Workshop stock goes wrong less because nobody records it and more because it gets treated as one kind of thing when it is two. Some of it is counted: filters, brake pads, spark plugs, batteries. Some of it is measured: engine oil, transmission fluid, coolant, brake fluid.
Both are legitimately “parts”. They leak through different routes and need different habits. This article separates the two, and flags the one decision worth making once at the start so your stock card does not quietly lie for a year.
Two kinds of stock, two kinds of trouble
| Fluids (oil, coolant, brake fluid) | Parts (filters, pads, plugs, batteries) | |
|---|---|---|
| Known by | measuring | counting |
| Source of drift | the unit changes mid-journey | the wrong lookalike gets picked |
| Leftovers | always, and hard to pin down | none — whole or gone |
| Biggest risk | unit cost is wrong | money sits on the shelf unused |
Two columns, two habits. Flattening them into one is where most workshop drift begins.
The oil problem: the unit changes on the way in
Oil usually arrives large and leaves small. Illustratively: in as a 200-litre drum, out as 3.7 litres for a car and 0.8 litres for a scooter. If the purchase is recorded as “1 drum” while consumption is recorded in litres, the stock card can never reconcile — not because the arithmetic is wrong, but because what came in and what went out are not the same thing.
This kind of drift never sets off an alarm. It accumulates quietly and eventually shows up as profit that looks better than it is, because the oil cost consumed was never fully counted.
Decide the oil unit once, then leave it alone
This is an operational decision, not a technical one, and it only has to be made once:
- Pick the smallest unit you genuinely sell or consume. For most workshops that is the litre — not the drum, not the bottle.
- Record purchases in that same unit. One 200-litre drum is entered as 200 litres at a per-litre cost, not as one “drum”.
- If you sell sealed bottles and decant from a drum, treat them as two products — for example “Engine Oil 10W-40 (1 L bottle)” and “Engine Oil 10W-40 (bulk, per litre)”. It sounds fussy; it is far lighter than reconciling two units every month.
Automan gives you your own list of units, so litres, bottles and pieces can stand as separate units. What the app does not do — and what we will not claim — is convert a drum into litres for you. That conversion is your decision, made once, at the moment you record the purchase.
The parts problem: one part, many vehicles
Parts behave in the opposite way. They are whole or gone; there is no half. The difficulty is not measurement, it is finding the right one in seconds while a mechanic waits at the lift.
Oil filters that fit several models look identical on the shelf. If the product is named “Oil Filter”, whoever is picking will guess — and a wrong guess ends up inside somebody’s engine.
Automan gives you categories, subcategories and brands. The seeded car workshop catalogue already separates Lubricants and Fluids (engine oil, transmission fluid and ATF, coolant, brake fluid) from Spare Parts (oil, air and cabin filters, brake pads and braking systems, spark plugs and ignition, batteries, suspension and steering, belts and timing). What it does not include is a vehicle fitment database — the app does not know which filter fits which model.
That makes fitment something you encode in the product name, which turns naming into stockroom work rather than tidiness. The general principles of searchable naming are in structuring parts product data; for a workshop, add one rule on top: the product name carries the fitment marker your own mechanics say out loud, not a manufacturer code nobody uses on the floor.
Workshop parts are rarely sold over a counter
In a parts shop, goods move through sales. In a workshop, most goods move through jobs — that oil left the drum because a car was being serviced, not because someone bought it at the front desk.
Which makes one requirement absolute: parts consumption has to be recorded inside the job it belongs to, never as a separate stock adjustment. Otherwise two things vanish at once — stock accuracy, and per-job profit, because the cost of goods never attaches to the work that consumed it.
The stock-card mechanics behind that, and how consumed cost follows usage, are covered in batch stock cards for parts. What differs in a workshop is only the proportion: here the job route is the main road, not a side street.
Counting what cannot be counted
Stocktaking in a workshop carries a difficulty an electronics counter never faces: a half-full drum cannot be counted. It has to be measured, and the result is always approximate.
The workable approach is to set your own tolerance for fluids and record the rounding as an adjustment rather than letting it hang. For counted parts there is no tolerance — twelve is twelve. A counting rhythm that does not close the shop for a day is covered in how often to count stock; the only thing to add for a workshop is keeping the two lists apart, so fluid uncertainty never contaminates part numbers that are genuinely exact.
The stock nobody looks at
Workshops have a characteristic disease: a part ordered for one specific car, never fitted, then resident on the shelf for two years. The money left, the item stayed, and nobody remembers it.
Stocktaking does not help here — it only confirms the item exists. What helps is a warning that runs on its own: low stock, dead stock (nothing moving for a defined period) and negative stock. All three exist as automatic alerts in Automan, and they are addressed to you, not to a customer.
The dead stock alert is the one people least often switch on, and the one that most often finds money that stopped moving a long time ago.
A realistic order of work
You do not have to fix all of this in one week. The sequence that tends to hold:
- Decide the fluid unit and record the next purchase in it. Leave history alone — do not backfill.
- Rename the twenty parts you reach for most, not the whole catalogue.
- Enforce consumption through the job document. Highest impact, lowest cost.
- Turn on low stock and dead stock alerts, then let them run for a month before judging them.
Once those four are in place, “how much capital is asleep on that shelf?” becomes a question with a defensible answer. The stock card and cost mechanics behind it are on the spare parts inventory page.