orders keep moving. Cycle counts find and fix errors through the year; a stock take confirms total stock at a date.
Which cycle counting methods are there (ABC, random, location-based)?: ABC counting counts high-value or fast-moving items most often. Location-based counting walks the warehouse aisle by aisle so every bin is covered each cycle. Random sampling counts a random set of bins to measure accuracy without bias. Opportunity-based counting checks a bin when it is nearly empty or a pick finds a shortage. Most D2C warehouses combine ABC with opportunity counts.
How often should you cycle count?: Count your top-selling or highest-value SKUs (A items) about monthly, B items quarterly and C items twice a year, then adjust to your error rates. Divide each class by the working days in its cycle to get a daily count. In our example, 1,000 SKUs need about 15 counts a day. Count more often before sale events.
How do you calculate inventory accuracy from a cycle count?: Divide the number of bin-SKU counts that matched the system by the number of counts made. If 950 of 1,000 counts match, location accuracy is 95%. Also track the gross variance value, the total value of every overage and shortage without netting, because a net figure close to zero can hide large errors in both directions.
What do you do when a count doesn't match the system?: Don't adjust straight away. Have a second person recount the bin. If the difference holds, check recent picks, putaways, returns and receipts for that bin to find the cause, such as a mis-pick or stock put in the wrong bin. Then post the adjustment with a reason code, and fix the process that caused it.
Cycle count vs perpetual inventory: how do they relate?: Perpetual inventory means the system updates stock with every receipt, pick, return and adjustment, so it always shows a current number. Cycle counting is how you check that number against the shelf. The two work together: perpetual inventory gives every channel a current figure, and cycle counts catch where it has drifted from what is physically there.
Can cycle counting replace the annual physical stock count, and what do auditors need?: Sometimes, but that is your auditors' call, so ask them before dropping the year-end count. Make their decision easy: count every location at least once a year, keep dated count sheets or scans, record every variance, recount and adjustment with a reason code, and show your accuracy trend. A well-run cycle count programme usually makes any remaining year-end count faster.

GRN (goods received note)
Putaway: stock put in the wrong bin is the classic count variance
WMS (warehouse management system)
FIFO and FEFO
Batch and lot tracking
RTO (return to origin): unprocessed RTO parcels are a common count variance
Bin location, inventory shrinkage and overselling: in the Commerce glossary (full guides in later batches)
Backorder
Dark store
Overselling
Inventory shrinkage
Bin location
Inventory turnover ratio
Inventory reconciliation

See CommerceOS on your own data.

Book a demo →

Last reviewed: 5 November 2026 by Fretron.

Sources

No statutory or marketplace rules are stated on this page. Formulas and examples are illustrative.