How a wholesale distributor stopped over-ordering and started trusting their own data

There’s a particular kind of institutional knowledge that builds up inside businesses over time. Not the useful kind – the kind that gets written into onboarding documents and process guides. The other kind. The unwritten rules. The workarounds. The habits that nobody formally agreed to but everyone just does.

At a mid-sized wholesale distribution firm, one of those unwritten rules had been running quietly for years: never trust what the inventory system says. Always assume it’s wrong. Order more than you think you need, because the number on screen probably isn’t the number in the warehouse.

It seemed like common sense. Experience reinforced it. Yet it was costing them hundreds of thousands of pounds a year.

 

When the Workaround Becomes the Process

The purchasing manager had been with the company long enough to remember when the inventory discrepancies first started. A new warehouse management system, a rushed implementation, a period where the goods-in process wasn’t quite right. Stock levels drifted. The system said one thing, a physical count said another. People learned to compensate.

That was three years ago. The implementation issues had long since been resolved. But the habit of distrust had not.

By the time an external review was initiated, the company was operating with three separate versions of inventory data running in parallel. The legacy ERP held the official stock record. A warehouse spreadsheet, maintained by two members of the operations team, tracked day-to-day movements independently. And a manual goods-in log – originally created as a temprary fix during the NetSuite implementation – had never actually been retired.

Each of these records told a slightly different story. Nobody knew which one was right. So the purchasing team had developed an informal rule: assume the system is running about 20% light, and order accordingly.

That rule had never been written down. It had never been formally agreed. It had simply become the way things were done.

 

What Three Years of Over-Ordering Looks Like

When the full audit came, the numbers were uncomfortable.

£340,000 of excess inventory was sitting across two warehouse sites. Not slow-moving stock that had accumulated by accident – stock that had been deliberately ordered, repeatedly, because nobody trusted the figures well enough to order only what was needed.

The carrying costs alone were significant. Warehouse space costs money whether it is full or not, but filling it with inventory that exists primarily because of data anxiety is a particularly expensive way to run a business. Factor in the capital tied up in that stock – cash that wasn’t available for anything else – and the picture became harder to look at.

There were softer costs too. Supplier relationships had become strained by erratic order patterns. Volume would spike one month and drop the next, depending on which member of the purchasing team had done the calculation and how pessimistic they’d been about the system figures that week. Suppliers noticed. Pricing conversations became more difficult.

And perhaps most significantly, the business had never formally calculated what any of this was costing them. Nobody had sat down and worked out the carrying cost of the excess stock, the opportunity cost of the tied-up capital, the premium being paid on inconsistent order volumes. It had simply become the cost of doing business – invisible because it had always been there.

 

What the Assessment Actually Found

When EcobSoft came in to assess the situation, the expectation from the operations team was that the inventory system would be found to be fundamentally unreliable. That was the story the business had been telling itself for three years.

That wasn’t what the assessment found.

The first step was mapping all three data sources – the ERP, the warehouse spreadsheet, and the manual goods-in log – and identifying specifically where they diverged and why. This took time. The divergence points were not random. They followed a pattern.

Almost every discrepancy traced back to the same place: a process gap at goods-in. When deliveries arrived, the physical receipt was being logged in the manual goods-in register immediately. But the update to the ERP system was being done later – sometimes hours later, sometimes the following morning – by a different person, working from the paper log. In the gap between those two events, the warehouse spreadsheet often captured a movement that the ERP had not yet reflected.

The ERP wasn’t wrong. It was delayed. And the delay was being interpreted as inaccuracy.

When the assessment team ran a reconciliation across all three data sources and adjusted for the timing gaps, the ERP figures were accurate to within 2% of physical counts. The purchasing team had been applying a 20% buffer to a system that was 98% correct.

The problem was never the data. It was the process feeding it.

 

Building One Source of Truth

The remediation work had two components: fixing the process that had been creating the perception of unreliability, and building a system that made that perception impossible to sustain.

On the process side, the team redesigned the goods-in workflow from scratch. Every receipt now updates the system as soon as goods arrive rather than hours later. The team completely retired the manual goods-in log, which had outlived its purpose by approximately three years. They also decommissioned the warehouse spreadsheet that staff had maintained as a parallel record because they did not trust the main system. Once the new process proved reliable, the spreadsheet no longer served any purpose.

On the system side, as part of the NetSuite implementation, EcobSoft configured NetSuite as the single inventory record of truth across both warehouse sites.

This sounds straightforward. In practice, the business had to make decisions it had avoided for years about inventory categorisation, bin location tracking, receipt processing, and responsibility at each stage of the goods movement process.

Bin-level tracking was implemented for the first time. Previously, the system knew that 847 units of a product were on hand somewhere across the warehouse. Now it knew exactly which bin they were in, which site they were at, and when they had arrived. This changed the pick process, reduced errors, and gave the warehouse team a level of visibility they hadn’t previously had.

The team configured reorder points and reorder quantities inside NetSuite as part of the NetSuite implementation, using actual consumption data from the previous twelve months. Instead of relying on buyer instinct, a rough rule of thumb, or a spreadsheet managed by one person who might be absent when an order needs placing. Actual data, built into the system, triggering the right action at the right time.

The team formally retired the informal 20% buffer rule and replaced it with demand planning logic that accounted for lead times, consumption rates, and seasonality, updating automatically as new data came in.

 

The First Quarter on Clean Data

The results of the first quarter operating on clean inventory data were, depending on how you looked at them, either surprising or entirely predictable.

Purchasing volume dropped 18% against the same quarter the previous year. Not because the business was smaller – revenue was up slightly. But because the reorder points were accurate, the consumption data was reliable, and the purchasing team no longer needed to add a buffer to protect against a system they didn’t trust.

There were zero stockouts. The assumption had always been that trusting the system meant risking running out of stock. In practice, the opposite was true. Accurate reorder points, triggered automatically at the right moment, meant that replenishment happened when it needed to – not weeks early because of a vague anxiety about whether the numbers were right.

The warehouse team reported a measurable reduction in pick errors within the first eight weeks of bin-level tracking going live. When stock is in the right place, and the system knows exactly where that place is, the margin for error at fulfilment narrows significantly.

But the moment that perhaps best illustrated what had changed came about six weeks into the new system going live. The purchasing manager – the same person who had developed the habit of never trusting the inventory figure – ran a check on a fast-moving product before placing an order. The system showed 847 units on hand across both sites. She looked at the number.

And she didn’t order.

That sounds like a small thing. It wasn’t. The simple experience of trusting accurate data overrode three years of institutional habit.

 

The Actual Problem Is Rarely the System

The thing that businesses in this situation most often discover – and the thing that takes the longest to accept – is that the system is rarely the problem.

People often blame systems because they provide a convenient explanation for discrepancies that actually stem from surrounding processes. When teams enter data late, inconsistently, or through three different people maintaining parallel records, they make the system produce unreliable output. As a result, users attribute that unreliability to the system and stop trusting it. They then create workarounds, which gradually become part of the process. Those workarounds introduce new distortions, and the cycle continues.

The wholesale distributor in this story had a functional inventory system. What they didn’t have was a clean, consistent, single process feeding it. Once that was fixed, the system performed almost exactly as it was supposed to.

The £340,000 sitting in excess inventory wasn’t the result of a bad system. It was the result of a business that had stopped believing its own data – and had been quietly paying for that disbelief, month after month, without ever calculating the cost.

 

What Good Inventory Management Actually Looks Like

The goal isn’t a system that produces a number and hopes you believe it. The goal is to embed the system into the process so the team can rely on the number, act on it confidently, and achieve better outcomes than by ignoring it.

That’s the shift that matters. Not the software. Not the configuration. The experience of trusting the data and being right.

Once a purchasing team has gone through a full quarter of accurate reorder triggers, zero stockouts, and declining carrying costs, the conversation about whether to trust the system tends to become fairly short.

The number on screen is the number in the warehouse. Order accordingly.EcobSoft helps growing businesses implement and optimise NetSuite, so the numbers they’re running on are numbers worth trusting. If your team has developed its own version of the 20% buffer rule, it might be worth finding out why.