4 minute read
Choosing a point of sale system: what actually matters
Most POS comparisons list features nobody uses. Four things decide whether a till works in a Kenyan shop, and only one of them is on the brochure.
Every point of sale system claims the same features. Here is what actually separates the ones that work from the ones that get abandoned after four months.
## It has to work when the internet does not
This is first for a reason. A till that stops when the line drops is not a till; it is a liability that appears at the worst possible moment, which is when the shop is busy.
Ask the vendor directly: what happens when the connection goes? The answer you want is that sales are recorded on the device and sync when the connection returns. Any other answer, including a confident "our uptime is excellent", means no.
## Recording a sale has to be faster than not recording it
Every system that fails, fails here. If ringing up a sale properly takes longer than writing it in a book, staff will write it in a book, and within a month your stock figures are fiction.
This is what a barcode scanner actually buys you. It is not about accuracy; it is about making the correct action the quick one. Watch somebody complete a sale on the system before you buy. If it takes more than a few seconds for a normal item, keep looking.
## Stock has to fall as things sell
A point of sale that records sales but does not touch stock levels is a cash register with extra steps. The reason to computerise the counter is that it gives you stock control for free: you find out what is running low without counting, and what is not selling without guessing.
If stock is a separate module you have to update by hand, you have bought two jobs instead of one.
## It has to raise an eTIMS invoice
From the till, once. The alternative is somebody re-typing the day's invoices into a second system every evening, and that person will stop doing it.
## What matters much less than you would think
The number of reports. You will use three of them.
Whether it is cloud or installed locally. What matters is what happens when the connection drops, which is the first question, not this one.
Loyalty programmes, promotions engines, and the rest of the feature list. Buy them when you need them.
## On hardware
Buy a receipt printer rated for the volume you actually have. The cheap thermal printers are fine for a low counter and wear out quickly on a busy one, and the difference shows up after a year rather than on day one.
Buy a 2D scanner rather than a 1D laser if there is any chance you will ever scan a code off a customer's phone. The price difference is small and the capability is not retrofittable.
And put the till on a UPS. Not to trade through a blackout (fifteen minutes will not do that) but so a power cut in the middle of a transaction does not corrupt the day's data.
## Who looks after it afterwards
A till is not a purchase you make once. Prices change, staff come and go, a printer jams on a Saturday, and a tax rule changes with a few weeks of notice. Before you buy, ask who you call when something stops working, what hours they answer, and whether they can connect remotely or have to visit.
Ask too what happens to your data if you stop paying the subscription. You should be able to take your sales history and your product list away in a spreadsheet. A system that holds your records hostage is expensive however cheap it was to start.
## The test to apply
Ask to watch a shop of about your size use the system for an hour, with real customers. Not a demonstration: a shop. Ten minutes of that tells you more than every brochure.