Inventory Discrepancies in Multi-Warehouse Operations

A Practical Framework for Australian Distributors

If you're running more than one warehouse, you've accepted that your stock count will never perfectly match your system count. The question isn't whether discrepancies exist — it's whether they're small enough to ignore or large enough to be quietly costing you margin, service level, and staff hours every month.

Why Multi-Warehouse Operations Are Structurally More Exposed

A single-site business can get away with informal controls because visibility is high — staff can physically walk the floor and reconcile by eye. Multi-warehouse operations don't have that luxury. Every additional site multiplies the number of handoffs, and every handoff is a point where stock and system can diverge:

• Inter-warehouse transfers create a window where stock is in transit but recorded as either "still at origin" or "already at destination," depending on when the transfer is processed relative to the physical movement.
• Each site may have slightly different practices for receiving, put-away, and cycle counting, even under one ERP.
• Local overrides — a warehouse supervisor manually adjusting a quantity to "fix" a picking problem without root-causing it, won't get visibility at head office until the variance shows up in a stocktake.
• Consolidated reporting can mask site-level problems. Aggregate inventory accuracy might look acceptable while one warehouse is carrying a serious control gap that's being offset by another site's surplus.

This is why "we'll tighten up processes" rarely works as a standalone fix. The processes usually aren't the problem.

The lack of a structured way to see where control breaks down across sites is.

The Root Causes Worth Diagnosing

Most recurring stock discrepancies trace back to one of five categories. It's worth working through these deliberately rather than assuming you already know which one applies — in our experience running gap analyses for distribution clients, businesses are frequently surprised by which cause is actually dominant.

1. Transaction timing gaps. Physical movement and system movement don't happen simultaneously. Goods are received on the dock before they're booked in. Picks are completed before the transaction posts. In a single warehouse this washes out within a day. Across multiple warehouses, especially with inter-branch transfers, the timing gap compounds. And if your ERP doesn't have a proper in-transit state for stock, you'll see phantom variances that self-correct a day later and confuse everyone in the meantime.

2. Manual entry and workarounds. Any point where a person types a quantity instead of scanning it is a point of risk. This is common in returns processing, damaged stock write-offs, and "quick fix" adjustments made to unblock an order. These workarounds are rarely malicious. They're usually a symptom of a process gap (no clear procedure for handling a damaged carton, for example) rather than carelessness. But they erode inventory accuracy steadily because they bypass the controls that would otherwise flag the discrepancy at the time it occurs.

3. Cycle counting cadence and method. Many mid-market distributors still run an annual full stocktake and call it a control. That's a lagging indicator, not a control — by the time you find the variance, the transactions that caused it are months old and unrecoverable. Businesses with strong inventory accuracy run continuous cycle counting, typically ABC-weighted (your highest-value or highest-velocity SKUs counted more frequently than slow movers), so discrepancies are caught and root-caused within days, not once a year.

4. WMS/ERP integration gaps. If your warehouse execution — scanning, picking, put-away — sits in a system that isn't tightly integrated with your ERP, you're relying on batch syncs or manual reconciliation between two sources of truth. Any latency or failure in that sync shows up as a discrepancy that has nothing to do with the physical stock at all. This is one of the most common issues we see in Acumatica and Business Central implementations where a legacy WMS or a spreadsheet-based process has been bolted on rather than properly integrated.

5. Negative inventory and system tolerance settings. Some ERP configurations allow negative inventory as a default to avoid blocking sales orders. That's a legitimate business decision in some contexts, but it also means the system will let you sell stock that isn't physically there, and the resulting variance gets buried until the next count. If this setting exists in your environment, it should be a deliberate, reviewed decision. Not a default nobody's revisited since go-live.

A Diagnostic Framework for Locating the Control Gap

Rather than auditing everything at once, work through this sequence per warehouse:

Step 1: Isolate by transaction type, not by SKU. Pull your variance data and categorise it by the transaction type most recently associated with each affected SKU: transfer, sale, purchase receipt, return, adjustment. Patterns by transaction type point directly at the process to fix. Patterns by SKU alone usually just tell you what's fast-moving.

Step 2: Compare variance timing against your cycle count schedule. If discrepancies cluster around specific counting windows, the issue may be cadence or method rather than the underlying transactions. If they're spread evenly, the issue is more likely transactional.

Step 3: Check for site-specific process divergence. Run the same variance analysis per warehouse. If one site is consistently worse, you have a training or local-process issue, not a systemic one, and it's a much cheaper fix.

Step 4: Audit manual adjustment logs. Every ERP worth using logs who made manual stock adjustments and why. If that reason field is routinely blank or generic ("stock count correction"), you don't have visibility into your own control gaps. What you have is a blind spot with a paper trail attached.

Step 5: Map your in-transit and inter-warehouse transfer process end to end. This is the single most common structural gap in multi-site distribution. If stock in transit isn't a distinct, trackable state in your system, you will never fully close this variance category. You'll just keep managing around it.

Reducing Manual Reconciliation Going Forward

Once the diagnosis is done, the fix is rarely "hire more people to count more often." The businesses that get inventory accuracy under control durably do three things:

• Push data capture to the point of transaction, via barcode or RFID scanning at receipt, pick, and put-away, so the system record and the physical movement happen in the same action rather than being reconciled after the fact.
• Build exception-based reporting rather than relying on periodic full counts. A well-configured ERP can flag negative stock, ageing in-transit transfers, and adjustment frequency by user or site automatically — surfacing problems in days rather than at year-end stocktake.
• Standardise the transfer and adjustment workflow across every warehouse, so a discrepancy at one site is diagnosed the same way as at another, and head office has one consistent data set to work from rather than five slightly different ones.

Where This Usually Lands

In our experience working with Australian wholesale and distribution businesses on both MYOB Acumatica and Microsoft Business Central (via Wiise), the businesses carrying the worst inventory accuracy problems are almost never the ones with bad staff or bad intentions. They're the ones running mature operations on ERP configurations that haven't been revisited since go-live — inventory settings, transfer workflows, and counting cadence that made sense for a single-warehouse business and were never redesigned for a multi-site one.

If your last full stocktake produced a variance you couldn't fully explain, that's the signal to run a structured diagnostic rather than tightening procedures informally and hoping it holds until next year.

Next steps: Pull your last two stocktake variance reports and categorise the discrepancies by transaction type as outlined above. If a clear pattern emerges, particularly around inter-warehouse transfers or manual adjustments, then that's your starting point for a targeted process and configuration review, rather than a full re-implementation.