How to choose a restaurant POS

Buying a point-of-sale system is usually a three-to-five-year decision. It touches every order, every payment, every shift and your menu data, and switching later means retraining staff and migrating data. A common way it goes wrong is in the first meeting: a vendor shows a polished order screen, everyone compares the monthly subscription, and nobody tests the parts of service that actually break.

This guide is a process you can follow in order. It works for a single counter-service shop or a multi-location group, because the steps stay the same even though the answers change.

1. Write down how your restaurant runs before you talk to anyone

Spend an hour with your manager and your busiest server or cashier. List what happens on your worst night, not your average one:

• How orders come in. Counter, table, bar, phone, your own website, delivery apps, catering.
• Your hardest menu items. Pick the five that cause the most mistakes: half-and-half pizzas, combination plates, build-your-own bowls, market-price seafood, drinks with three modifiers.
• How tickets reach the kitchen. Which stations, printers or screens, and what has to be fired together.
• How guests pay. Split checks, tips on the card, tabs, gift cards, house accounts, cash.
• What you count at close. Cash drawer, tips, comps and voids by manager, sales by category.
• What you can't live without. These are your veto items. More on those below.

This list becomes your demo script. Without it, every demo shows the vendor's best case instead of yours.

2. Decide what you are really comparing

A restaurant POS is several purchases bundled together. Separate them before you compare prices:

The full cost picture over three years is in what a restaurant POS really costs.

3. Build a shortlist of three

Three vendors is enough to see real differences without losing a month to demos. Pick systems that publish support for your restaurant type, check whether your must-haves are on the plan you can afford, and drop any vendor that won't tell you its processing terms in writing before a demo. Our comparison of restaurant POS systems groups the main vendors by the kind of restaurant they fit, and the guide for restaurants under 50 seats covers the cheaper end.

4. Run the same six tests in every demo

Ask each vendor to run these with your own menu, on the hardware you would actually use. A system that can't run your worst night in a demo won't get there after you sign.

1. Menu test. Build your five hardest menu items with their real modifiers, sizes, taxes, discounts and kitchen routing, then ring them up.
2. Payment test. Run a split check, a partial refund, a void after payment, a tip adjustment, a gift card and a declined card.
3. Offline test. Disconnect the internet in the middle of taking an order. Note exactly what keeps working, what stops, and what happens to card payments taken while offline.
4. Close test. Ask for yesterday's sales by category, labor as a share of sales, comps and voids by manager, and the cash drawer count.
5. Support test. Call the support line during your real Friday-night hours before you sign. Ask how outages are handled, not how sales questions are.
6. Exit test. Ask for sample exports of your items, customers, sales, gift cards and loyalty balances. If you can't get your data out, the vendor owns more of your restaurant than it should.

5. Score each system on the same criteria

Use the weights below, which are the same ones we use in every comparison on this site (see how we compare systems). Give each vendor a score from 1 to 5 on each line, multiply by the weight, and add them up. Change the weights if your restaurant is different. A food truck might weigh reliability higher, a group of five locations might weigh reporting higher. But decide the weights before the demos, not after.

A vendor scores 4 on fit (25%), 3 on cost (25%), 2 on payment terms (15%), 4 on reliability (15%), 5 on support (10%) and 3 on data portability (10%). Its weighted score is 1.00 + 0.75 + 0.30 + 0.60 + 0.50 + 0.30 = 3.45 out of 5.

Veto items beat averages

Some failures can't be averaged away. If you run a bar, weak tab handling is disqualifying no matter how good the reports are. If most of your sales come through delivery, a system that can't take delivery orders into the kitchen without retyping is a no. If you run several stores, no central menu control means every new item is entered by hand at each location. Mark these as veto items on your list from step 1, and drop any system that fails one.

6. Read the contract against your scores

Every promise that earned a vendor points should appear in writing: offline behavior, support hours, hardware ownership, processing rates and fees, data export, the setup timeline and the cancellation terms. If the contract is weaker than the demo, lower the score before you sign, not after.

Two questions that are easy to forget:

• If you change card processors later, does the hardware keep working, and what does it cost to switch?
• When the contract ends or you cancel, what happens to stored customer cards, gift card balances and loyalty points?

Card processing is the line that compounds. For the arithmetic, see the example below.

A restaurant takes $120,000 a month in card payments. A rate that is 0.30 percentage points higher costs $360 a month, $4,320 a year, or $12,960 over a three-year agreement, before any monthly fees.

7. Plan the switch before you sign

Pick a go-live date in a slow week, not before a holiday. Ask the vendor who builds the menu, who trains staff and for how long, and whether someone will be on call for your first weekend. Run the old and new systems side by side for a shift if you can. Card security is part of the switch too: the PCI Data Security Standard applies to merchants and every other business that stores, processes or transmits cardholder data, so ask the vendor what it covers and what remains your responsibility.