How to choose a POS system in Uganda

What makes choosing a POS in Uganda different?

A point-of-sale (POS) system is the software and hardware that rings up a sale, takes payment, prints a receipt and updates stock. Choosing one in Uganda is different from choosing one abroad for three reasons the glossy feature lists rarely address: the internet is not reliable, most customers pay by mobile money, and much of a shop’s trade runs on credit written in a notebook.

So the right question is not “which POS has the most features”. It is “which POS still works on the afternoon the network is down, the customer is paying by MTN, and they want to add it to what they already owe”. The seven questions below are ordered by how often they decide whether the software survives contact with a real counter.

What should I compare POS systems on?

Feature lists are written to be compared favourably, so compare on the answers a shop actually lives with instead. These are the seven checks this guide walks through, and what a good answer looks like for a Ugandan counter.

Does it keep working when the internet goes down?

This is the first filter, and it eliminates most of the field. A cloud-only till stops selling the moment the connection drops, and in most Ugandan trading centres it will. Look for a POS that is offline-first: the till keeps a full copy of your catalogue, prices and customers on the machine itself and sells at full speed with no connection, syncing later.

Be specific when you ask. “Works offline” can mean anything from a genuine local database to a browser tab that white-screens after five minutes. Ask exactly what keeps working offline (selling, receipts, credit, repayments, reports) and what needs a connection. An honest vendor will tell you where the line is.

Does it take mobile money properly?

Most Ugandan customers pay by MTN Mobile Money or Airtel Money. A POS should record mobile money as its own payment type, separately from cash, so your end-of-day totals show how much came in by each method and your drawer count actually balances.

Watch for two traps. First, a till that only knows “cash” and “card” forces you to log every MoMo sale as cash, which makes reconciliation impossible. Second, be clear-eyed about what “mobile money integration” means: very few tills confirm the transaction with MTN or Airtel automatically. Most, sensibly, record the tender and let you confirm the payment on your own line. That is fine; just know which one you are buying.

Can it handle the credit book?

In a Ugandan shop, credit is not an edge case. It is a large share of trade, and today it usually lives in a paper notebook that never quite collects. A POS worth buying treats credit as a first-class payment method: it records who owes what, lets a customer pay part now and the rest on credit in the same sale, and keeps a running balance you can act on.

Ask whether the credit book works offline too. If it only updates when the till is online, it is the notebook with extra steps.

Does it actually file EFRIS, or just say it does?

If you are EFRIS-registered with Uganda Revenue Authority, this is the claim to scrutinise hardest, because it is the easiest to fake on a feature list. A settings screen for your TIN and fiscal device is not the same as software that transmits a completed sale to URA’s EFRIS system and stores the real fiscal number (FDN) that comes back.

Ask for proof, not a checkbox: “show me a sale going through and the FDN URA returned.” Some products, Stem included today, have the EFRIS settings and the surrounding structure but do not yet transmit sales to URA. That is a legitimate roadmap position; being misled into thinking you are filing when you are not is how you end up with a URA penalty. Insist on a live demonstration before you rely on any POS for EFRIS compliance.

What hardware will you actually need?

A POS should run on hardware you can buy locally and already half-own. At minimum that means any Windows laptop or a browser on a device you have. Beyond that, the common additions are a thermal receipt printer (the standard 80mm ESC/POS ones), a barcode scanner and a cash drawer.

Avoid software that only runs on proprietary, imported terminals. The hardware lock-in costs more than the software ever will, and a replacement after a lightning storm can take weeks. Standard USB and network peripherals are the safe bet.

How is it priced, and in what currency?

The price model matters as much as the number. For a Ugandan shop, the right shape is a free or low entry tier, monthly billing, and payment by mobile money, rather than a large annual licence payable only by card or bank transfer. A free tier also lets you test the software on your real stock before you commit a shilling.

Add up the true cost: software plan, any per-device or per-user fees, hardware, and support. A till advertised as “cheap” that charges per device across three counters is not cheap.

One shop today, or branches tomorrow?

If you might open a second location, check how the POS handles it before you are locked in. The things that matter are per-branch stock (so each shop’s count is its own), verified transfers between branches (so stock moving between them is accounted for, not just deducted and hoped for), and reporting you can read per branch and across the business.

A single-shop till bolted into a multi-branch setup later is painful. If growth is plausible, buy the one that already thinks in branches.