How to Do a Stock Take Without Closing Your Shop

Somewhere in Lagos this Sunday, a shop owner is taping a handwritten sign to the door: closed for stock take. Inside, the staff count cartons in the heat, tired and rushing, while the regulars walk two streets over and buy from someone else. By evening everyone wants to go home, so the last shelves get counted fast and badly.

That sign costs you twice. Once in the sales you never made. Again in the wages you paid people to produce numbers you only half trust. And you get to do all of it maybe once a year, which means a problem that started in February sits quietly in your stock until December.

You do not have to close to count. A provisions shop, a pharmacy or a boutique can keep serving customers and still know its stock better than the shop that shuts down every year. The method is called cycle counting, it costs almost nothing but discipline, and this guide walks you through it week by week.

A stock take answers the one question your sales report cannot: did the stock you paid for become money in the drawer? Sales can look healthy all month while cash quietly leaks, because goods you bought never turned into recorded revenue.

Counting is how you find the leak. Some of it is theft, by staff or by customers. Plenty of it is not. A supplier delivers ten cartons, the delivery note says twelve, and someone signs anyway. A cashier rings up the wrong size. A broken bottle gets swept up and never written down. Expired sachets go in the bin without a record. None of this appears anywhere until you physically count the shelf and compare it with what your books say should be there. Skip the comparison and you are running the shop on faith. Faith does not pay rent, and it does not pay NEPA either.

Be honest about what an annual count really is: a confession that you lost track eleven months ago and only now got round to checking.

A carton that went missing in February surfaces in December. By then the delivery note is buried, the staff on shift that week may have left, and the supplier stopped answering questions about that invoice long ago. You end up with a number, but no story behind it and nothing you can fix. The problem is not that you counted. It is that you counted too late for the count to change anything.

The alternative has a name: cycle counting. Instead of tallying the whole shop in one exhausting day, you count one section at a time, on a rotation, while the shop trades normally. Beverages this week. Tinned food the next. Toiletries after that. Over a few weeks the rotation covers everything, then it simply starts again.

A single shelf of forty products takes fifteen to thirty minutes, and a person counting forty items gets them right. The same person counting four thousand items on a hot Sunday starts drifting by the third hour, and the last aisles of the day are guesswork with a straight face.

The other prize is time. When a count shows three cartons missing from a shelf you counted two weeks ago, the gap opened in the last two weeks. Not sometime in the last year. Two weeks is a window you can actually investigate: the delivery notes are still on top of the pile and the staff still remember who worked which shift.

Not every product deserves equal attention. In most shops a small slice of the lines, often around a fifth, brings in the bulk of the money: the cold drinks, the popular sachets, the recharge cards, the phone accessories. Retailers call the ranking ABC.

Your A items are those fast movers. Count them every week, because that is where errors and theft cost the most and stay hidden the shortest time. B items, the steady middle, get a monthly count. C items, the slow tail (the odd spice, the stationery nobody has asked for since Easter), can wait for a quarterly pass and nobody suffers.

Do not rank by feeling. Pull the sales report from your till and let the numbers name your A list. Owners are routinely surprised by what is actually carrying the shop, and that surprise is exactly why the report beats instinct. Then build your counting rotation around the ranking instead of walking the aisles in whatever order they happen to sit.

Timing does half the work. Counting a shelf while customers pull items off it is how you manufacture fake shortages, so never count a section at peak.

Every shop has dead spots: the lull after the morning buyers leave, the slow stretch before evening pickup, the calm half hour after you lock up while the staff are still around. A single section fits inside any of them. If you must count during trading, choose an aisle nobody is standing in, and if a customer needs something from the shelf you are counting, serve them first and recount that spot. A moving shelf gives a moving number.

Counting is only half a stock take. The count tells you what you have. You also need what you are supposed to have, and that is the job of your point of sale software. One clarification for Nigeria: not the little card machine everyone calls a POS, but the system that records your sales.

Good software keeps a running figure for every product, called theoretical stock. It starts from what you received, subtracts each sale as it rings up, and adjusts for returns. At any moment digabloPos can tell you it believes there are 37 cartons of a drink on your shelf. You go and count 34. Those three missing cartons are your variance, and finding it is the entire point of the exercise.

Count without that reference and you are counting into a void. You know you have 34, but 34 compared with what? Paper alone gives you numbers. The comparison gives you information.

Here is the trap that catches people counting with the doors open. You count a shelf at 10am, you sit down to compare at noon, and six of those items have sold in between. The figures disagree, you smell theft, and all that actually happened is two hours of normal trading.

The fix is to freeze the reference at the moment you count. Note the theoretical stock at 10am, count at 10am, compare those two numbers and no others. Whatever sells afterwards belongs to the next cycle. Software that timestamps counts does this for you. On paper it means writing the time on the sheet and reconciling straight away, not after lunch. Skip the freeze and every open-door count drowns in false alarms until you stop trusting the whole method.

The slowest version of a count is one person squinting at a shelf calling out numbers while another writes them down. A barcode scanner removes most of that. Point, beep, the product identifies itself, you key in the quantity. No hunting through a list, and no mixing up the 50cl with the 75cl, which is a classic source of phantom shortages: one size shows missing, its sibling shows surplus, and you spend an evening chasing a thief who does not exist. A basic Bluetooth scanner costs less than the errors it prevents in its first month. If your goods carry barcodes, scan your counts.

Who counts matters as much as how. The rule is simple: the person who manages a section should not be the only one who counts it. If your storekeeper runs the drinks and also counts the drinks, any shortage he caused, whether by helping himself or by sloppiness, is a shortage he gets to quietly smooth over on the sheet.

Rotate the counters so a fresh pair of eyes covers each area, or let one person count while a second spot-checks the expensive lines. This is not about treating your staff as suspects. It protects the honest ones, because when the numbers are clean nobody can whisper that they were cooked. A system that logs each stock adjustment with the name of the employee who made it tightens things further. People count more carefully when their name sits next to the figure.

The count says 34, the system says 37. Now what? Resist the reflex to shout theft, and resist the lazier reflex to silently correct the number and move on. Both destroy the value of the count.

Investigate first. Three missing cartons can be a supplier who delivered short while the full invoice got signed. Breakage nobody reported. Expired stock thrown out without a record. Two similar products counted into each other. Or yes, sometimes, someone helping themselves. Each cause has a different fix, and calling everything theft poisons the shop while the real leak keeps dripping.

Check the delivery notes. Ask who was on shift. Look at the wastage record if you keep one, and start keeping one if you do not. Only then adjust the stock, and write the reason down with the adjustment. After a few cycles those written reasons become a pattern you can act on: one supplier who repeatedly delivers short, one shelf that always comes up light on Fridays, one product that seems to walk on its own. A vague suspicion becomes a specific problem with a name.

There is no single answer, which is exactly why the ABC ranking exists. A items weekly, B items monthly, C items quarterly is a rhythm a small shop can hold without strain: one or two sections a day, fifteen minutes each, and the rotation takes care of itself.

Run that for a year and you will have counted your fast movers around fifty times while the shop across the road counted everything once. That difference in frequency, not some expensive machine, is the entire advantage.

You can start all of this on paper today. Print a sheet per section with product names and empty boxes, count, then compare against your records by hand. A paper cycle count beats an annual shutdown every single time, and if paper is what you have, do not wait for anything better.

But know where paper stops. It cannot tell you the theoretical stock, so someone has to look it up line by line. It does not timestamp anything, so the freeze depends on discipline you will sometimes forget at 6pm. It does not log who counted or who changed a figure. And it will never warn you that the same drink has come up short three cycles running.

An app closes those gaps: theoretical stock on the screen, scan to count, timestamped snapshots, adjustments tied to an employee name, variance history over time. digabloPos does all of this and keeps working when the network drops or NEPA takes the light mid-count; the numbers sync once you are back online. Paper gets you started. Software makes the habit light enough that you actually keep it.

Do not reorganise the whole shop in one go. Here is a first month any shop can run.

Week one: pull the sales report and rank your products A, B and C. You are not counting anything yet, just finding out which lines pay the rent.

Week two: put a rotation on the calendar, weighted so A items come round weekly, and count your first one or two sections in the quiet hours. Compare each count against the theoretical stock and write down every gap, however small.

Week three: add the barcode scanner if your goods are barcoded, and stop silently correcting figures. From now on every adjustment carries a written reason.

Week four: sit down with the variance list and look for repeats. A product, a shelf or a supplier that keeps coming up short is your first real investigation, and it is usually worth more money than the counting cost you.

By the end of the month you will have counted the entire shop without hanging a single sign on the door, and you will almost certainly have found one leak you did not know about. The Sunday shutdown never gave you that. A quiet weekly habit does.