Bramvia — Business Central Experts
You added inventory on purpose. Now it's holding cash you need — and nobody can tell you which items
63% of manufacturers raised inventory buffers, 22% of distributors hold over 90 days of stock, and 26% report 6-10% deadstock. That was a rational decision. What wasn't planned is that nobody can now say which SKUs are the dead ones.
Short answer: between tariff turbulence, freight volatility and supply delays, US manufacturers and distributors made a deliberate call: stop running lean, protect service levels, carry more stock. 63% of manufacturers increased safety stock or buffers over two years; 22% of distributors now hold more than 90 days of inventory; 70% manage over 5,000 SKUs. That was rational. What nobody planned for is the bill: 26% of distributors report 6-10% deadstock, inventory carrying cost has climbed with rates and storage, and 35% name inventory accuracy and visibility as a leading operational problem — which means the cash is trapped and nobody can say in which items. Meanwhile warehouse turnover runs around 49% a year, so the people who knew what was in the back corner have left. This article is about getting the cash back without breaking service.
What the data actually says
| Finding | Figure | Source |
|---|---|---|
| Manufacturers that raised safety stock or buffers | 63% | WSI survey |
| Distributors holding more than 90 days of stock | 22% | Phocas, distributor survey |
| Distributors reporting 6-10% deadstock | 26% | Phocas |
| Distributors managing over 5,000 SKUs | 70% | Phocas |
| Naming inventory accuracy and visibility a leading challenge | 35% | WSI |
| Wanting new demand planning techniques | 54% | Phocas |
| Annual warehouse worker turnover | ~49% | KPI Solutions |
| Manufacturers naming skilled labor their biggest external challenge | 79% | 2026 Manufacturing Outlook |
| Saying tariffs and unclear trade policy make planning harder | 47% | 2026 Manufacturing Outlook |
Read those together and the picture is specific: more stock, more SKUs, less visibility, fewer experienced people, and planning assumptions that changed three times this year.
Why "just reduce inventory" is bad advice
The instinct is to cut. It's the wrong instruction, because the buffer is doing a job: it is absorbing lead-time variability that got worse, and protecting customers who will switch supplier over one stockout. Cutting across the board re-creates the problem the buffer was built to solve.
The right instruction is narrower: separate the buffer that earns its keep from the stock that is simply stuck. They look identical on a balance sheet and completely different in the item ledger.
- Working buffer: items with demand, held deliberately, sized against real variability.
- Dead stock: no movement in 12 months. Cash already spent, plus space, insurance and handling, forever.
- Slow and mis-sized: items that move, but where you hold 300 days of them because the reorder point was set in 2021 on different lead times.
The third category is usually the biggest and the least discussed. Nobody audits reorder points.
Four questions that free cash in 30 days
1. Which items haven't moved in 12 months, and what are they worth? On-hand value versus 12-month consumption, per item. Simple query, rarely run. In companies that have never done it, the dead portion is typically 10-25% of inventory value. That's your liquidation list — and it funds everything else on this page.
2. Which reorder points and safety stocks are still based on pre-2024 lead times? If your supplier moved from 4 weeks to 9 and back to 6, and nobody updated the parameters, you are either over-holding or firefighting. Recalculate against actual received lead times from the last 12 months, not the ones in the vendor card.
3. What does your deadstock actually cost you per year? Carrying cost is not a theory at current rates: capital, storage, insurance, handling, obsolescence. Put a number on it and the liquidation conversation gets short.
4. Which SKUs are you protecting for customers who don't pay for it? Cross inventory value against customer contribution margin. Holding 90 days of a slow item for a customer who is already margin-negative after freight and rebates is a decision — just make it consciously. How to compute that margin properly.
The labor problem is a data problem
Nearly half of warehouse staff turn over each year. That has a consequence people underestimate: process knowledge that lives in people leaves with them. Where things are stored, which supplier is reliable, which customer always calls before cutoff, which part number is really two parts — none of that is in the system, and it walks out the door twice a year.
We saw exactly this at a US distribution client: high turnover meant procedures existed in heads rather than in the ERP, so every departure cost re-learning and errors. It never appeared as a cost line. It was larger than the freight increase everyone was discussing.
The practical countermeasure is not a retention programme — it is putting the knowledge in the system: directed putaway and picking, bin logic that a new hire can follow on day one, documented approval rules, and item data clean enough that the right part is picked without tribal knowledge. That is also what makes automation possible later; 58% of manufacturers are investing in automation, and automation over dirty master data automates the errors.
Reshoring is redrawing warehouse networks
This is the quieter change. Per the September 2026 reshoring survey, momentum keeps building — limited by policy uncertainty and skilled-worker shortages, but real. Downstream: 35% of manufacturers are adding US warehouse capacity to support reshored production, 34% are repositioning facilities closer to new manufacturing sites, and three quarters are rethinking their warehouse strategy at regional or national level.
If that's you, two things matter more than the real estate:
- Multi-location inventory has to be real in the system — stock per site, transfers, and a planning engine that knows which warehouse serves which customer. Not one aggregate number with a spreadsheet for the split.
- A new site is usually a new entity or at least a new dimension: intercompany flows, consolidated reporting, and tax by jurisdiction. Designing that after go-live is expensive.
Where AI actually helps (and where it doesn't)
49% of manufacturers are expanding AI-powered analytics for real-time monitoring, and the useful applications are narrow and boring: demand forecasting per item family, flagging items whose pattern changed, detecting duplicate part numbers, and reading two years of history to find the dead stock nobody knew about. That last one is cheap, immediate, and the reason a diagnosis that used to be a consulting project is now a fixed fee.
Where it doesn't help: 82% of manufacturers cite a lack of AI-ready skills as their top workforce challenge, and no model fixes a process that was never defined. AI over clean data answers questions. AI over a warehouse nobody has counted produces confident nonsense.
How to start this month
- Run the dead stock report. On-hand value versus 12-month movement. One query, biggest prize.
- Recalculate reorder points against actual lead times.
- Cost your carrying and take the liquidation decision with a number in front of you.
- Cross inventory against customer margin before protecting anything else.
- Write down what the leavers knew — and then configure it, so the next hire doesn't need it.
If you'd rather have this done than build it, that is exactly what our fixed-fee Profit Leak Audit produces: you send an extract from whatever system you run, and in 10 working days you get the dead stock, the mis-sized parameters, the negative-margin items and the manual-work hours, quantified in dollars. It works on SAP, Epicor, Infor, NetSuite, Dynamics or QuickBooks, and it is not a proposal to replace any of them.
FAQ
Isn't deadstock just a write-off decision for accounting? Accounting writes down the value. This is about getting cash and space back, and about stopping the purchasing pattern that created it — which is an operational decision.
We know we have dead stock, we just don't have time to find it. That is the normal answer, and it is why the number keeps growing. The query takes minutes once the data is extracted; the work is in the extraction and the follow-up.
Will reducing inventory hurt service? Not if you separate dead stock from working buffer. Liquidating items with zero movement in a year has no service impact by definition.
Our ERP can't report on this. Then that is a finding. Plenty of mid-market systems hold the data but cannot present it; some genuinely don't capture receipts against lead times. Knowing which is your case is worth more than another dashboard.
Cash tied up and no list of which items? Free assessment, no commitment — first reply within one working day.