Van stock: why engineer boot stock never reconciles

Ask any field service manager what their van stock is worth and you will get a number. Ask how confident they are in it and you will get a pause. It is usually somewhere between a guess and a hope, and the gap between the two is frequently larger than the entire main store.

The instinctive explanation is that engineers are taking things. Occasionally true, overwhelmingly not. The structural explanation is better and more useful: a van is a stock location with no storeman, no door, no counter and no quiet moment. Every control that makes a main store work is absent, and then everyone is surprised the numbers do not tie.

The four ways van stock actually leaks

1. Fitted but not recorded. The part went into the customer’s boiler. The job was closed on the phone in the rain. Nobody typed the part number. This is the biggest one by a distance, and it is a recording problem, not an honesty problem.
2. Van to van. One engineer is short, another has three, a transfer happens in a car park, and no system on earth knows about it. Entirely sensible behaviour that destroys the record.
3. Returns that never came back. A part is taken for a job, the job does not need it, and it lives in the footwell for four months before returning as a mystery.
4. Replenishment counted against the wrong thing. Stock is issued to the van as a bulk top-up, and the record now says the van holds a delivery rather than the sum of what it has.

None of those is fixed by counting more often. Counting tells you the number is wrong; these tell you why, and they need different remedies.

Treat the van as a site, not as a person

The most common modelling error is holding van stock against the engineer. It feels right — one engineer, one van — and it breaks the first time somebody is on leave and a colleague drives their van, or when a van is swapped for a service loaner.

Hold the stock against the van. Hold accountability against whoever is driving it. Those are two facts and they change at different rates.

Stock lives in a vehicle. Responsibility lives with a person. Conflate them and every holiday becomes a reconciliation problem.

Make recording cheaper than not recording

This is the whole game, and it is worth being blunt about the arithmetic. An engineer finishing a job in the wet has perhaps twenty seconds of goodwill for admin. If your recording step costs sixty, it will not happen — not out of defiance, but because the job comes first and the admin is always something to do later, and later has other jobs in it.

So the target is not compliance, it is duration. Scanning a barcode on the part as it comes off the shelf in the van takes about three seconds and requires no typing and no signal. Selecting from a list of what that van is supposed to hold takes about ten. Typing a part number from memory into a form takes forty and produces errors that take an hour to unpick.

• Record consumption at the van, at the moment it happens, not at the office afterwards.
• Make it work with no signal and reconcile when there is one. Vans go into plant rooms and basements.
• Show the engineer what the system thinks they are carrying. People correct a wrong list far more readily than they fill in a blank one.
• Give van-to-van transfers a two-tap path. If the honest action is hard and the dishonest one is easy, you have designed a leak.
• Let a return be recorded as a return rather than as a negative issue, so the history reads correctly later.

Replenishment: max-min, not top-up

Most van replenishment is a request-driven ritual — the engineer asks for what they noticed they were short of, which is a list of what they ran out of last week rather than what they will need next week.

A van list is a good candidate for straightforward max-min, because the demand is more predictable than people expect and the cost of a stockout is enormous: a second visit, a lost slot, sometimes a lost customer. Set a maximum by what physically fits and a minimum by how long a resupply takes, and replenish to the maximum rather than to the request.

Counting a van without losing a day

Do not attempt a full count of every van. Cycle count them: a short list per van, counted at the depot on a rota, blind — without showing the expected figure.

The right cadence is roughly monthly per van on the top twenty lines by value, which takes ten minutes and catches drift while somebody can still explain it. Anything counted a year later is just a number.

And track accuracy per van rather than in aggregate. Aggregate accuracy hides the important finding, which is that eleven vans are fine and two are not — and the two are a coaching conversation, not a policy change.

What good looks like

You should be able to answer three questions without ringing anybody: what is in van 7 right now, what did van 7 consume last month, and which vans are carrying stock they have not used in six months. That third one is usually where the money is — vans accumulate parts for a job type the engineer no longer does, and nobody ever takes them back out.

Nilstock treats each van as its own site with its own stock and its own minimums, records consumption at the point it happens, and reconciles the ledger nightly so drift shows up as a number rather than a surprise. The fleet and field service page covers the depot-and-van shape specifically. For setting the minimums, see setting reorder points; for the counting rota, cycle counting.