WTR PRO AMD, ST13-U and ST11-U: A Five-Year View of Store Hardware

Store hardware is judged twice: once on the invoice, and again over the years that follow, when most of the money is actually spent. The two judgements rarely agree, and the gap between them is what this piece is about. Three units cover the work in a typical store. The ST13-U lives at the service desk where a transaction has to be reviewed on a screen large enough for two people to read. The ST11-U is carried around the shop floor by whoever is checking shelves. The fixed compute behind both is our rugged mini PC solutions range, represented here by the WTR PRO AMD. Scanning and payment peripherals that sit alongside them are collected under retail POS terminals.

Rather than compare specifications, this is a walk through the life of a store installation, in the order the costs actually arrive.

The Bill Arrives Before the First Sale

The first expenditure at a new site happens before a single transaction is taken, and it is almost entirely installation. Counter cut-outs, cable runs, power points, network drops, mounting brackets and the labour to fit all of it. The shape of the equipment decides how large that bill is, because a small sealed computer can be attached to the inside of an existing counter while a full-size machine needs a cabinet, a vented enclosure and somebody to plan where the heat will go. A large service desk screen needs a mount, but it also removes the need for a customer-facing second display, which is a saving that shows up in the same week. Installations that run late are almost never late because of software; they are late because a bracket did not fit or a cable route was not planned.

The desk screen has to be reviewed with a customer standing alongside, which is a job no carried device performs well.

What the Opening Weeks Reveal

The fortnight after opening is when a store learns whether its equipment will actually be used. A carried device is either picked up as a habit or left on a shelf, and the deciding factor is usually physical rather than technical. If it is awkward to hold in one hand while the other hand reaches for a box, staff will write the question down and deal with it at the desk, which quietly returns the store to the process the device was bought to ｒｅｐｌａｃｅ. This is why the carried unit should be chosen on weight and grip before anything else, and why the desk unit should be chosen for legibility rather than portability. Both decisions look trivial on a specification comparison and both determine usage rates that no software update can recover.

The carried unit is either picked up as a habit or left on a shelf within a fortnight, and weight decides which.

Peak Concurrency Sets the Ceiling

Fleet size is normally calculated during a calm week and then tested during a busy one. The figure that matters is concurrency at the peak hour: how many positions must be live at the same moment, how many staff are on the floor, and how many units are unavailable because they are charging or awaiting repair. A store with exactly as many units as it has positions has no cover at all, and the buffer that looks like overspending in a quiet month is the reason a busy month does not become a complaint. Floor devices follow the shift pattern rather than the trading pattern, so their count should be taken across a full week instead of a single day.

When a Counter Goes Dark

A failed unit does not merely lose the transactions it would have handled. It pushes a queue into the next lane, and a customer who has queued once is not inclined to queue twice, so the loss is larger than the arithmetic of one position suggests. The measurement that reflects this is available minutes per position per week, which is a different number from mean time between failures and much closer to what a store manager experiences. It also explains why passive cooling earns its place where dust is unavoidable: a machine with no intake filter and no fan bearing has removed the components that fail in the second and third years, which is exactly when attention has moved on to newer sites. Management from a distance matters for the same reason, since a unit that can be restarted without a site visit stops generating a whole class of journeys rather than postponing them.

The store controller is deliberately unremarkable and out of reach, which is where a machine that must never be noticed belongs.

The Quiet Costs of the Second Winter

By the second year the equipment is no longer new and the costs become steady rather than dramatic. Dust accumulates wherever cardboard is handled, battery capacity in carried units quietly declines, charging docks multiply, and the spare parts cupboard fills with whatever the previous generation used. None of these items is large on its own, and together they form a fixed overhead that a chain can only reduce by owning fewer distinct types of hardware. Standardising on one small form factor is therefore not an aesthetic preference; it is the difference between one bracket, one power supply and one set of staff knowledge, and three of each.

The Swap That Eventually Happens

Every installation is eventually replaced, and the replacement is the most expensive day in its life because trade stops while it happens. A store built on easily remounted units can be converted in a short slot, ideally outside trading hours, where a store built on cabinets and bespoke mounting needs a full day of rebuilding and testing. The decision that determines which of those two days a chain experiences is taken years earlier, when the mounting standard was chosen. The same decision also settles whether retired units are kept as spares or moved on, and teams that leave that until the swap weekend generally regret it.

A Ledger, Not a Purchase Order

Read the fourth column and the pattern is that the largest numbers are the recurring ones, which is why a slightly higher purchase price is often the cheaper decision. Read the second column and it becomes clear that two of the five stages have nothing to do with hardware performance at all.

Recording Enough to Repeat It

The purpose of recording anything is to make the second site easier than the first. Five items are worth keeping from a pilot store: the number of hours it took to open, the proportion of floor tasks completed away from the desk, unavailable minutes per position per week, support calls per store per month, and the hours taken for the first hardware swap. A chain that holds those five can open the next store with a defensible estimate instead of an optimistic one, and can tell within a month whether a change of equipment made any difference. A chain that holds none of them repeats the first site indefinitely, which is why the second rollout is so often harder to fund than the first.