A practical guide to building an inventory system that earns its place: why spreadsheets fail at stock, the two numbers that prevent both stockouts and overstock, modeling movement instead of just counts, and building a tracker your team keeps accurate.
Retailers, studios, workshops, ecommerce teams, and product businesses tracking stock in spreadsheets that keep going wrong.
- A stock model built on movement, not just current counts
- Reorder points that prevent both stockouts and dead stock
- A tracker accurate enough that the team actually trusts it
Inventory is where two expensive mistakes hide: running out of the thing customers want, and drowning in the thing they do not. A spreadsheet of current counts catches neither until it is too late, because it shows a number without the movement and the thresholds that make the number mean something. This guide is about building a tracker that does not just say how much you have, but tells you what to reorder, when, and what to stop buying.
Because a spreadsheet shows a count, and inventory is about movement and thresholds. A cell that says "42" tells you nothing about whether 42 is dangerously low or wastefully high, how fast it is selling, or when to reorder. And because a spreadsheet trusts every hand equally, the count drifts the first time someone forgets to update it after a sale or a delivery, so the number you are staring at is quietly wrong. Inventory needs a system that records movement and flags thresholds, not a grid of last-known guesses.
The failure is expensive in both directions at once. Trust a stale count that reads high and you sell something you do not have, disappointing a customer and scrambling to fulfill. Trust one that reads low and you reorder stock you already have, tying up cash in a shelf. The spreadsheet cannot warn you about either, because it has no idea what "too low" or "too high" means for each item, and no record of the movement that would let it learn.
This is exactly the kind of problem software is for: not a prettier count, but a system that watches stock levels against thresholds and tells you what needs action before it becomes a problem.
You do not need a complex forecasting engine to fix most inventory pain. You need two numbers per item, and a system that watches them for you.
A homeware shop kept a spreadsheet of counts and lurched between selling out of bestsellers and overbuying slow movers. Setting a reorder point and quantity per item, and letting the system flag anything below its point, they stopped discovering stockouts at the till and stopped reordering the candles nobody bought. The counts were the same data they always had; the two thresholds are what turned it into decisions.
The single change that keeps inventory accurate is recording movement, not editing a count. Every sale, delivery, return, and adjustment is a movement that changes the level; the current count is the result of the movements, not a number someone types. Model it that way and the count is always derivable, auditable, and hard to silently corrupt.
The reason movement-based inventory stays accurate while spreadsheets drift is that nobody ever overwrites the count; they record what happened, and the count follows. A miscount becomes a traceable movement to correct, not a mystery number. This is the difference between a system you can trust at the till and one you double-check by walking the shelves.
An inventory system only works if updating it is faster than not updating it, so the whole design goal is to make recording a movement effortless at the moment it happens: a quick action at the till, on the receiving dock, at the point of a return. If logging a sale is slower than the sale itself, the system drifts; if it is a tap, it stays true.
Building this as an application you own, shaped to your products, locations, and suppliers, is what lets the tracker match how you actually hold stock rather than forcing your business into a generic template. Describe your items, the movements that change them, and the two numbers per item, and the system is generated to watch your stock and tell you what needs action, yours to keep accurate and to change as the range grows.
Because they show a current count with no sense of movement or thresholds, and they trust every hand equally, so the number drifts the first time someone forgets to update it after a sale or delivery. A count of "42" cannot tell you whether that is too low, too high, or how fast it is selling, which is exactly what prevents the expensive mistakes.
Two numbers per item and movement tracking. A reorder point flags when to order more so you avoid stockouts, a reorder quantity sets how much so you avoid dead stock, and recording each sale, delivery, and return as a movement keeps the count accurate. Even rough first thresholds remove most of the pain immediately.
Record movement instead of editing a count. Every sale, delivery, return, and adjustment is a logged event, and the current level is the sum of those events rather than a number someone types. Nobody overwrites the count, so a miscount becomes a traceable movement to correct, not a mystery, and the number stays trustworthy at the till.
Not anymore. You can describe your items, the movements that change them, your locations and suppliers, and the reorder point and quantity per item, and get an application that watches stock and flags what needs action. Building it as software you own lets it match how you actually hold stock and change as your range grows.