Stock counting is a dreaded task performed once a year in most businesses. The warehouse is closed, the team works over the weekend, and the resulting discrepancies are heavily debated.
The problem is not the count itself, but its frequency. When errors accumulated over an entire year are put in front of you all at once, finding the root cause becomes impossible.
In this article, we explain a system that breaks the count into small pieces, the rules for counting day, and methods for interpreting discrepancies.
Table of Contents
Why partial counting?
Annual full-scale counting has three major problems. First, it halts operations; second, it exhausts the team; third, and most importantly, it makes finding the root cause of discrepancies impossible.
If a product has not been counted for twelve months, you cannot know when a discrepancy of ten units occurred. Tracing back records becomes meaningless.
With partial counting, however, the same product is counted every month or every quarter. Discrepancies turn out small, and the cause is sought in the recent past. The probability of finding it is much higher.
In addition, operations do not stop. While one aisle is being counted, shipments continue from other aisles.
How to set up a count plan?
There is no need to count all products at the same frequency. Group products based on value and turnover speed.
High-value and fast-moving products can be counted once a month. This group usually accounts for a small portion of the product count but makes up the majority of the inventory value.
The middle group can be counted every three months, and low-value, slow-moving products once a year. This grouping significantly reduces the total counting effort.
Turnover rate is the primary input for this grouping; we covered the calculation in the inventory turnover rate article.
In multi-warehouse structures, the plan must be established on a warehouse basis; you can check out the multi-warehouse inventory management article.
Pre-count preparation
The quality of a count is determined during preparation. Three things must be completed before counting.
Close pending receipts. If there are unentered waybills or unprocessed transfers, discrepancies will stem from them. Clearing the queue before counting is essential.
Ensure physical organization. Having the same product located on three different shelves lengthens the count and leads to duplicate counting.
Resolve missing barcodes. A product without a barcode means manual entry; manual entry means errors. See the Barcode management article.
Counting day rules
Movements during the count are the biggest source of discrepancies. There are two ways to manage this.
The first way is freezing: inbound and outbound movements are halted in the counted zone. This is applicable in partial counting because only a few shelves are affected.
The second way is recording: movements made while counting continues are marked separately and taken into account in discrepancy calculations. This method does not stop operations but requires strict discipline.
Whichever method you choose, write down the rule in advance and share it with the team. A rule debated on counting day invalidates the count.
Who performs the count and who approves it must also be separated. Allowing the person who counts to fix discrepancies single-handedly eliminates control; authorization ensures this separation.
Barcoded and mobile counting
Counting with paper lists means double data entry: first on paper, then into the system. Every transfer generates new errors.
In mobile device counting, the product is scanned, the quantity is entered, and the record is created instantly. The person counting may not see the expected quantity on the screen—this is a deliberate choice; if they see it, they look at the screen instead of counting.
Barcode scanning is particularly valuable for products with variants. Distinguishing colors and sizes visually leads to mix-ups; check out the variant product tracking article.
For products with lot and serial tracking, counting verifies not only the quantity but also which batch is present; the batch and lot tracking article addresses this topic.
Post-count: reading the discrepancies
When the count is over, the real work begins. The discrepancy list is not a correction list, but a diagnostic tool.
Before immediately closing discrepancies, classify their source: is it missing records, physical loss, miscount, or shrinkage? Each requires a different action.
Once the adjustment receipt is issued, the discrepancy disappears; therefore, analysis must be done first. We detailed the method in the how to close stock count discrepancies article.
In retail, a significant portion of discrepancies consists of shrinkage and waste; you can check out the shrinkage and waste tracking article.
Count and adjustment records are linked to receipts in the inventory module; this is how retrospective querying becomes possible.
Building the counting team
The quality of a count depends on the person doing the counting. Therefore, who makes up the team and how they will work must be determined in advance.
A fundamental principle is the tendency of individuals counting their own area to overlook discrepancies. When the warehouse supervisor counts their own shelf, they are likely to brush off a missing product by assuming it is "around somewhere." Cross-counting reduces this risk.
Working in pairs is also a common method: one counts, the other records. When working with mobile scanning, this need diminishes; a single person both scans and records, increasing speed.
A brief briefing should be held before the count. When what will be counted, how partial packages will be recorded, and what to do with damaged products are not discussed, every counter acts differently.
The approval step must rest with a separate person. Allowing the counter to close discrepancies single-handedly removes the control function of the count; authorization supports this separation.
Measuring count quality
Counting is also a process and can be measured. Without measurement, there is no improvement; every count will reproduce the same problems.
The most fundamental indicator is the accuracy rate: what percentage of counted items had no discrepancies? If this rate is not increasing over time, there is a bottleneck in the process.
The second indicator is the discrepancy value. A small number of discrepancies in terms of quantity can be large in terms of value; small discrepancies in high-value products must be monitored separately.
The third is the repetition rate: how many consecutive counts showed discrepancies for the same product? Recurring discrepancies are the clearest sign of a process issue and require root cause analysis.
These indicators can be extracted from count receipts. Recording discrepancy reasons categorically enables analysis; we explained this recording discipline in the count discrepancy article.
Frequently asked questions
Does partial counting replace annual counting?
Operationally, yes. Consult your financial advisor regarding legal inventory obligations.
Who should perform the count?
If possible, someone not responsible for that area. The person counting their own area may overlook discrepancies.
Can sales continue during the count?
In partial counting, yes. Movements outside the counted zone are not affected.
How do count results reflect in reports?
Adjustment receipts are recorded as movements and appear on the statement; check out the stock statement article.
Transforming counting from a major once-a-year event into a weekly routine both minimizes discrepancies and makes their causes discoverable.
By meeting with the EQLEM team, you can design your warehouse counting plan and mobile counting workflow.