The best POS for a small shop in Pakistan: how to actually choose

What should a small shop in Pakistan look for in a POS?

For a small shop in Pakistan, five things decide whether a POS works: does it keep billing when the internet drops, does it handle udhaar as a normal part of a sale, can it produce a proper tax invoice for your province when a customer asks, can a new cashier learn it in a day, and can you get your data out if you leave. Everything else — the feature lists, the dashboards, the screenshots — is secondary, because a shop that can't take money during a signal drop has an expensive problem no report will fix.

Below is each of those five, what "good" looks like, and how to test it before you pay.

1. Does it work offline?

This is the first filter, and it eliminates a lot of otherwise good software. Ask the vendor plainly: what happens to a sale when the internet is gone? If the answer is "you wait for the connection," the tool was built for an office, not a shop. What you want is a POS where prices, stock and sales live on the device, so ringing up a customer never needs a round-trip to a server, and the day's sales sync themselves when the signal returns.

The test: install it on the device you'll really use, in the shop, on your real connection. Turn airplane mode on. Ring up three sales — one on udhaar — and print a receipt. Turn the connection back on. If all three are there and correct, it passes.

2. Does it handle udhaar properly?

Credit is half the business in most Pakistani shops, and a POS that treats it as an afterthought pushes you back to the paper khata within a month. "Properly" means specific things:

• You can complete a sale as udhaar in the same flow as cash — not through a separate module you have to remember.
• Each customer has a running balance that updates itself on every sale and every payment, with no manual re-totalling.
• Every entry is date-stamped automatically, so an old debt visibly looks old.
• You can see who owes the most and who has owed the longest, because that list is exactly who's worth a reminder this week.

If any of that requires a second app or a spreadsheet on the side, you now have two ledgers, and two ledgers means neither is trusted.

3. Can it produce a tax invoice for your province?

Most small shops run day-to-day on informal receipts, and that's fine. But when a customer, a corporate buyer or an auditor asks for a proper sales-tax invoice, you need one — and the sales tax that applies in Pakistan depends on whether you sell goods or services and which province you're in, because the provinces run their own authorities. A POS should let you issue both a quick receipt and a proper tax invoice from the same sale, without keeping two registers.

If you rarely or never issue a tax invoice, don't over-weight this. If you sell to businesses, weight it heavily — it's the criterion most likely to be discovered too late.

4. Can a cashier learn it in a day?

This is the criterion most people underrate and most regret. Staff turnover is real. If a new person needs a week to become useful, that's a recurring cost every time somebody leaves, and it's paid in slow evenings and wrong bills. Judge the till screen, not the admin dashboard: how many taps to sell three items and take cash, and can somebody who isn't comfortable with computers do it without you standing there?

The test: don't demo it yourself. Hand it to whoever will actually be on the counter, say nothing, and watch. Where they hesitate is where you'll lose time every single day.

5. What happens to your data if you leave?

Ask before you start, not after. Can you export your products, customers and sales in a normal file? Is your data backed up somewhere other than the shop's computer? Software you can't leave isn't cheaper — it just charges you later, in a currency you can't budget for.

Who should NOT buy NexusTill

Being useful here means saying this plainly. There are shops that should keep their money:

• A single-owner shop with a small, mostly-cash turnover, a handful of udhaar customers you settle weekly, and stock you can see from where you stand. A notebook and a calculator genuinely work. Software adds a monthly cost and a login for problems you don't have.
• A shop with no reliable power at all. Offline billing survives a dead internet connection, but nothing survives a dead device. If you can't keep a phone or laptop charged through your trading hours, sort the power out first — a UPS or a power bank is a better first purchase than any software.
• A shop that only needs the features on a higher tier but can only afford the lower one. If you need multi-branch or stock transfers, those are on Scale — buying Starter and hoping won't work. Restaurant mode with recipes, food cost, tables and the kitchen display starts on Growth, as does pharmacy batch-and-expiry, and the AI assistant isn't on Starter at all. Price the tier that has what you need, and if that number doesn't work for the shop today, wait.
• Anyone who wants a system somebody else runs for them. NexusTill is software you use, not a bookkeeping service. If what you actually want is a person to keep your accounts, hire that person.

The shops NexusTill fits are the ones with a real inventory to keep straight, udhaar that runs longer than a few days, more than one person touching the till, or a second shop on the horizon. If none of that is you yet, come back when it is.

A short buying checklist

1. 1Write down your three actual problems — not features. "I don't know what I'm owed." "Stock runs out without warning." "I can't tell if the evening shift is short."
2. 2Filter to software that works offline. That's a hard yes/no, not a trade-off.
3. 3Check udhaar is in the sale flow, not a bolt-on.
4. 4Check the tier that has what you need, and confirm that price works — not the tier you wish you needed.
5. 5Test it in the shop for a week, on your connection, with your staff.
6. 6Confirm you can export your own data before you commit.

The best POS for your shop is the boring one your staff stops noticing. If they're still thinking about the software after two weeks, it's the wrong software.

Written by The NexusTill team.