Stock control software

Stock control software

Searching for "stock control software" almost always returns the same thing: lists of programs, comparisons and pages promising to sort out a small company's inventory in an afternoon. The problem is that the tool is not the starting point. If nobody records what leaves the store, if no one can say which document a box left on, and if the physical count happens once a year, no program will fix the mismatch: it will only make it visible, which is not the same thing. A system organises what is already being done somehow; it does not invent discipline where none exists.

This article tries to do the opposite of an advert. Before talking about programs, it explains what controlling stock really means, why an inventory that "does not match" is almost always a process problem rather than a system one, and what to ask before buying any stock software. It then sets out the functions such a system should show every day, a numerical example so each reader can run their own numbers, and an honest closing: in some cases no software is needed at all.

Transparency note: this article is published by the team that develops Kardex Tauro, an inventory and invoicing system for small and medium-sized businesses. We do not compare brands, we do not promise prices, and we will not describe features that do not exist. The intention is for readers to decide on their own terms, even if they decide to buy nothing.

What controlling stock really means

Having a list of products with quantities on it is not stock control. A list says what somebody believed was there on the day they wrote it, and from that moment on it starts to age. Stock control is something else: it is being able to answer four questions at any moment, without opening a folder, without calling anyone and without waiting for the month-end count. Those four questions are the following.

• What is there. How many units of each product, each reference and each presentation, counting both what is available and what is already committed.
• Where it is. In which warehouse, on which shelf and, where relevant, under which cost centre.
• What it is worth. At what cost that inventory sits today, and how that cost moved with the most recent purchases.
• What will be needed. What is running out, and what is sitting still, going stale or approaching its expiry date.

What matters is not each question on its own, but that all four can be answered together, at the same instant. Knowing that two hundred units of a product exist without knowing which warehouse holds them is half an answer. Knowing how many units there are and what they cost, but not knowing which ones are running out, serves the accounting report and not the operation. Real control appears when all four answers come out of the same record, and that record is fed every day with whatever happens in the store.

There is a second, less glamorous condition that decides whether control works: speed. A correct answer from last week will not help you dispatch today's order. If finding out how many units are left means asking the person who knows, then this is not stock control, it is asking a person. The system can be very simple; what it cannot do is arrive late.

Controlling money and controlling units are not the same thing

There is a very common confusion in small businesses: believing that stock control is an accounting task. It is not, or at least it is not only that. Accounting needs to know what the inventory is worth at the end of the period; operations need to know how many units there are and where they are. They are two different questions, both are legitimate, and the mistake lies in using one answer to cover for the other.

Controlling value without controlling units produces reports that balance on paper and warehouses that do not hold what they claim to hold. Controlling units without controlling value produces a tidy storeroom and a manager who has no idea how much money is sleeping on the shelves. The middle ground is the same record feeding both views: every movement changes a quantity and, in the same act, changes a value. The table below sums up the dimensions worth watching and what breaks when one of them is missing.

The table has a practical reading. The first two rows are the ones almost everybody looks at; the remaining four explain most of the surprises. A business can have the value of its inventory worked out perfectly and, at the same time, have no idea where the product a customer came to claim actually is.

Why stock that does not match is almost always a process problem

When a physical count finds fewer units than the system reports, the most common reaction is to blame the program or look for a thief. Both explanations exist, but in a small business they are the minority. Most mismatches come from a single source: the system does not know something that happened in the real world, because nobody told it, or told it too late.

The number one cause is outgoing movements that are never recorded. A box leaves for an urgent order, it is delivered, it is invoiced, and the record is left for later. That "later" never arrives. The second cause is batched recording: instead of writing each movement down as it happens, everything is written up on Friday from memory. At that point what gets recorded is not what happened but what is remembered, and what is remembered is always rounded off.

Then come the quieter causes, which are almost never recognised as causes at all. Goods that leave with no document because they were used for an internal repair, for a sample, or for the office's own consumption. Products that were damaged and never written off, so they still sit in the system as though they were available. Unit-of-measure mismatches: the system counts boxes while the salesperson dispatches loose units, or the other way round. Typing errors when quantities are captured, which cancel each other out one month and pile up the next. And finally, a count done badly: two people count the same shelf, one of them skips a row, and the difference shows up as if it were real.

None of these causes is fixed by buying a program. They are fixed by changing the process: whoever hands goods over records them; whoever records leaves the document; whoever damages writes it off with a reason; whoever counts counts from a printed list, in a fixed order. Software can make that process much faster and much harder to forget, and that is a real contribution, but it cannot do it in place of the person.

The seven questions to ask before buying any stock software

If after reading the above the process is more or less clear, the decision about the program becomes far easier. These seven questions work for any option, and none of them requires technical knowledge to ask.

What the system should show every day

Beyond any brand, there is a short list of things the person responsible for the storeroom should be able to see without asking anyone's permission. If the system does not show them, control depends on somebody's patience and ends up depending on their memory. These are the items consulted daily in an inventory that is taken seriously.

• Current stock by product and by warehouse, available to consult at any moment.
• The stock ledger per product: every movement with date, document, concept, quantity, average cost and resulting balance.
• Incoming and outgoing movements through their own modules, without having to rebuild the day by hand.
• Manual adjustments, both those that add and those that subtract, each with its reason: shortages, surpluses, damage or recording errors.
• Physical counts and their comparison against what the system says.
• Batches, with their expiry dates and the days they have left.
• Serials, up to three identifiers per product, with their traceability and warranties.
• The stock report and the inventory ageing report, which shows the products that have stalled.
• The negative stock warning, which lets you see such balances and correct them before they turn into an accounting problem.

That list is exactly what Kardex Tauro covers. It also works with several warehouses and cost centres, allows transfers between warehouses, imports from and exports to Excel, and handles per-user roles and permissions and backups. The database can be local, on a network or on a web server, depending on how the business is set up. It is worth stating what is not there: there is no automatic minimum-level or reorder-point alert, so that part of the work remains a decision for the person doing the buying.

A worked example with numbers

To stop the discussion being abstract, it is worth doing the arithmetic on a small, concrete case. Take a product that the system reports as holding one thousand five hundred units of theoretical stock. The physical count arrives and finds one thousand four hundred and sixty-three. The difference is thirty-seven units. At an average cost of eight thousand four hundred per unit, that difference is worth three hundred and ten thousand eight hundred.

It is worth looking at the proportion before drawing conclusions. Thirty-seven units out of one thousand five hundred is about two and a half for every hundred: a small mismatch, the kind almost nobody reports because it looks normal. And that is precisely why it matters. A small shortfall that appears once is an episode; the same shortfall repeating every month stops being a counting problem and becomes a problem in how outgoing movements are recorded. Counting better will not fix it, because the count was already done properly: the wrong number is the one in the system, and the system is wrong because nobody told it about every outgoing movement.

The exercise also works the other way around. If the difference does not repeat, if it shows up one month and disappears the next, then it is worth reviewing the counting procedure, the calibration of the scale, the unit of measure, or the product that was captured under the wrong code. The same figure leads to two opposite decisions depending on what happened in the previous months, and that is the reason the stock ledger of each individual product is worth more than the total of the whole inventory.

How to tell whether control is improving

A stock system is not judged by the number of reports it produces, but by three simple signals that can be reviewed every month without being an accountant.

The first is inventory accuracy: the difference between what the system says and what the count finds, measured product by product rather than as one big average. A business can have a small total difference and a handful of products with enormous differences that cancel each other out; looking only at the total hides the problem. The second is turnover: how many times the inventory sells through and is replenished in a period. A product with high turnover rarely goes stale, and a product with low turnover is a candidate for sitting still. The third is stockouts: the occasions when a product that normally sells was not available when somebody asked for it. A stockout does not appear in the inventory report, because the inventory was at zero and zero is a valid number; it appears at the till, when the sale did not happen.

All three signals have something in common: they cannot be calculated on day one. Several months of consistent recording are needed to establish a baseline, and that baseline is what later allows you to say whether control improved or got worse. A business that starts recording today will not know whether it has improved for some time; what it can know from the first day is whether the recording is complete, which is the only condition that depends on it.

Common mistakes

Almost every attempt at stock control that fails does so for one of these reasons, and none of them has anything to do with the quality of the program chosen.

• Buying the software before deciding the process, in the hope that the tool will organise what nobody has decided to organise.
• Using the system only to invoice and leaving the inventory for later, so that the invoice and reality drift apart from day one.
• Recording everything at the end of the month from memory, which is the fastest way to turn a record into an estimate.
• Counting once a year, when counting should be part of the routine rather than an annual event.
• Merging similar products under one code to "simplify", which loses exactly the detail you wanted to have.
• Not using batches when the product expires, and discovering that something has gone past its date only once it can no longer be sold.
• Adjusting the system so that it matches the count without investigating the cause, which is the way to erase the evidence and keep the same problem.
• Letting anyone change quantities without leaving a trace of who did it and why.

An honest closing

After all of the above, it is worth saying plainly: some businesses do not need stock control software. If the catalogue is short, if one person manages it and that person knows by heart where everything is, a well-kept notebook or a tidy Excel template is enough, and it is cheaper. Software starts to pay for itself when one of these conditions appears: more than one person touching the inventory, more than one warehouse, products that expire or carry serials, counts that no longer get finished, or a value of losses that comfortably exceeds the work of keeping the record up to date.

When those conditions are present, Kardex Tauro is an option that covers what this article described: stock by product and warehouse, a stock ledger with average cost, adjustments with a reason, compared counts, batches with expiry, serials, stock and ageing reports, a negative stock warning, transfers between warehouses and export to Excel. It promises nothing beyond that, and that is the point: a tool does not fix a disordered process, but a tidy process with no tool only holds up for as long as the memory of the person running it lasts.

The useful decision is not which program to buy, but which single movement is going to be recorded today. If that answer exists, any reasonable tool will do. If it does not exist, none will help, neither the most complete nor the most expensive, because the problem was never in the software. Kardex Tauro is built for the moment that answer already exists and you need to stop holding it up with memory alone.