Warehouse Stocktaking and Cycle Counting: How to Resolve Inventory Discrepancies
Practical guide to warehouse stocktaking, cycle counts, variance investigation, recounts, shrinkage controls and inventory reconciliation.

Practical guide to warehouse stocktaking, cycle counts, variance investigation, recounts, shrinkage controls and inventory reconciliation.
Why Warehouse Stocktaking Needs More Than a Single Number
Warehouse stocktaking is not just an exercise in counting boxes. It is a controlled process for comparing what the warehouse should hold with what is physically present at a specific time. In a system such as Rigi24, the core concepts that matter are warehouse, location, stock count, and movement. Those basics are enough to organize a reliable reconciliation process, but the discipline around them must come from warehouse practice: precise cut-off timing, careful counting, and documented investigation of differences.
Shortages and overages are inevitable in real operations. They may come from mispicks, damaged packaging, misplaced stock, receiving errors, or incorrect unit conversions. The practical goal is not to “make the numbers fit” at any cost. The goal is to establish a trustworthy inventory snapshot, explain every difference that can be explained, and only then approve a correction where appropriate.
Physical Inventory Count, Cycle Count, and Spot Count
Physical inventory count: the full snapshot
A physical inventory count is the most complete inventory verification method. It usually covers the entire warehouse at one time, or at least a defined inventory scope such as one facility, one section, or one product family. Its purpose is to establish a full stock position against which accounting and operational records can be reconciled.
This method is appropriate when the warehouse needs a formal year-end or period-end balance, after major layout changes, after a system migration, or after repeated discrepancies have made smaller checks unreliable. The strength of the physical count is completeness. The weakness is cost: it consumes time, labor, and operational flexibility, especially if order movements continue during counting.
Cycle count: smaller scope, more frequent control
A cycle count is a scheduled count of a subset of locations, SKUs, or value bands. Instead of stopping everything for one large event, the warehouse counts targeted items regularly. For example, a fast-moving router SKU might be counted every week, while slow-moving cables may be counted monthly or quarterly.
Cycle counts are useful when the warehouse wants ongoing control without full shutdown. They are cheaper per session and allow quicker detection of problems. Their weakness is that they may miss broad structural issues if the selection is too narrow or if exceptions are ignored. A cycle count works best when it is part of a deliberate control plan, not a random look-around.
Spot count: a quick verification tool
A spot count is a narrow, often ad hoc check of one SKU, one bin, one location, or one suspicious difference. It is commonly used to verify a receipt, confirm a pick, investigate a complaint, or validate a doubtful location after a scanner discrepancy.
Spot counts are practical, fast, and low-cost. They are not a substitute for a formal count because they do not provide a complete snapshot. Their value is in speed and diagnosis. If a location that should hold 40 units shows 4 units, a spot count can confirm whether the issue is a misshelving problem, a damaged pack, or a recording error.
| Method | Scope | Main purpose | Strength | Typical limitation |
|---|---|---|---|---|
| Physical inventory count | Entire warehouse or major section | Official snapshot and reconciliation | Most complete view | Disruptive and labor-intensive |
| Cycle count | Selected SKUs or locations | Ongoing control and error detection | Regular oversight with lower disruption | May overlook wider issues |
| Spot count | Single item or location | Immediate verification | Fast diagnosis | Not a full inventory picture |
Preparing an Exact Inventory Snapshot
A count is only as good as its cut-off. The warehouse must define the exact date and time at which the inventory position is frozen for reconciliation purposes. Every movement after that point belongs to the next period or must be separately identified. Without this discipline, the warehouse can count the same goods twice or miss them entirely.
Good practice is to record the count date, start time, and end time for each counted area or batch. If the warehouse is counting over several hours, that timing detail matters. A pallet moved from receiving to storage at 10:15 should not be counted in the receiving area if the snapshot was taken at 10:00, even if the pallet physically passed through the door before the team arrived.
Managing movements during the count
Movements during a count create the most common reconciliation failures. A warehouse may be tempted to keep shipping, receiving, and putaway running while the team counts. If that happens, every movement must be tightly documented and separated from the count records. The safest operational approach is to minimize movement in the counted area while the count is in progress, then record any unavoidable exceptions with time stamps and reference notes.
For example, if two cartons of access points were received at 11:20 while the count started at 11:00, the count sheet should clearly show whether they were excluded from the snapshot or counted separately as post-cut-off movement. A vague note such as “arrived during count” is not enough. The record should identify who moved the goods, from where, to where, and at what time.
Count sheets and location ownership
Count sheets are the working documents used to record the physical result. A well-designed count sheet should include the SKU, location code, condition, expected quantity if it is hidden or on a separate control sheet, counted quantity, timestamp, observer, reason for variance if known, and any attachment such as a photo or receiving document reference. The purpose is not only to capture the number but also to preserve the context around it.
Location ownership matters because one location should have one accountable status at the moment of count. If three teams are using the same shelf area for temporary staging, the count becomes unreliable. Each counted location should have a clear owner or operational status: storage, quarantine, returns, damaged, or staging. A location code without ownership discipline can lead to duplicate counting, missed stock, or false shortages.
Case Study: A Tbilisi Warehouse with Cables, Access Points, and Routers
Consider a hypothetical warehouse in Tbilisi holding networking products in three locations: A-01, B-14, and C-03. The stock includes sellable cables, sellable access points, and damaged routers awaiting review. The warehouse uses Rigi24 to manage warehouse and location records, stock counts, and movements, but the count process itself still depends on disciplined physical procedure.
Suppose the system shows the following expected inventory before the count:
| SKU | Location | Condition | Expected |
|---|---|---|---|
| CABLE-5M | A-01 | Sellable | 120 units |
| AP-AX1200 | B-14 | Sellable | 38 units |
| RTR-AC300 | C-03 | Damaged | 6 units |
The first count is done blind, meaning the counters do not see the expected quantity while counting. The A-01 shelf is physically found to contain 12 inner boxes of cables, each box containing 10 units, plus 3 loose units. That equals 123 units. If the expected was 120, there is an overage of 3 units. The team should not immediately assume the system is wrong. They should first check whether one box was opened and re-packed, whether a carton was placed in the wrong shelf, or whether the conversion from boxes to pieces was entered incorrectly.
At B-14, the team finds 37 access points, not 38. A one-unit shortage may come from a pick error, a damaged unit moved to quarantine, or a receipt that was counted twice elsewhere. At C-03, the damaged router shelf holds 5 units, not 6. That difference could be legitimate if one unit was scrapped after the expected quantity was recorded, or it could indicate that a damaged unit was moved out without a movement record.
Blind Counting, Independent Recount, and Segregation of Duties
Blind counting reduces bias. When the counter does not know the expected amount, the physical observation is less likely to be influenced by the system number. This is especially useful when a location has a history of recurring differences. The counter should read the location code and SKU, count the physical stock, and record the result without seeing the expected quantity.
An independent recount is a second physical count by a different person or team. It is not the same as simply changing the first number. The purpose is to verify whether the discrepancy is real. If the first count says 123 and the second says 123 again, confidence increases. If the second says 121, the warehouse must investigate why two trained people saw different results. The difference may be caused by hidden stock behind a front row, mixed packaging, or a misread label.
Segregation of duties is crucial. One person should not both count and approve the correction without review. A practical workflow is: counter one records; counter two independently verifies; supervisor reviews evidence; only then is an approved adjustment made in the system if justified. This is a recommendation for control, not a claim about any automatic feature.
SKU, Batch, GTIN, Location Code, and Packaging Conversions
Counting errors often begin with identity confusion. A SKU is the warehouse’s internal item identifier. A GTIN identifies a trade item in a standardized way. A batch or lot ties stock to a production or receipt group. A location code identifies where the stock is stored. These are not interchangeable.
Imagine the warehouse has two product presentations: boxes of 10 cables and single loose cables. If the location shows 12 boxes, the counted quantity may be entered as 12 or 120 depending on the unit of measure. That is why conversion discipline matters. A count sheet should state the unit clearly: boxes, pieces, or both. If the shelf contains 11 full boxes and 4 loose pieces, the physical count is 114 pieces, but only if each box is confirmed to contain 10 pieces. If one box is partially open or damaged, the conversion assumption may be wrong.
Common conversion failure cases
- Counting packaging units as individual units without confirming inner pack quantity.
- Recording sellable and damaged stock together in one line.
- Mixing two SKUs with similar labels in the same bin.
- Using the wrong location code after stock was relocated temporarily.
- Assuming a GTIN on the carton always matches the interior unit count without verification.
Shelf Labels, Barcode Scanning, and Verification Limits
Shelf labeling is one of the simplest but most powerful control tools. A clear label on the shelf and a readable barcode can reduce picking mistakes and help counters confirm that they are in the right place. However, scanning a label or an ID is not proof that the physical stock has been reconciled. It only proves that a code was read.
For example, a counter may scan location C-03 and then scan a router label. That confirms the location and item identification, but it does not prove the quantity, condition, or ownership status. The warehouse still needs a physical count and a documented decision process. A barcode scan can support verification, but it does not replace the count result or the reconciliation note.
Good practice is to scan after counting, not instead of counting. The scan can help avoid writing the wrong SKU or location on the sheet, but the core evidence remains the physical observation and the recorded quantity.
Count Sheet Fields That Actually Matter
A useful count sheet is simple, but not simplistic. It should capture enough detail to explain the count and support later review. The following fields are especially valuable:
- SKU – the item identity being counted.
- Location – the exact shelf, bin, or zone code.
- Condition – sellable, damaged, quarantined, returns, or other local status.
- Expected – hidden during blind count or kept on a separate control sheet.
- Counted – the physical number observed.
- Timestamp – when the count was made.
- Observer – who performed the count.
- Reason – known or suspected cause of variance.
- Attachments – photos, receiving slips, damage notes, or transfer references.
Checklist for a reliable count workflow
- Define the count scope and the exact cut-off time.
- Freeze or clearly document any unavoidable movements during the count.
- Prepare labels, count sheets, and location maps in advance.
- Use blind counting where practical.
- Separate the first count from the independent recount.
- Record condition, unit of measure, and location code on every line.
- Investigate every shortage or overage before approving any correction.
- Keep attachments and notes for traceability.
From Difference to Decision
Shortages and overages are not just accounting problems; they are operational signals. A shortage may reveal picking loss, misplacement, unrecorded damage, or a transaction cut-off problem. An overage may reveal duplicate receipts, cross-location mixing, or a packaging conversion error. The right response is to investigate first and adjust only after evidence supports the decision.
In a warehouse like the hypothetical Tbilisi site, the most useful habit is to treat every discrepancy as a question: Was the location ownership clear? Was the count blind? Was the unit of measure correct? Did a movement happen during the snapshot window? Was the damaged stock isolated from sellable stock? Those questions turn a simple count into a controlled reconciliation process.
Why stocktaking errors matter in a warehouse
Warehouse stocktaking is not only a counting exercise; it is a control process that tests whether the book balance in the system matches the physical quantity on the shelf, in the yard, or in another storage area. For Rigi24 users, the core concepts involved are straightforward: warehouse, location, stock count, and movement. When the count does not match the book, the difference must be understood before any adjustment is approved.
In practice, discrepancies usually come from ordinary operational causes: goods received but not entered, items issued but not posted, transfers in transit, broken units set aside, or stock sitting in the wrong location. The purpose of stocktaking is not to “force” the numbers to agree. It is to explain why they do not agree and then correct the record responsibly.
How to calculate shortages and overages
The arithmetic should be simple and consistent. Use the same sign convention every time so the team can read reports without confusion:
- Shortage = physical count is lower than book count.
- Overage = physical count is higher than book count.
- Discrepancy can be shown as physical minus book, so negatives indicate shortages and positives indicate overages.
Consider the first item: book count 125 units, physical count 117 units. The difference is 117 minus 125, which equals -8. That means a shortage of 8 units. The second item: book count 40 units, physical count 46 units. The difference is 46 minus 40, which equals +6. That means an overage of 6 units.
Many warehouses prefer to record both the direction and the amount so there is no ambiguity. For example:
- Item A: book 125, physical 117, variance -8, shortage 8
- Item B: book 40, physical 46, variance +6, overage 6
This convention is especially useful when multiple locations are counted on the same day. The count sheet, the recount sheet, and the approved adjustment should all use the same logic.
Investigating differences systematically
A discrepancy is not solved by guessing. A disciplined review should ask a sequence of questions. The same checklist works for both shortages and overages.
- Was the item counted in the correct warehouse and location?
- Was any transfer created but not completed?
- Were returns received physically but not recorded, or recorded but not physically present?
- Was an issue to production, a customer, or another branch posted correctly?
- Was the supplier receipt entered with the right quantity and item code?
- Is damaged, quarantined, or scrap stock mixed with sellable stock?
- Could the same movement have been entered twice?
A practical example from a fictional warehouse in Tbilisi: the system shows 60 cartons of a detergent, but the physical count in the main aisle is only 52 cartons. Before assuming loss, the team checks a receiving note, a transfer note, and the issue register. They discover 8 cartons were moved to a back room location during repacking but the movement was never recorded. The shortage is not real loss; it is a location error.
Wrong location, unrecorded transfers, and duplicate operations
One of the most common causes of apparent shortage is stock sitting in the wrong place. In a warehouse with several locations, an item may be counted in the receiving area, bulk rack, quarantine zone, or dispatch lane even though the system expects it elsewhere. If the count team only checks one location, the item can appear missing.
Unrecorded transfers create a similar problem. Suppose 30 units are moved from central storage to a picking bin, but the movement is done physically before it is entered in the system. The receiving location appears short, while the destination appears overstocked. The total warehouse quantity may still be correct, but the distribution by location is wrong.
Duplicate operations can distort the book balance in the opposite direction. If a receipt is entered twice, the system may show 100 units when only 50 were actually received. During stocktaking, the physical count may look short by 50 units, but the real problem is a duplicated receipt. The corrective action should address the source document, not just the final variance.
Returns, broken goods, and scrap stock
Returned items require special attention because they are often physically present but not yet fit for sale. If a customer return is placed on the rack with saleable inventory, the count may look correct while the quality status is wrong. The reverse also happens: a return may be sitting in a separate area but still absent from the system.
Broken goods and scrap should be separated from sellable inventory as a daily discipline, not only during stocktaking. For example, a branch in Batumi may have 12 units of glassware on the floor shelf, but 3 are cracked and placed in quarantine. The physical count of sellable stock is 9, while the total on site is 12. If the system does not distinguish these statuses, reconciliation becomes confusing.
A useful operational rule is to label these groups clearly:
- Sellable stock: available for normal issue or sale
- Quarantine stock: held for inspection or decision
- Repair stock: temporarily unusable but recoverable
- Scrap stock: not intended to return to saleable inventory
This separation helps the count team avoid mixing true inventory with unusable items. It also helps management understand whether a shortage is actually a quality-status issue rather than a loss.
Physical count versus book balance during online sales
Simultaneous online sales can make stocktaking more difficult because orders, reservations, and shipment events occur while the count is in progress. The main risk is counting stock that has already been committed to a customer but not yet physically removed, or excluding stock that is still present but already reserved.
Imagine a warehouse with 100 units of a fast-moving item. Before the count begins, 18 units are reserved for online orders and 7 units are picked during the day for shipment. If the count happens while these operations are active, the book balance may change multiple times. A count team may physically see 82 units on the shelf, while the system shows 75 available after reservations and picks. The difference does not automatically mean theft or error; it may simply reflect timing.
The operational lesson is to review the event timing carefully. Ask when the reservation was created, when the pick was confirmed, when the shipment left, and when the count was taken. A count snapshot must be compared with the correct book snapshot, otherwise a legitimate movement can be mistaken for a discrepancy.
Rigi24’s confirmed concepts of warehouse, location, stock count, and movement are relevant here because every movement changes the inventory picture. However, decisions about how to pause work, sequence counts, or control online order timing are operating procedures, not confirmed software automation features.
How to triage discrepancies by materiality
Not every variance deserves the same level of investigation. A warehouse should prioritize differences based on monetary importance, stock value, materiality to operations, and customer promises. A high-value product with a small unit difference may matter more than a larger unit difference on low-cost goods.
For example, a shortage of 2 units of an expensive electronic part may affect a pending customer order and require immediate review. By contrast, an overage of 10 units of a low-cost consumable may be less urgent if the branch has no immediate demand risk. The point is not to use a universal tolerance, because warehouses differ in value, turnover, and service commitments. Instead, assess the operational impact case by case.
A practical triage checklist may look like this:
- Does the discrepancy affect a promised customer shipment?
- Is the item high-value or strategically important?
- Does the difference suggest a likely process failure that could repeat?
- Is the stock in a sensitive status such as quarantine, return, or scrap?
- Is the issue confined to one location or spread across several locations?
This approach avoids spending the same effort on every variance. It also helps management focus on the risks that matter most to service and inventory integrity.
ABC cycle count prioritization in practice
Cycle counts work best when the warehouse counts more frequently where the risk is highest. A simple ABC approach can help prioritize without promising a rigid formula. In a fictional Georgian distribution operation, A items are high-value or high-demand products, B items are moderate priority, and C items are routine low-risk items.
A practical weekly and monthly pattern could be:
- Every week: count the top A items in central Tbilisi, especially fast-moving or high-value stock.
- Every two weeks: count selected B items or items with recent discrepancy history.
- Every month: count a rotation of C items by location, including reserve stock and slow movers.
- After unusual events: count affected items after damages, returns processing, or major transfers.
The benefit of cycle counting is that errors are found earlier, while the quantity is still small and the cause is easier to trace. The limitation is that cycle counting does not eliminate the need for a full physical stocktake from time to time, especially when the warehouse has many locations or frequent movements. It also requires discipline: if counts are done casually, the results will not be trustworthy.
Cross-location stock between central Tbilisi and Batumi
Cross-location inventory creates special timing problems. Suppose central Tbilisi sends 20 units of a product to the Batumi branch. If the transfer leaves Tbilisi physically on Tuesday but is entered only on Wednesday, the Tbilisi count on Tuesday evening may show a shortage, while Batumi may also look short because the incoming transfer has not yet been received in the system. For a short period, both locations can appear wrong even though the stock is simply in transit.
Another common case is double counting. If the transfer is counted physically in Batumi before it is officially received, and the same cartons are still listed as in transit, the team may believe the stock exists in two places. The correct interpretation depends on the movement timing and the agreed transfer process.
To avoid confusion, the warehouse should confirm whether the count is being done at the sending location, receiving location, or during transit. If the inventory is physically on the truck, it should not be counted as available stock in either branch unless the process specifically treats in-transit goods differently. The key is to align the physical status with the system status.
Working through a full reconciliation example
Assume a Tbilisi warehouse has the following items during stocktaking:
- Item 1: book 125, physical 117
- Item 2: book 40, physical 46
- Item 3: book 18, physical 18
The team first computes the variance for each item: Item 1 is -8, Item 2 is +6, and Item 3 is 0. They then investigate Item 1 because the shortage may affect a customer order. The inquiry reveals 5 units were issued to dispatch but not posted, and 3 units were moved to a repair shelf without a movement entry. Item 2 is traced to a duplicate receipt of 6 units from a supplier document. Item 3 needs no action.
This example shows why a count is only the starting point. The final approved adjustment should follow the investigation, not replace it. The team should keep a record of the physical count, any recount, the reason found, and the approved adjustment. That record becomes the basis for future improvement.
Practical stocktaking checklist
- Confirm the warehouse and exact location before counting.
- Separate sellable, quarantine, repair, returned, and scrap stock.
- Check for active transfers, receipts, issues, and shipments.
- Use a consistent sign convention for all variances.
- Prioritize high-value and customer-critical discrepancies first.
- Recount when the first count and records do not align.
- Investigate timing if online sales or reservations are active.
- Approve adjustments only after the cause is understood.
Done well, stocktaking is not just about finding errors. It is a structured way to protect service levels, improve data quality, and keep warehouse movements aligned with real inventory on the ground.
Recount Thresholds, Supervisor Approval, and Immutable Audit Trails
A good stocktaking process does not depend on “getting the right number the first time.” It depends on knowing when a number is uncertain, when a recount is required, who can approve the next step, and how every action is preserved for review. In a warehouse, the moment a physical count disagrees with the system, the process must slow down, not speed up. That is why recount thresholds, supervisor signoff, reason codes, and a protected audit trail are central controls.
Rigi24’s inventory module is confirmed to work with warehouses, locations, stock counts, and movements. Those are the core concepts you should rely on when planning your workflow. Any use of recount thresholds, approval gates, or adjustment controls should be treated as operating procedures designed around the system, unless your own verified Rigi24 configuration explicitly supports them. Do not assume that the software itself enforces every control you need.
A practical threshold can be very simple. For example, if a location count shows 100 units in stock and the physical count returns 97, the variance is 3 units or 3%. A team may decide that any variance over 2 units, any line with negative stock, or any line with a value above a fixed amount must be recounted by a second person. The threshold is not a financial rule; it is a control rule that says, “This result needs another look before anyone changes the books.”
Supervisor approval should come before any adjustment is posted. The supervisor does not merely “agree with the number.” The supervisor reviews the evidence: the first count sheet, the recount result, the reason code, and any movement history around the location. In a Georgian warehouse handling bottled products, for example, a first count of 240 and a recount of 236 might be explained by four damaged bottles already moved to a quarantine area. Without that explanation, the variance remains unresolved and should not be closed casually.
An immutable audit trail means that the record of what happened must not be overwritten by later edits. Best practice is to keep the original count, the recount, the reviewer’s notes, the chosen reason code, the approver’s identity, and the timestamp for each action. If a correction is needed, the correction should add to the record rather than erase the previous result. That way, an auditor can see the sequence: first count, discrepancy detection, recount, analysis, approval, and posted adjustment.
From First Count to Recount: A Controlled Reconciliation Flow
Reconciliation is the disciplined path from physical observation to approved stock correction. It should never be a shortcut from “system says 50, warehouse says 44” to “change it to 44.” The correct sequence is: count, validate, recount if needed, analyze, approve, and only then post an adjustment through the proper inventory process.
Start with the first count. The counter records the quantity by warehouse location, using the unit of measure that actually matches the stored stock. If the item is tracked in cartons but the team counts pieces, the result must be converted before comparison. A crate might contain 12 bottles, so a physical count of 5 crates and 6 bottles equals 66 bottles, not 5.6 or 56. Small unit mistakes create large reconciliations later.
Next comes the discrepancy review. Suppose the system shows 180 units in Location A and the physical count shows 174. The difference is 6 units. If your threshold requires recount above 2 units, a different person repeats the count. The recount should be independent, meaning the second counter should not simply confirm the first number by memory. They should physically inspect the same location, same packaging, and same labeling conditions.
If the recount returns 175, the discrepancy is now 5 units against system stock. At this stage the team should analyze why two counts produced two results. Common explanations include mixed packaging, one pallet in quarantine, a return that was received but not yet posted, or a double-scanned movement. The analysis step is not optional. It is what separates controlled inventory management from guesswork.
Only after analysis should an adjustment be considered. The approval record should show the quantity difference and the reason code. For example, “Packaging conversion error,” “Damaged goods in quarantine,” “Unposted receipt,” or “Location misplacement.” The reason code matters because it tells finance and operations why the adjustment happened and whether follow-up action is needed. A missing reason code is a weak control because it hides repeat patterns.
Do not automate write-offs without control. Even if a cycle count suggests a shortage, the process should not instantly expense the difference. There should be human review, especially for recurring differences, high-value items, or unusual patterns by location. Automated posting can make the books look tidy while hiding real operational problems such as receiving errors, mis-slotted stock, or unrecorded movements.
Operational Checklist: The Day Before Stock Count
The day before counting is when most failures can still be prevented. A clean count starts with disciplined preparation, not with a faster counter on the morning of the event.
- Confirm the count scope: which warehouses, which locations, and which item groups are included.
- Review open movements, pending receipts, pending dispatches, and any goods in temporary staging areas.
- Decide whether sales will be frozen or whether a controlled movement window will be used for critical operations only.
- Verify that all location labels are readable and that mixed or damaged labels are corrected in advance.
- Segregate quarantined, damaged, consignment, and returns stock so it is not counted with sellable inventory.
- Check unit-of-measure conversions for items stored in boxes, bundles, liters, or kilograms.
- Print or prepare the count list, ensuring each line is tied to the correct warehouse and location.
- Assign roles: counter, verifier, supervisor, finance contact, and operations manager.
A sales freeze is the cleanest option because it removes movement noise. If business conditions do not allow a full freeze, then a controlled movement window should be announced clearly. That means only specific movements are allowed, such as emergency shipping for priority customers, and every exception must be logged immediately. The goal is not zero activity; the goal is controlled activity.
For a warehouse in Tbilisi preparing for a Friday count, the team may decide that from 18:00 onward no normal picking will occur, and all receiving stops by 19:00. If a pallet must still be received after that time, it is moved to a clearly marked holding location and counted separately or excluded by written rule. The most important point is consistency. If one team member thinks “late receipts are okay” and another thinks “nothing moves after 18:00,” the count will lose credibility.
Operational Checklist: Count Day and Immediate Control Steps
On count day, the warehouse should work with a single objective: collect accurate physical evidence without mixing it with normal operations. Speed is useful only after accuracy is protected.
- Open the count session or count paperwork for the selected warehouse and locations.
- Confirm that all counters understand the counting unit and the labeling rules.
- Walk the warehouse to verify that quarantined, damaged, and returns areas are physically separated.
- Count by location, not by memory, and record exactly what is seen.
- Do not default missing quantities to zero unless the process rule explicitly says a verified empty location may be recorded as zero.
- Escalate uncertain lines for recount rather than guessing.
- Record any abnormal condition: broken pallet, unreadable label, mixed SKU, or unposted movement.
- Keep the verifier independent from the counter where possible.
Double scanning is a common failure case in barcode-driven operations. If the same pallet is scanned twice, the system record may appear correct while the physical count is too low, or vice versa. The team should therefore verify sequence and location, not just barcode presence. If a barcode is damaged and manually keyed, that manual entry should be clearly marked for review.
Another recurring mistake is counting a converted packaging unit incorrectly. A location may hold 8 cases, each with 24 units, plus 6 loose units. The correct total is 198 units. If the counter writes “8.6 cases” or counts 86 units, the variance will appear larger than it is. Unit conversion must be standardized before count day, especially for mixed-pack items.
After Count: Discrepancy Analysis, Approval, and Accounting Handoff
After the physical count is complete, the warehouse still is not done. The next phase is a disciplined review of every variance line. This is where operations, finance, and supervision must work together.
Begin by separating small explainable variances from larger unresolved ones. A shortage of 1 unit on a slow-moving item may be a labeling slip. A shortage of 120 units on a high-value item demands deeper analysis. The review should compare the first count, recount if performed, recent receipts, dispatches, returns, and any movements between locations.
Take this example: System stock is 500 units in Location B. First count is 487. Recount is 489. The warehouse finds that 10 units were moved to staging for dispatch but only 8 were posted before the count. That leaves a true variance of 2 units. The approved adjustment should reflect the actual condition after movement review, not the first emotional reaction to the shortage number.
The accounting handoff should include the item code, warehouse, location, quantity difference, value of adjustment, reason code, approver, date, and any supporting notes. Finance needs this information to post the entry correctly, while operations needs it to fix the process. Adjustment records should never be treated as paperwork only; they are also operational evidence.
If the discrepancy remains unexplained, that should be stated plainly in the incident report. Do not invent an explanation merely to close the file. An unresolved shortage is still useful information because it tells management where control failures may exist, such as poor location discipline, weak receiving checks, or uncontrolled temporary storage.
Rigi24 Inventory Scope: What Is Confirmed and What Is Process Design
Rigi24 inventory is confirmed to include warehouse, location, stock count, and movement concepts. That means the system can represent where stock is held, how it moves, and how counts are recorded against those structures. Use those core capabilities as the foundation of your process design.
However, certain control features should not be assumed unless your own verified setup shows them. Do not presume that Rigi24 automatically schedules counts, locks movements during counting, enforces blind counts, applies tolerance approvals, detects anomalies with AI, or posts financial adjustments without review. Those may be good process ideas, but they are not confirmed here as product features.
The practical approach is to design the operating procedure around the confirmed objects: warehouse, location, stock count, and movement. For example, if you want a controlled movement window, that is an operational rule your team follows. If you want a recount threshold, that is a policy your supervisor applies. If you want adjustment approval, that is a control workflow around the system, not a feature you should assume exists by default.
This distinction matters because a process can fail when teams believe software will handle a control that actually depends on people. If a movement is made during the count and nobody documents it, the system data may no longer match the physical state. Good inventory governance always separates what the software records from what the team must manage manually.
Inventory Variance KPIs for Management Review
Variance KPIs help management see whether stocktaking is improving. They should be simple, repeatable, and tied to the actual count process.
| KPI | What it measures | Simple example |
|---|---|---|
| Count accuracy | How close physical counts are to expected stock | 98 matched lines out of 100 counted lines |
| Gross adjustments | Total absolute quantity adjusted, ignoring sign | Shortages 12 plus overages 5 equals gross 17 |
| Net adjustments | Shortages minus overages | 12 shortage and 5 overage equals net shortage 7 |
| Adjustment value | Financial value of the approved correction | 7 units at 14 GEL each equals 98 GEL |
| Repeat discrepancies | Items or locations that fail again in later counts | The same SKU shows shortage twice in one month |
Do not attach invented benchmark targets if your business has not defined them. A KPI becomes useful when it compares current performance to your own historical performance, not to a random external number. For example, if one warehouse in Batumi reduces repeat discrepancies from 18 lines in one month to 9 the next month, that trend is meaningful even without an industry benchmark.
Gross and net adjustments tell different stories. Net may look small while gross is large, which can mean many opposite errors are hiding inside the warehouse. If shortages total 40 units and overages total 39 units, the net is only 1 unit, but the gross is 79 units. That is not a healthy count result; it means the operation is unstable.
Training Plan for a Five-Person Warehouse Team
A small team can run a strong count if each role is trained clearly. The team should know what to do before, during, and after the count.
- Counter: Physically counts stock by location, records what is seen, and escalates unclear cases.
- Verifier: Independently rechecks selected lines, especially variances and high-value items.
- Supervisor: Reviews discrepancies, applies reason codes, and approves or rejects adjustments.
- Finance: Reviews valuation impact and confirms accounting treatment for approved adjustments.
- Operations manager: Owns the process, resolves cross-functional issues, and drives corrective actions.
Training should include a short practical drill. First, show the team how to count a mixed location. Second, show how to handle a missing label. Third, show how to document a discrepancy without guessing. Fourth, show how to escalate a questionable count to a recount. A team that has practiced these cases will make fewer emotional decisions on the day of the count.
The counter should be trained not to “fix” the system number. The verifier should be trained not to rubber-stamp the first count. The supervisor should be trained to ask, “What evidence supports this adjustment?” Finance should be trained to look for valuation impact and repeat patterns, not just entries. The operations manager should be trained to ask whether the root cause is process-related, location-related, or movement-related.
Incident Report and Corrective Action Plan for Recurring Shortages
When shortages recur without a clear explanation, the response should be structured and neutral. The report should not accuse a person of theft unless there is verified evidence and an authorized investigation process. In normal operations, the focus is on process failure, control weakness, and evidence-based correction.
An incident report should contain the following elements:
- Warehouse, location, item code, and item description
- Count date, first count, recount, and final approved quantity
- Shortage or overage amount and value
- Relevant movement history, receipts, returns, and quarantined stock notes
- Observed conditions such as damaged packaging, mixed pallets, or unreadable labels
- Reason code if one is approved, or “unexplained” if no cause is confirmed
- Immediate containment action and planned corrective action
A root-cause corrective action plan should move from symptom to cause to fix. If repeated shortages appear in one location, ask whether the issue is receiving error, picking error, mis-slotting, conversion error, or uncontrolled temporary movement. Then assign a corrective action with an owner and due date. For example: relabel the location, retrain the receiving clerk, require second-person verification for returns, or separate quarantine stock with a visible marker.
Consider a repeated shortage of 3 units on the same SKU across three monthly counts. The cause might not be loss at all; it may be that the item is received in packs of 10 but stored in loose units after partial picking, and the system is not updated consistently. The corrective action is therefore process discipline, not blame. That is the kind of reasoning that improves inventory health over time.
A strong stocktaking program is built on verification, not assumption. When the team uses threshold rules, independent recounts, reason codes, and a clean audit trail, it can resolve shortages and overages in a way that supports both operations and finance. Rigi24’s confirmed warehouse, location, stock count, and movement concepts provide the data structure; the warehouse team provides the discipline that makes the numbers reliable.

Frequently asked questions
How does cycle counting differ from a full stocktake?
Cycle counts verify selected items repeatedly; a full stocktake checks a broader inventory population at a chosen time.
What is an inventory variance?
The difference between a verified physical count and the book balance at the same reference time.
When should a count be repeated?
An independent recount can confirm or challenge an unexpected first result.
How should zero stock be recorded?
Record a confirmed zero explicitly; do not confuse it with an item not counted.
Does barcode scanning automatically adjust stock?
No. Scanning identifies an item; a controlled inventory transaction changes stock.
How should goods in transit be handled?
Record the transfer cut-off and actual custody so quantities are not doubled.
Should damaged items be counted?
Yes, but separately by condition and sellability.
Who should approve inventory adjustments?
An authorized supervisor should sign off on documented adjustments.


