A plastics manufacturer in the West Midlands discovered last year that 38% of their warehouse held raw materials they wouldn't use for at least 90 days. That's working capital sitting on shelves, doing nothing, while the business struggled to fund a second production shift. Their spreadsheet said stock levels were fine. The spreadsheet was wrong by over 12,000 units across three material categories. Sound familiar?

Why Manufacturing Inventory Is Nothing Like Retail Inventory

Retail tracks products in and products out. Simple. Manufacturing tracks raw materials in, transformations through multiple production stages, sub-assemblies waiting for the next operation, waste generated at each step, and finished goods out. At any given moment, a manufacturer carries inventory in four or five different states simultaneously.

That complexity is why generic inventory management software built for warehousing or e-commerce falls apart in a factory. It wasn't designed for work-in-progress. It doesn't understand that 500kg of steel in the warehouse isn't the same as 500kg of steel that's been cut, bent and is halfway through powder coating.

According to a Cranfield University supply chain study, UK manufacturers hold an average of 45 days of inventory, but the optimal figure for most SMEs sits between 20 and 30 days. The gap represents thousands of pounds in unnecessary carrying cost.

What Is the 80/20 Rule in Inventory?

The 80/20 rule says that 20% of your inventory items account for 80% of your total inventory value. It's also called ABC analysis, and it matters because not all stock deserves the same attention.

Your A items (the top 20% by value) need tight controls. Regular counts. Accurate demand forecasting. Reliable suppliers. These are the materials where a stock-out stops production and overstocking burns cash.

Your C items (the bottom 50% by value) can be managed with looser controls. Order in bulk, reorder when low, don't lose sleep over exact counts. They cost little per unit even if you hold extra.

Most manufacturers know this intellectually. Few actually apply it. Their spreadsheet treats every item the same. So the planner spends equal time managing 2p washers and 200-pound aluminium billets. That's not inventory management. That's busywork.

What Are the 4 Types of Inventory in Manufacturing?

Every manufacturer carries four types, whether they track them separately or not.

Raw materials. Steel, plastic pellets, fabric, electronic components. Whatever comes in from suppliers before any production happens.

Work-in-progress (WIP). Materials that have entered production but aren't finished. A circuit board with components soldered but not yet tested. A garment cut but not sewn. WIP is the hardest to track and the easiest to lose visibility on.

Finished goods. Complete products ready for dispatch. These should move out quickly. If finished goods pile up, either sales isn't keeping pace or production is building to forecast instead of demand.

MRO (Maintenance, Repair and Operations). Drill bits, lubricant, packaging materials, safety equipment. Not part of the product, but production stops without them. Often tracked separately, often poorly.

A manufacturing ERP tracks all four types in one system. A spreadsheet tracks whichever type the person maintaining it remembered to update.

When Demand Prediction Beats Reorder Points

Traditional inventory management uses fixed reorder points. When stock drops below 500 units, order more. Simple. Also wrong, most of the time.

Reorder points don't account for seasonal swings, lead time variability, supplier reliability changes or shifts in customer demand. A static reorder point set in January might trigger too many orders in summer (when demand drops) and too few in autumn (when it spikes).

AI-driven demand prediction analyses historical consumption, current order pipelines, supplier lead times and seasonal patterns to forecast what you'll need and when. It adjusts automatically. No one reviews a spreadsheet or updates a formula.

According to research by the Institution of Engineering and Technology (IET), manufacturers using AI-driven demand forecasting report 20 to 35% reductions in excess inventory within the first 12 months. That's real money freed up for the business.

The Single Source of Truth Problem

Here's where most manufacturers actually fail. Not in the methodology. Not in the software selection. In the data.

When sales maintains one stock spreadsheet, procurement maintains another, and the warehouse has a third, discrepancies are inevitable. A customer order updated in sales doesn't flow to production. A goods receipt processed in the warehouse doesn't update the procurement record. By Thursday, three departments are making decisions based on three different numbers.

The fix isn't better spreadsheets. It's a single system where every receipt, issue, transfer and adjustment updates one record. When the warehouse scans a delivery, the stock count updates everywhere. When production draws materials, inventory adjusts in real time. When a customer returns goods, the record reflects it immediately.

One number. Everywhere. That's the baseline for everything else.

Connecting Inventory to the Rest of the Factory

Inventory management in isolation is a counting exercise. Connected to production planning, procurement and fulfilment, it becomes an operations engine.

When a sales order is confirmed, the system checks material availability against the BOM. Materials in stock? Schedule production. Not in stock? Trigger a purchase order with the right lead time. Finished production? Notify fulfilment. No emails. No phone calls. No walking to the warehouse to check.

That closed-loop is the difference between inventory management as a cost centre and inventory management as a competitive advantage.

What This Looks Like in Practice

Arcflow provides inventory management built for how manufacturers actually work. Multi-warehouse visibility across raw materials, WIP and finished goods. AI-driven demand prediction that replaces static reorder points. Automated replenishment that routes purchase orders to the right supplier based on 37 KPI variables. Connected to production, sales and fulfilment in one platform, so the single source of truth problem disappears on day one.

Book a demo to see how inventory management works for your factory.