A distribution business can report a healthy gross margin on Friday and discover on Monday that a high-value product line is missing, mislocated or already allocated twice. The immediate issue looks like inventory. The commercial consequence is wider: rushed purchasing, delayed fulfilment, unreliable financial reporting, dissatisfied customers and management decisions made on false assumptions.
This inventory accuracy case study examines how a mid-market Australian wholesaler improved control over stock held across a central warehouse, satellite locations and service vehicles. The scenario is illustrative, but the operating issues are familiar to organisations managing growth through a mix of spreadsheets, legacy processes and disconnected systems.
Inventory accuracy case study: where margin began to leak
The business supplied maintenance and project-critical parts to industrial customers. Its catalogue had expanded quickly, with around 18,000 active stock keeping units, variable supplier lead times and customer service commitments that often required same-day dispatch.
On paper, inventory records appeared acceptable. The enterprise resource planning system showed an accuracy rate of 96 per cent when measured against periodic stocktakes. Yet warehouse leaders did not trust available-to-promise figures, customer service staff regularly called the warehouse to verify stock, and finance was processing increasingly material stock adjustments at month end.
The gap between the reported figure and operational reality mattered. A broad stocktake percentage can conceal where errors sit and what they cost. A $5 fast-moving fitting recorded incorrectly may be inconvenient. A small number of misrecorded specialist components, each worth thousands of dollars and needed for a shutdown project, can damage revenue, margin and customer confidence at once.
The executive team set a clear objective: improve inventory accuracy not as a warehouse housekeeping exercise, but as a control and performance program. They wanted reliable stock visibility, disciplined replenishment and financial reporting that reflected the physical business.
The symptoms were connected, not isolated
Initial investigation found that errors were entering the process at several points. Goods were sometimes received into a staging area before being posted in the system. Transfers between locations were recorded late or not at all. Pickers made practical substitutions to meet urgent orders, but the system transaction followed later, if it occurred. Returned goods were placed in convenient locations while their condition and ownership were assessed.
The warehouse team was under pressure to keep orders moving. Finance was focused on reconciling valuation adjustments. Customer service was managing the consequences of unavailable stock. Each group was acting rationally within its immediate responsibilities, but no one owned the end-to-end accuracy of the inventory record.
This is a common control failure. Inventory accuracy is not created by a stocktake alone. It is created every time goods are received, moved, picked, returned, counted, written off or reclassified. If the physical movement and the system transaction are separated, uncertainty accumulates.
Diagnosing the causes before selecting technology
The program began with process observation, transaction analysis and interviews across operations, finance, procurement and customer service. The aim was to establish a usable baseline, rather than assume the ERP system was the problem.
The analysis separated inventory variances into meaningful categories: receiving timing, put-away errors, location transfer failures, picking discrepancies, returns, master data issues and obsolete or damaged stock. This made patterns visible. Most adjustments were not random. They clustered around a small group of locations, high-velocity product families and periods of peak dispatch activity.
The business also reviewed its measures. Counting accuracy by SKU was retained, but it was supplemented with inventory record accuracy by value, stock adjustment value as a proportion of sales, order lines affected by stock exceptions, aged unposted transactions and cycle count completion. These measures connected warehouse discipline with service and financial outcomes.
A key finding concerned master data. Some products had inconsistent units of measure between purchasing, warehousing and sales. A carton received as one unit could be sold as individual items, with conversion rules understood by experienced staff but not consistently represented in the system. Better scanning would not resolve that issue on its own.
The diagnosis led to a practical principle: technology should enforce a sound process, not automate an unclear one. The business decided to simplify workflows before extending mobile capability and improving its Microsoft Dynamics 365 Business Central configuration.
Building control into everyday warehouse work
The redesigned process established clear transaction points. Goods could not move from receiving to available stock until quantity, condition, purchase order and location were confirmed. Exceptions were given a visible status rather than being managed through informal notes or memory.
Put-away was directed to defined bin locations. Inter-location transfers required confirmation at both the dispatching and receiving point for higher-risk movements. Returns were routed to a quarantine location until inspection determined whether stock could be returned to saleable inventory, sent to a supplier, repaired or written off.
For urgent customer orders, the business created an expedited fulfilment process rather than allowing normal controls to be bypassed. Staff could act quickly, but every substitution, partial supply and stock movement had a defined approval and system record. This was an important trade-off. A process that is too rigid can encourage workarounds; a process that allows unlimited exceptions loses control. The right design makes the compliant path the fastest practical path.
Mobile scanning was introduced progressively, beginning with receiving, put-away, picking and cycle counts. Business Central was configured to improve bin-level visibility, item tracking and exception reporting. The technology reduced manual entry, but the larger benefit came from making transactions occur at the point of physical work.
Role design also changed. Warehouse supervisors owned daily exception review. Finance monitored adjustment trends and valuation impacts. Procurement used more reliable demand and available-stock data to reduce unnecessary expedites. A cross-functional inventory control forum reviewed root causes monthly, with actions assigned to the person best placed to change the process.
Why cycle counting changed the quality of decisions
The annual stocktake had previously been treated as the main assurance activity. It was costly, disruptive and too late to prevent recurring errors. The new approach introduced risk-based cycle counting.
High-value, high-velocity and project-critical items were counted more often than slow-moving, low-value stock. Locations with repeat discrepancies received additional attention. Counts were blind where appropriate, so staff were not influenced by the quantity expected in the system.
Just as importantly, every significant variance triggered investigation. The purpose was not to assign blame for a single mispick. It was to identify whether the cause was training, bin labelling, product identification, system configuration, workload planning or an approval weakness.
Within six months, the business improved its value-weighted inventory record accuracy from 91 per cent to 98.7 per cent. More tellingly, stock adjustment value fell by 43 per cent, emergency purchasing reduced, and customer service calls to verify stock availability became the exception rather than routine practice. Month-end reconciliation was faster because finance was resolving fewer unexplained movements.
The numbers should not be treated as a universal benchmark. A business with regulated traceability obligations, serialised assets or multiple third-party logistics providers may need tighter controls and a more substantial integration program. Conversely, a smaller operation may achieve meaningful improvement through location discipline, better master data and weekly cycle counts before investing in advanced warehouse functionality.
What executives should test in their own operation
Leaders should be cautious when an inventory accuracy percentage appears reassuring but operating teams still compensate for unreliable data. Ask whether the metric is weighted by value and business risk, whether variances are classified by cause, and whether the system records the physical movement when it happens.
It is also worth testing the governance around exceptions. Can staff identify stock in receiving, quarantine, consignment, service vehicles and third-party locations? Is there a defined owner for unresolved transactions? Do finance and operations review the same source of truth, or are reconciliation spreadsheets carrying critical control information outside the ERP environment?
For organisations using Microsoft Business Central, the opportunity is not limited to replacing spreadsheets. Well-designed item, location, bin, item tracking and approval controls can provide a stronger operating foundation. Reporting and analytics can then focus management attention on the exceptions that affect margin, customer commitments and working capital.
i3 Australia approaches these programs as connected business improvement work. Inventory control needs process discipline, clear accountability, usable data and technology configured around the way the organisation should operate. Addressing only one of those elements may improve a metric temporarily without reducing the underlying uncertainty.
Reliable inventory is a practical source of commercial confidence. When leaders can trust what is on hand, where it is held and what it is worth, they can commit to customers with greater certainty, protect margin without unnecessary buffer stock, and make decisions before a variance becomes an expensive surprise.