How to run a POS trial that actually tells you something

A free trial is the cheapest information you will ever get about a POS, and most merchants waste it. They install it, click around on sample data for twenty minutes, and conclude that it "seems fine".

Here is a protocol that produces an actual answer in five days.

Day 1: real data, nothing else

Import your real catalogue. Not a sample — the actual file, with its inconsistencies.

Then measure three things:

1. How long did the import take, including your time?
2. How many rows failed, and did the system tell you which and why?
3. Spot-check twenty products: are prices, costs, VAT treatment and barcodes right?

A system that silently drops rows has already told you something important. A system that reports exactly what it could not match has told you something better.

Also on day 1: create a user for each staff member with real permissions, and configure your receipt template with your actual logo and legal name.

Day 2: the quiet shift

Run real sales on one terminal during your quietest period, with your normal staff.

Watch for:

• Taps per sale on your three most common items
• Where staff hesitate
• Whether anything requires a manager who is not there

Print a receipt and scan the QR with a phone. See theQR guide.

At the end, close the shift and reconcile the drawer. Compare the report to your existing system's for the same hours.

Day 3: the awkward transactions

Work through the list you wrote inhow to choose a POS — your five hardest transactions. Typically:

• A bill split three ways with one cash payer
• A return of one item from a multi-item sale two days old
• A B2B sale needing a standard tax invoice with a VAT number
• A mixed-VAT basket with a percentage discount
• An exchange for a different size

Time each one and note the number of steps. If any of them is impossible, you have your answer already.

Day 4: the outage

Pick a moment you can afford andunplug the router for thirty minutes during real trading.

• Can you complete sales?
• Does a signed, scannable ZATCA invoice print?
• Can you take a card payment?
• What does the screen tell staff?

Reconnect and verify every transaction arrived, once each. See theoffline comparison.

If the system fails this and you cannot tolerate that, stop the trial here. Nothing later will compensate.

Day 5: the peak shift and the exit

Run the busiest shift you have on the trial system, with a manager present. This is the real test: a system that is pleasant at 11am and painful at 9pm is a painful system.

Then, at the end of the day:

1. Run every report you care about — sales by product, by hour, by cashier; gross margin; stock on hand
2. Compare them to your current system for the same period. Differences are findings.
3. Export everything and open the files. This is the single best predictor of whether you will be able to leave later.

What to decide on

Not whether you liked it. Three concrete things:

1. Did your five hardest transactions work, in an acceptable number of steps?
2. Did it keep trading during the outage?
3. Could you get all your data out at the end?

Three yeses is a system you can buy. A no on number 2 or 3 is a structural problem that no feature compensates for.

Trial hygiene

• Do not run it as your only till during the trial. One terminal in parallel.
• Do not let the vendor set up your data. You need to know how hard it is, because you will be doing it for the next three years.
• Do involve your slowest staff member. They are the real user.
• Do write things down daily. By day 5 you will not remember day 2.
• Do ask support a real question during the trial, at a realistic hour, and time the response. This is the only chance you get to test support before you depend on it.

That last one is underrated. How a vendor answers a trial-user question at 8pm on a Thursday is a very good predictor of how they will answer a customer question at 8pm on a Thursday.