Four causes, in the order you should rule them out.
By Salyntra · 26 August 2026
When a salon’s stock figures stop matching the shelf, the first suspicion is usually theft and it is almost always wrong. The gap is nearly always one of four ordinary things: product used during services and never deducted, refills received but not recorded, sales rung up against the wrong item, and a physical count that has not happened in long enough that nobody trusts either number. Each has a different fix, and working through them in that order will close most of the gap before you have to consider anything uncomfortable.
This is the biggest one, and it is invisible because it never involves a transaction. Colour, developer and treatment go onto a client during a service; nothing is sold, so nothing is deducted, and the shelf quietly empties faster than the system believes.
The fix is to attach expected product use to the service itself, so performing a colour deducts what a colour consumes. It will never be exact — stylists mix by eye — but an approximation that runs automatically beats an exact figure nobody records.
A box arrives during a busy afternoon, somebody puts it on the shelf, and the paperwork waits for a quieter moment that does not come. Now the system is under-counting, which is the opposite error from the one above and hides it.
This is a process problem rather than a software one. The only reliable answer is that stock is entered when it is unpacked, by whoever unpacks it, before it reaches the shelf.
Two shampoos from the same range, different sizes, similar bottles. One is sold and the other is deducted, and both counts are now wrong in opposite directions — which is worse than one being wrong, because the total still looks plausible.
Photographs on the product records fix more of this than any amount of staff training.
Every system above drifts. A physical count is what resets it, and a salon that has not counted in a year does not have a stock figure at all — it has an opening balance and a hopeful arithmetic.
Count the fast-moving lines monthly and everything else quarterly. A partial count you actually do is worth more than a full one you plan.
Then you will only see it once the four causes above are closed, because until then the noise is larger than the signal. That is the practical argument for doing this in order rather than starting with the accusation: a clean count is what makes the question answerable at all.
Fast-moving lines monthly, everything else quarterly. The point is regularity rather than completeness — a count you do beats a count you schedule.
Yes, because the alternative is not a more accurate number but no number. An approximation that updates on every service tells you when a line is running out; an exact figure from six months ago does not.
Thirty days, every feature, no card. Import your clients and services and judge it against how you actually work.