The warehouse team's objection to inventory software is almost always the same: the system count never matches the physical count. They are not wrong. In the first few months of any inventory implementation, the system count and the physical count will differ, sometimes significantly. Opening balances are estimated. Goods in transit are double-counted. Old stock that has been written off mentally but never formally is still on the books. These are real discrepancies, and they take real time to resolve.
What makes this objection worth examining is that in businesses running without any perpetual inventory system, the gap between system and physical has been present all along - it was simply invisible. The register showed what was entered. Nobody knew whether what was entered matched what was on the shelf. The first stock count after an ERP implementation does not create discrepancies; it makes pre-existing ones visible for the first time.
Understanding this distinction does not make the reconciliation work easier. But it changes the question being asked. The question is not "why does the system not match?" It is "what does the gap tell us, and how much of it would exist in any environment versus how much is specific to this one?"
What Warehouse Resistance Looks Like
Warehouse resistance to inventory software takes several forms. The most common is persistent reporting of system-to-physical discrepancies as a system problem rather than a reconciliation task. If the system shows 500 units and the physical count finds 487, the warehouse team's framing matters: "the system is wrong" versus "we have a 13-unit discrepancy to investigate." Both statements describe the same situation. Only one treats it as something to resolve.
The second form is selective data entry. Goods arriving are entered into the system. Goods leaving - whether for internal use, minor write-offs, or small quantities that "just went" - are entered inconsistently. The result is a system that perpetually reads higher than physical, with the gap growing gradually as unrecorded outflows accumulate. When the stock count eventually happens, the adjustment entry is large enough that it looks like a system error rather than the accumulated result of incomplete recording.
Third is resistance to unannounced physical counts. A warehouse team comfortable with the perpetual inventory system accepts a stock count as a routine calibration exercise. A warehouse team uncomfortable with the gap between system and physical tends to have strong views about when counts should happen and needs notice to "prepare." Preparation for an inventory count means reconciliation and recording of pending entries - which is legitimate work. When it means other things, the count itself is the control.
The pattern to watch is not whether discrepancies exist - they always do in early implementation. It is whether the warehouse team treats them as reconciliation tasks to resolve or as evidence that the system does not work. The response to discrepancy reveals more than the discrepancy itself.
What Manual Stock Systems Leave Unrecorded
In a manual warehouse register, every line represents what the person making the entry chose to record. Inbound stock from a supplier is almost always recorded - it comes with a delivery note and someone is paying for it. Outbound stock issued to production, transferred between locations, returned to a supplier, or removed for sampling is recorded with more variability. Small quantities that leave the warehouse without a corresponding document often go unrecorded entirely, particularly in fast-moving operations where the recording step feels disproportionate to the transaction size.
Over time, the manual register and the actual physical stock diverge. The register reads higher. The gap grows. At the annual physical count, an adjustment entry is raised that brings the system into line with reality. That adjustment entry, in a manual environment, carries no detail about how the gap accumulated - only its magnitude. Nobody goes back to understand whether the gap was normal operational variance, damage and spoilage, or something else.
Nepal businesses with multiple locations - a Kathmandu head office, a Biratnagar godown, a Butwal branch - often have genuinely independent inventory records at each location. The person managing each godown effectively controls the stock record for that location, with no real-time visibility from the head office. Stock transfers between locations are recorded when the head office reconciles, not when the transfer happens. This creates windows between a transfer being physically made and being reflected in any system - windows where the stock exists on neither location's record simultaneously, or on both.
The most significant gap in manual inventory environments is not the small daily variance from unrecorded minor issues. It is the structural gap that opens when stock adjustments can be posted without authorization or documented reason. When a warehouse manager can post a stock reduction entry - "adjustment," "damage," "expired" - without anyone else reviewing it, the reasons for stock movements are visible only to the person who made them. In a system with an authorization requirement for adjustments above a threshold, every stock reduction above that threshold requires a second person to review and approve the entry, and the reason is recorded permanently.
The critical control is not perpetual inventory tracking alone - it is stock adjustment authorization. Any reduction in stock quantity above a defined threshold should require a documented reason and a second authorization. Without this, perpetual inventory creates visibility but not accountability for how the inventory changes.
What Digital Inventory Makes Traceable
A perpetual inventory system with full transaction logging records every stock movement - inbound from suppliers, outbound to customers, inter-location transfers, adjustments, returns, and physical count corrections. Each movement carries a user attribution, a timestamp, and a reference to the source document. The history of any item can be traced from the moment it entered the warehouse to the moment it left.
For management, the most practically useful output of this traceability is not the item-level history - it is the pattern analysis. Which stock categories show the highest adjustment frequency? Which location has the most GRN-to-system discrepancies? Which items consistently read higher on the system than on physical count? These patterns, visible in aggregate from a system report, are not visible from a manual register without a significant manual analysis exercise.
Stock adjustment authorization should be configured before go-live, not added later. Specifically: set a threshold above which any stock reduction requires approval from the finance manager or operations director, and require a reason code on every adjustment regardless of amount. The reason codes - "damaged," "expired," "customer return," "internal use," "count variance" - should be defined in advance and selected from a list rather than typed freely. Free-text reasons are not analysable; coded reasons are.
Unannounced physical counts are the calibration mechanism that keeps the system honest. A business that counts stock once a year at the financial year-end (Ashadh) is measuring a point-in-time gap that has accumulated over twelve months without any intermediate check. A business that conducts spot counts of specific categories quarterly - rotating through categories so everything is counted at least twice a year - detects variances when they are small and traceable rather than when they have grown to a magnitude that requires a large unexplained adjustment.
Unannounced spot counts are more effective than annual physical counts for inventory control. They catch variances when they are small and still traceable to specific transactions, rather than at year-end when the accumulated gap requires a write-off with no understanding of where it came from.
How to Establish Opening Balances and Move Forward
The most technically challenging part of an inventory implementation is establishing correct opening balances. For businesses that have been running manual records for years, the physical count at go-live will find discrepancies between the manual record and what is actually on the shelf. The temptation is to use the manual record as the opening balance to avoid a large adjustment entry. This imports the existing gap into the new system and the cleaning-up process begins from a compromised baseline.
The correct approach is to use a physical count as the opening balance, regardless of the adjustment it generates. The variance between the physical count and the old register is documented as a transition adjustment with a clear narration. From that point, the system starts from a known position. Every subsequent variance is measured from a verified baseline, which makes it meaningful rather than ambiguous.
Use the physical count at go-live as the opening balance, not the manual register. Starting from a verified physical position gives the system a meaningful baseline. Starting from an unverified manual register imports accumulated inaccuracies and makes every subsequent discrepancy uninterpretable.
Twelve months of unrecorded minor movements accumulate into a material adjustment with no traceable cause
Rotating category counts detect discrepancies when they are small and still traceable to recent transactions
Any warehouse staff member can reduce stock quantities without documentation or second review
Reductions above a threshold route to finance manager with a mandatory reason code from a defined list
Head office has no real-time stock position for branch locations - only what was last reported
Every location's stock position visible from the head office dashboard without calling the godown
Stock exists on neither location's record during transit - or on both simultaneously
In-transit stock tracked as a separate status - no double-counting or invisible period
Accumulated inaccuracies from years of manual recording imported into the new system
System starts from a known position - every subsequent variance is meaningful and measurable
Frequently Asked Questions
In the first three months, expect variances as the team builds entry discipline and the opening balance is refined. A system-to-physical variance of 1-2 percent of total stock value in this period is within normal range for a business establishing its first perpetual inventory. By month four, variances above 0.5 percent warrant investigation. By month six, a well-run inventory system should reconcile within 0.2-0.3 percent of total value at any spot count. Persistent high variances after the first quarter are an operational signal, not a system signal.
Record what actually arrived on the GRN - not what was ordered. A purchase order for 100 bags that delivers 93 intact and 7 damaged produces a GRN for 93 usable units and a damage note for 7. The system then knows the actual stock received. The payment is authorised against the GRN quantity, not the PO quantity. A debit note or return document handles the shortfall commercially with the supplier. Recording what arrived rather than what was ordered is the foundation of accurate inventory - and it protects the business in the supplier dispute that usually follows.
Spot counts work best when you count one category at a time rather than the full warehouse. Select a product category representing 10-15 percent of total SKUs, freeze movements for that category for two hours, count and compare to system. The operation continues on all other products. This takes half a day per category rather than a full warehouse shutdown. Rotating through categories every 6-8 weeks means everything is counted at least twice a year without a single disruptive full count. The unannounced element matters - give no advance notice to the warehouse team beyond "we are counting category X today."
Perpetual Inventory With Full Movement Audit and Multi-Location Visibility
MISAC's inventory module tracks every stock movement with a full audit trail - user, timestamp, source document reference, and reason code on every adjustment. Every GRN auto-posts the corresponding accounting entry, so the inventory record and the financial record are always in sync without a separate reconciliation step. Multi-location businesses see real-time stock positions across all godowns from a single dashboard - no need to call each location and aggregate manually.
Stock adjustment authorization in MISAC is configurable by category and amount. Damage write-offs above a defined threshold route to the finance manager for review before the system stock is reduced. Reason codes are selected from a managed list rather than entered as free text, making adjustment pattern analysis straightforward. The physical stock count module supports partial counts - count one category while the rest of the warehouse continues operating - and automatically generates the variance entry for management review before it is posted.
Businesses that implement MISAC inventory as a standalone starting module - without HR, payroll, or full ERP at the same time - find the adoption curve manageable precisely because the warehouse team is learning one system for one purpose. MISAC Intelligence Pvt. Ltd.'s modular architecture means inventory can go live first, prove its value, and create the operational confidence that makes the next module's adoption straightforward.
Ready to See MISAC in Action?
Talk to us about establishing opening balances and configuring inventory controls for your Nepal warehouse operation.