Is an M-Pesa Till Enough, or Does Your Business Need a POS?

It’s past nine on a Saturday and you’re closing the duka. Your phone shows a good day: KSh 23,000 came in through the till, plus whatever cash is in the drawer. Then the questions start. Which products actually sold? Why did the cooking oil run out at 3 p.m. while nobody told you? The girl who covered the afternoon shift, how much did she sell, and does the cash she left match anything? The till number cannot answer a single one of those questions. It moved the money, full stop.

This is the confusion at the heart of “M-Pesa till vs POS”. Plenty of owners believe that because payments come in smoothly, the business is under control. It isn’t. A till and a POS do two different jobs, and this article will be direct about it: the till is excellent at collecting money, and past a certain size, it is nowhere near enough on its own.

A till is a channel for receiving business payments. The common one is Buy Goods: the customer opens Lipa na M-Pesa, enters your till number, and the money lands in your business account within seconds, with a confirmation on both phones. For smaller traders, Safaricom offers Pochi la Biashara, which keeps business money on your phone but separate from your personal M-Pesa, so a customer paying for tomatoes doesn’t send to the same wallet your cousin borrows from.

Getting a till is straightforward, and it immediately makes you look more serious than telling customers to “send to my personal number”. If you take payments in Kenya, you should have one. That part is not up for debate.

Credit where it’s due. Payment clears in seconds, even on a weak network. The money skips the cash drawer entirely, which means less to count, less to lose, and less to tempt anyone. Customers trust it completely; nobody in Kenya hesitates at “Lipa na M-Pesa”. And your business money stays separate from your personal money, which your accountant, or your future self, will thank you for.

Now the other side. The till knows that KSh 500 arrived at 4:17 p.m. It does not know that the 500 was two loaves, a packet of milk and airtime. It has no idea how many bottles of soda are left on your shelf, so it will never warn you that you’re about to run out of the thing that sells fastest. It cannot say which employee served the customer, because as far as the till is concerned, there are no employees, only money arriving. It cannot print an itemised receipt for the customer who asks for one, and an M-Pesa confirmation SMS is not a KRA receipt, which matters more every year now that eTIMS is spreading.

Think of the till as a pipe. Water flows in, and you can see how much flowed. What you cannot see is where each litre came from, what it was for, or who turned the tap. A shop run on a pipe alone is a shop run blind.

Here is the position, plainly. If you sell mandazi from a basket at the stage, one product, no staff, no shelf, a Pochi la Biashara is genuinely all you need. Enjoy the simplicity.

The moment you carry real stock, even a small duka’s worth of bread, unga, sugar, oil and sodas, or the moment one person other than you touches the money, the till stops being enough. Not “might stop”. Stops. You now have shillings sleeping on shelves and hands you don’t watch all day, and the till tracks neither. What you need at that point is a POS: the system that records each sale item by item, cuts the stock automatically, ties every transaction to a person, and shows you at the end of the day whether the shop made money or just moved it around.

Ask any owner who finally counted properly: the losses were never where they expected. A POS reduces stock with every sale, so at any moment you know what should be on the shelf. When bread or cooking oil drops below the level you set, it alerts you before you run out, not after the customer has crossed the road to the competitor. It also shows the opposite problem, the slow movers: those cartons that have sat since March, quietly tying up cash you could have put into fast sellers.

Then there is shrinkage, the polite word for stock that walks away. Compare what the POS says you should have against what you physically count, and the gap has a size and a date. Without that comparison you don’t even know you’re losing, let alone how much. On a till alone, none of this exists. The till saw the money that arrived; it never saw the goods that left.

The day you hire your first employee, your questions change. It’s no longer “how much came in?” but “how much came in while I wasn’t there, and does it match what left the shelf?”

A POS answers with PINs. Each person taps their code before selling, so every sale, discount and cancellation carries a name. You decide who can do what: the attendant sells, but only you cancel a sale or change a price. When the evening count is short, you don’t interrogate everyone; you open the day’s history and look. That protects your money, and it protects your honest staff too, because the record clears them as surely as it exposes anyone else.

Once you run a POS next to your till, you gain the one check that till-only businesses can never do. In the POS, M-Pesa sales are recorded separately from cash sales. Each evening, you put two numbers side by side: the M-Pesa total your POS recorded, and the money that actually landed on your till, from your statement or the official Safaricom messages.

The two should match to the shilling. When they don’t, something specific happened today: a sale nobody rang up, a payment that never actually arrived, or someone helping themselves. Because you check daily, the damage is one day’s worth, not one month’s. The whole exercise takes five minutes with a POS that shows totals per payment method, and it is the single habit that separates owners who control their money from owners who hope.

It still catches people every week. A customer, usually at rush hour, flashes their phone: “imeingia, see the message.” The SMS is fake, cooked up by an app, and the money never left their account. By the time you check your till that evening, they’re long gone with the goods.

The rule costs nothing: never release goods based on a message shown on the customer’s phone. Confirm on your own side first, on your till phone or your statement. Train whoever works your counter to hold that line even when the queue is long, because the fraudsters deliberately pick the busy moments. And your daily reconciliation is the backstop: a “paid” sale that never reached the till shows up as a gap that same evening.

KRA has been pushing businesses onto eTIMS, its electronic invoicing system, and the net keeps widening. The receipt it wants is a proper tax receipt, itemised and transmitted; an M-Pesa confirmation SMS does not qualify and never will.

A POS keeps the clean, itemised sales history that makes compliance manageable instead of terrifying: what sold, when, for how much, receipt by receipt. When your accountant or a KRA officer asks, you print a report instead of reconstructing a year from till statements and memory. Your exact obligations depend on your registration and turnover, so confirm them with KRA or your accountant. But whichever side of the threshold you’re on today, tidy records cost you nothing and messy ones eventually cost plenty.

Forget dashboards with twenty charts. Four numbers change how you run a small business: what sold today, what should be in the drawer and on the till tonight, what is about to run out, and which ten products bring most of the money. The last one surprises almost everyone the first time; the products you think carry the shop and the ones that actually do are rarely the same list.

A POS hands you these numbers every evening without you lifting a pen. Buy stock based on real sales instead of feeling, put the fast movers where hands reach them, drop the line that hasn’t moved since Easter. The till, by design, can contribute exactly one number to this picture: how much M-Pesa arrived. Useful, and nowhere near sufficient.

M-Pesa itself is impressively resilient, but apps that need a live internet connection for every sale are not. Anyone who has traded through a fibre cut or a mast outage knows the feeling of a queue forming while a spinner spins.

So make offline capability a hard requirement, not a nice-to-have. A POS built offline-first, which is how digabloPos works, records sales and stock movements on the device itself with no connection at all, then syncs everything when the network returns, losing nothing in between. Your customer pays the till as usual, your record survives whatever Safaricom or your ISP is doing that afternoon. Keep the phone charged through the power cuts and the shop simply keeps trading.

Less than most owners assume. The till side you already know: Buy Goods transaction charges are borne per payment and Pochi is cheap to run. The POS side needs a device and software, and the device is almost certainly the Android phone already in your pocket. A tablet is a comfort upgrade, not a requirement.

For software, be careful with subscriptions of KSh 3,000 to 5,000 per month charged from day one; over a year that is the price of a decent phone, paid before you’ve proven the system earns its keep. The saner path is to start free and pay only when a specific need shows up. digabloPos follows that model: selling, stock tracking, staff PINs and daily reports cost nothing, and optional modules run around KSh 1,300 to 2,000 per month each, cancellable anytime. If you want printed receipts, a thermal printer goes for roughly KSh 4,000 to 8,000 in Nairobi, a one-off purchase. In short: a duka can run a full till-plus-POS setup for the cost of a printer and otherwise zero shillings a month.

No need for a grand migration. Here is a plan that fits into one week alongside normal trading. Monday: install a free POS on your phone and enter your twenty or thirty fastest movers with their prices. Skip the rest for now. Tuesday to Friday: ring up every sale, choosing “M-Pesa” or “cash” as the payment method, exactly as the customer pays. Keep your exercise book going in parallel if it reassures you. Every evening: do the five-minute reconciliation, POS M-Pesa total against till statement, POS cash total against the drawer.

By Sunday you will know your real daily sales, your real best-sellers and whether your counts match, most likely for the first time. From there, add the rest of the stock shelf by shelf, give each staff member a PIN, and set low-stock alerts on the products that hurt most when they run out. Your till keeps doing what it does best, collecting the money. The POS does everything the till was never built to do.