Your warehouse tells you the truth about your processes

Nothing sits in the warehouse by chance. Every part is there because someone decided to order it, to make it or to leave it lying there. The decision may be years old, may be hidden in a system parameter, may be the quiet habit of a buyer. But it exists.

In the Toyota Production System, inventory counts as waste, and specifically as the waste that conceals other waste. The image is well known: the water level in the river is the inventory, the rocks are the problems. As long as the water level is high, you see no rocks. Lower it, and they come into view.

It follows that anyone who wants to reduce inventory without knowing the rocks runs aground. But anyone who knows the rocks can lower the inventory afterwards. Reading the warehouse means finding the rocks before you let the water out.

Over the years I have learned to match certain warehouse pictures to certain process problems. The most common:

• Large lots of a single part: changeover is expensive, or is perceived as expensive. People prefer to make a lot at once. The finding is: look at changeover time, not lot size.
• Buffer stock in front of a machine: the machine is the bottleneck, or is regarded as unreliable. The upstream steps produce ahead so as not to be blamed when it stands still.
• Buffer stock after a machine: the downstream step does not take the parts, because it is planned differently. Two planning logics, one pallet in between.
• Dead stock with a faded label: an order was cancelled, a variant discontinued, a safety stock never reviewed. Nobody has the responsibility to dispose of it.
• Pallets in the aisle: the official warehouse is full or too far away. The hall has built its own warehouse, because the process does not deliver what it needs.
• Duplicate stocks in the hall and the office: the clerks keep samples, spare parts or documents in reserve because they do not trust the hall or the system.

Each of these findings points to a place in the process. The inventory is the symptom you can touch.

Of all the findings, buffer stock is the most important to me, because it says something about people. A buffer stock arises when two places do not trust each other. Pre-assembly builds ahead because it does not believe final assembly will say in time what it needs. Final assembly keeps stock because it does not believe pre-assembly will deliver in time. Both have experiences that justify this.

The buffer stock thus becomes the cushion between two departments that should talk to each other and do not. As long as it is there, they do not have to talk. They reach into the shelf. The inventory replaces the conversation.

That is why the question for every buffer stock is not: how do we get rid of it? But: which conversation is not taking place here? Usually it is a simple, daily, short conversation about what is needed today. Once that is running, the buffer stock shrinks by itself.

In a plant with several assembly lines, shelves full of subassemblies sat between pre-assembly and final assembly. Plant management saw this as a space problem and planned an extension of the storage area.

Instead, we counted the subassemblies on the shelves and compared them with the demand of the next few days. Some of them were not needed at all in the foreseeable future. Another part belonged to variants that were just being phased out. Pre-assembly had produced to a weekly plan that had been drawn up weeks earlier. Final assembly worked to the daily plan from sales, which changed several times a week. Two plans, no coordination, a shelf in between.

The solution was a shared board on which final assembly enters the demand for the next two days every morning, and a pre-assembly that produces exactly to that. No shelf building. The shelves became emptier over the following months, and plant management had its space without spending a single franc on construction.

Inventory is not only made of metal. In order processing it sits as a stack of unprocessed orders in the in-tray, as a list of open queries, as an inbox with a hundred unread emails. In design it sits as drawings in the approval loop. In accounting as invoices waiting for a signature.

The rules are the same. A large pile in front of one place shows a bottleneck or a trust problem. A pile that always stays the same size shows that it is regarded as normal and nobody sees it any more. And a case that sits for a very long time is the office's dead stock: an order with a question attached that nobody wants to ask.

So anyone who wants to read their plant should also read the in-trays. They are the warehouse of the administration, and they lie just as little.

You do not need a stocktake or a system report for this. You need an hour and a notepad.

Walk through the plant and note every stock that catches your eye, with three questions: what is sitting here? Who decided that it sits here? Which problem is being concealed by it? You will think longest about the third question. That is a good thing.

Then sort the findings by the cause behind them: changeover, planning, interface, trust, missing responsibility. You will find that the causes repeat. A plant rarely has twenty different warehouse problems. It usually has two or three process problems that show up in twenty shelves. Solve one of them, and several shelves will become emptier. That is the moment when inventory reduction is no longer a goal, but a consequence.