Dead stock: where to start once the easy part is done

There is a very precise moment in the life of an inventory reduction program when everything gets harder. The first millions have come out painlessly. Then, as one director put it to me, "it starts to grind." That isn't a willpower problem. It's that the inventory that was easy to cut has already been cut.

Dead stock isn't where you go looking for it#

Instinct points at the big lines: they're the ones that weigh in euros, so that's where the money is. That's only true in year one.

On big items, the money sleeps briefly#

On fast-moving items, inventory is large but it turns. Carrying a hundred and fifty days on an item that sells every day is a choice, sometimes the right one, especially when replenishment is slow. The money sleeps there only briefly, and a stockout is immediately visible.

The seam is in the multitude#

The real seam is elsewhere: two or three hundred days of cover on class C items, which account for a few percent of revenue but occupy a considerable amount of value in aggregate, because there are so many of them. Nobody looks at them individually — each one is too small to deserve a meeting — and that is exactly why they pile up.

Three definitions of "dead", and the one to use#

Depending on the definition, you're not talking about the same goods or the same amount.

Inventory with no movement: simple, and misleading#

Say no issue in twelve months. Easy to calculate, and misleading: a safety spare that never moves is doing its job perfectly.

Excess inventory: more accurate, but it needs a target#

It is the portion above target cover. More accurate, but it requires a target cover per item, which few companies have.

How do you spot dead stock before it becomes unsellable?#

The item still moves, but half as fast as eighteen months ago. This is the most useful signal, because it arrives before the goods become unsellable.

We use the third, with a simple rule: for each item, compare issues over the last six months with those of the six months before. A drop of more than half, on an item carrying more than a hundred days of cover, deserves a decision.

On the first thirty lines, there are about twenty where I can tell you right away what happened.

That kind of reaction is a good sign. A useful list is one the director recognizes.

The four ways out, in order#

Once an item is identified, there are only four outcomes. Taking them in this order avoids discounting what could still have been sold.

1. Move it#

It's sleeping here and missing there. The most profitable exit when it's possible — and it isn't always, see below.

2. Sell it at full price, to someone looking for it#

A dormant item is almost always an item nobody actively offers any more. Putting it back into an offer, a template quote, a sales argument, costs less than a discount.

3. Discount it#

Losing twenty percent now costs less than carrying the goods for two more years.

4. Write it off#

Provision and clear. The only one that hurts, and the one that gets postponed too long, because it means admitting an old mistake.

The classic error is jumping straight to number three. Number two is almost always under-used.

The inter-site transfer trap#

Moving seems obvious: the goods are already paid for, you just have to shift them. Except that transport has a price, and it isn't proportional to distance.

Why doesn't freight follow distance?#

I have in mind a multi-site business where freight between two islands about a hundred and fifty miles apart costs twice the freight from the mainland port several thousand miles away. Under those conditions, transferring a class C item costs more than discounting it on the spot.

The calculation the application makes for you#

The rule we code into the applications we deliver fits in a single comparison, made automatically:

transfer cost + handling cost vs. net value of the goods × probability of selling them at the destination

If the first term exceeds the second, the application doesn't propose the transfer. It proposes something else. That is exactly the kind of judgment a human won't redo five hundred times a month, and a machine redoes without tiring.

Stopping it from happening again#

Clearing dead stock without changing what produced it guarantees a repeat in eighteen months. Three mechanisms are enough, and none of them is spectacular:

How often should you recalculate turnover classes?#

An item moving from B to C should see its target cover fall by itself. A frozen classification is a wrong classification within a year.

Alert on the trend, not the level#

"This item carries two hundred days" teaches nobody anything. "This item carried ninety days six months ago" triggers action.

Make carrying cost visible at purchase#

A buyer who sees, as they approve an order, what the proposed quantity will tie up and for how many months, adjusts of their own accord.

How we surface dead stock#

Dead stock is not something you go looking for; it announces itself. Every item carries its last issue date and its cover, and the ones past their threshold surface on their own, with their annual holding cost beside them.

That figure is what triggers the decision, because it makes comparison possible: keep, clear or transfer stops being a matter of feel. See the inventory and replenishment module.

Dead stock: what to remember#

The seam isn't in your big items, it's in the multitude of small ones nobody looks at one by one.

Spot it through deteriorating turnover rather than absence of movement: you gain twelve to eighteen months on the diagnosis.

And before transferring anything between two sites, do the transport calculation. Clearing on the spot often costs less than moving — which is counter-intuitive, and that's precisely why people get it wrong.