How to Do Inventory Counts and Cycle Counts
Full Physical Count vs Cycle Counting
A full physical inventory count means stopping all warehouse operations and counting every single product in your entire warehouse on a single day. This approach provides a complete snapshot of inventory accuracy but requires shutting down order fulfillment for a full day or weekend, which means no orders ship during the count. For ecommerce businesses where customers expect fast shipping, a full-day shutdown is costly in both lost sales and delayed orders. Full physical counts are appropriate once per year (often at fiscal year-end for accounting purposes) but should not be your primary accuracy maintenance tool.
Cycle counting counts a small subset of your inventory each day or week, rotating through your entire catalog over a defined period. On any given day, you might count 20 to 50 SKUs rather than your entire 500-SKU catalog. The counts happen during normal operations, typically in the morning before order picking begins, so fulfillment is not disrupted. Over 1 to 3 months, every SKU in your warehouse gets counted at least once, and your most important products get counted multiple times. Cycle counting maintains accuracy continuously rather than discovering problems once a year, which means errors get caught and corrected within days or weeks instead of compounding for months.
Setting Up a Cycle Count Program
The standard inventory accuracy target for ecommerce operations is 98%, meaning 98% of your SKUs should have a physical count that exactly matches the system count at any point in time. Some operations target 99% or higher, especially for high-value products or businesses with thin margins where a single mispick erases the profit on the order. Use ABC classification to set counting frequencies: A items (top 20% by revenue) get counted weekly, B items (middle 30%) get counted biweekly or every 3 weeks, and C items (bottom 50%) get counted monthly. This concentrates counting effort on the products where accuracy matters most while still covering everything.
Before counting a section of your warehouse, ensure that all pending inbound receipts for products in that section have been processed and put away in their correct bin locations. If 50 units arrived yesterday but have not been shelved yet, your count will show 50 fewer units than the system expects, creating a false discrepancy. Similarly, make sure all picked orders have been removed from shelves before counting. If an order was picked from bin A-03-02 but has not been shipped yet and the units are sitting on the packing table, those units are not in the bin but are still in your system as available. The ideal counting window is after morning receiving is processed and before order picking begins for the day.
Count every unit in each bin location for the scheduled section. The most important rule: do not look at the expected system quantity before counting. If you know the system says there should be 47 units, human nature pushes you to count 47 even if there are actually 44. Count blind, record the number, and then compare to the system. For products stored in large quantities, count by tens or by container: if you know each shelf box holds 25 units and you have 3 full boxes plus a partial, count the partial box and calculate the total (75 plus 12 equals 87). Use barcode scanning where available, because scanning the bin location and product barcode before entering the count confirms you are counting the right product in the right location.
After completing the physical count, compare each SKU's physical count to the system quantity. Any variance, even a single unit, should be investigated before adjusting the system. Common root causes of discrepancies include: receiving errors (counted wrong when inventory arrived), picking errors (wrong product picked for an order, reducing the count of one product and leaving the correct product in place), put-away errors (product placed in the wrong bin location, so it shows as missing in the expected bin but exists elsewhere), unreported damage or breakage, and returns that were not properly processed back into inventory. Investigate before adjusting, because the adjustment fixes the symptom but understanding the root cause prevents recurrence.
Once you have verified the physical count and identified the root cause of any discrepancy, adjust the system quantity to match the physical reality. Every adjustment should be logged with the date, the SKU, the old system quantity, the new quantity, the variance, and the identified or suspected root cause. Most inventory management platforms have built-in stock adjustment logs that record this information. Over time, these logs reveal patterns: if the same product consistently shows up short, there may be a systematic receiving or picking issue specific to that product. If a particular bin location frequently has discrepancies, the bin might be poorly labeled or too close to a similar product, causing confusion.
Common Counting Mistakes to Avoid
The biggest mistake is counting while operations are running and products are moving. If pickers are pulling products from shelves during your count, your numbers will not be reliable because the system quantities are changing in real time while you are trying to match them. Either count before operations start for the day, count after operations end, or freeze the counted section (no picks from those locations) for the 15 to 30 minutes it takes to complete the section count.
Another common mistake is adjusting the system quantity without investigating the discrepancy. If your system says 100 and you count 95, it is tempting to just change the system to 95 and move on. But those 5 units did not evaporate. They are either in the wrong location, were shipped to a wrong customer, were damaged and not reported, or were miscounted during receiving. If you do not find the root cause, the same error will likely recur, and you will be making adjustments every cycle count without ever fixing the underlying process.
Do not skip cycle counts during busy periods. The busiest fulfillment periods (holiday season, promotional events, product launches) are exactly when inventory accuracy matters most and when errors are most likely to occur because of the higher volume and pace. Maintaining your cycle count schedule during peak periods catches the errors that fast-paced operations inevitably generate, and prevents those errors from compounding across the busiest weeks of the year.
Tools for Cycle Counting
The simplest counting tool is a printed count sheet with bin locations and product names, and a pen. This works for businesses with under 100 SKUs and a single small storage area. For anything larger, barcode scanning with a mobile device or dedicated scanner speeds the process dramatically and reduces recording errors. Most inventory management platforms include a mobile cycle count feature that generates the count schedule, guides the counter to each location, accepts scanned or manually entered counts, flags discrepancies automatically, and processes adjustments with an audit trail.
For Shopify sellers without dedicated inventory software, Shopify's Stocky app (available on Shopify POS plans) includes cycle counting functionality. WooCommerce sellers can use plugins like ATUM Inventory Management that add cycle count features to the WordPress admin. For spreadsheet-based operations, create a cycle count template that lists the date, section to count, each SKU and bin location in that section, columns for system quantity and physical count, a variance column, and a notes field for root cause documentation.