
Key takeaways
- A POS is the front line of your restaurant. Choose it for speed at the till and reliability in your worst hour, not for feature count.
- Offline mode is non-negotiable if your Wi-Fi is ever unreliable. Test it by switching off the router during the demo.
- Inventory with recipes turns sales into real food-cost control.
- Multi-branch owners need central menus, prices and reports with per-branch overrides.
- Compare total cost over three years (fees, hardware, payments, add-ons), and build custom only when products cannot meet your integrations or workflow.
What a restaurant POS must do
A restaurant POS takes orders, sends them to the kitchen, accepts payment and records everything for reporting. Modern systems also manage inventory, staff, loyalty and delivery channels. Before comparing brands, write down how an order flows in your restaurant (dine-in, takeaway, delivery), because the best system is the one that fits that flow with the fewest taps.
Seven features to check before you buy
1. Offline mode
Wi-Fi fails at the busiest hour. A POS built offline-first stores orders and payments locally and synchronises later, so service never stops. Test it: ask the vendor to switch off the router, take three orders, print the kitchen tickets and close a cash bill. Then restore the connection and check that nothing was lost or duplicated.
2. Speed at the till
Count the taps for your five most common orders. Modifiers, split bills and table transfers should take seconds. Slow tills create queues and tired staff. Test it: give a new waiter the device with no training and time how long it takes to enter a burger with two modifiers.
3. Kitchen display and order routing
Orders should route to the right station, such as grill, drinks or pastry, with timers and status. This replaces paper tickets and shouting across the pass. Test it: send an order with items for two stations and confirm each station sees only its own items, and that a late item turns visibly overdue.
4. Inventory and recipes
Link menu items to ingredients so every sale deducts stock. You get real food-cost percentages, low-stock alerts and wastage reports instead of month-end guesses. Test it: sell a dish, then check that the exact ingredient quantities fell. Our guide to restaurant food cost control shows how to use these numbers.
5. Multi-branch management
Owners need central control of menus, prices and roles, with per-branch overrides plus consolidated and per-branch reports. Test it: change a price centrally and one branch-only price, and confirm both appear correctly in the right branches within minutes.
6. Payments and delivery integration
Check support for local card terminals, wallets and cash reconciliation, and whether orders from delivery apps land in the same queue as your own. Test it: ask for a live example from a customer in your country, not a slide.
7. Reporting and permissions
Role-based access protects discounts and voids. Reports should show sales by hour, item and branch, plus staff performance and tax summaries. Test it: log in as a waiter and try to void a paid item or open the sales report. Both should be blocked.
Three scenarios
These are illustrative composites, not client stories. They show how priorities change with the business.
Scenario 1: a single café with unreliable Wi-Fi
One counter, one printer, a small menu and an internet connection that drops twice a week. The priority is offline mode and speed. Inventory can start simple, with a handful of key ingredients. A ready-made product is the sensible choice, as long as the offline demo passes.
Scenario 2: a three-branch group with a central kitchen
A group prepares sauces and dough in a central kitchen and delivers to three branches. They need central menus and prices, stock transfers between locations, branch-level and consolidated reports, and staff roles that differ by site. The priority is multi-branch management and inventory with recipes and sub-recipes. A product can work if it supports transfers and consolidated costing, but if it forces workarounds the group should price a custom system.
Scenario 3: a delivery-heavy brand or cloud kitchen
Most orders come from delivery apps and the brand’s own app, with no dining room. The priority is integration: every channel in one queue, preparation time tracking, packaging as a cost item and stock that updates from all channels. Here the number of available connectors matters more than table management. If a key platform has no connector, custom integration becomes part of the project. Our guide on how to build a food delivery app covers the ordering side.
Hardware checklist
| Item | What to check | Why it matters |
|---|---|---|
| Terminals and tablets | Touch response, screen brightness, spill resistance and spare units | A frozen screen at the till stops service |
| Receipt and kitchen printers | Supported models, network or USB connection, paper size and spare rolls | Kitchen tickets must print even when the internet is down |
| Kitchen display screens | Mounting, heat and grease tolerance, readable text from the station | Cooks should read orders at a glance |
| Cash drawer | Opens only on cash sale or manager action | Prevents untracked cash handling |
| Card terminals | Integration with the POS or manual entry fallback | Reduces double entry and mismatches |
| Network and power | Router quality, a backup connection and a small UPS | Keeps the router and till running through short outages |
Compare total cost over three years
Subscription prices look small until you multiply them. The numbers below are invented to show the method, so replace them with real quotes. Imagine three branches with two terminals each and a fee of 25 dollars per terminal per month. That is 6 × 25 × 12 × 3 = 5,400 dollars over three years in fees alone, before hardware, payment processing fees, add-on modules such as inventory or loyalty, and support.
For every option, build a three-year sheet that includes monthly fees, hardware purchase or rental, payment processing, add-ons, installation, training and the cost of any workaround you would need. A custom system shifts the shape: higher upfront investment and then mostly hosting and maintenance, with no per-terminal fees.
Ready-made POS or custom POS?
Ready-made systems are the right answer for most single restaurants: quick to start, supported and predictable. A custom POS makes sense when you run several concepts, need integrations nobody offers, want loyalty or ordering logic specific to your brand, or want to own your data. A common route is to start with a product and move to custom when workarounds multiply.
| Question | Ready-made POS | Custom POS |
|---|---|---|
| Time to start | Days to weeks | Weeks to months, delivered in phases |
| Ongoing cost | Per terminal or per branch fees | Mostly hosting and maintenance after the build |
| Fit to your workflow | Configure within limits | Built around your flow |
| Integrations | Available connectors only | Any system with an API |
| Data ownership | Vendor platform | Yours |
Signs it is time to consider custom
- You pay for several add-ons that still do not fit your workflow.
- Your staff keep a parallel spreadsheet for something the POS cannot do.
- A delivery platform or payment method you depend on has no connector.
- Per-terminal fees grow faster than your revenue as you add branches.
- You want loyalty, pricing or ordering rules unique to your brand.
A ten-step demo script
Use the same script with every vendor so the comparison is fair.
- 1Sign in as a waiter and open a table.
- 2Add a dish with two modifiers and a note for the kitchen.
- 3Send to the kitchen and watch the ticket or display.
- 4Switch off the internet and take two more orders.
- 5Split one bill three ways and take cash and card.
- 6Void a paid item as a waiter, then as a manager.
- 7Change a price centrally and check it in two branches.
- 8Run the daily sales report and compare it with what you entered.
- 9Restore the internet and confirm that everything synchronised once.
- 10Export your data and check that you can open it.
Questions to ask every vendor
- Does it work fully offline, including printing and cash payments?
- Can I export all of my data at any time?
- How are menu and price changes pushed to every branch?
- What happens to pricing and support if I add branches?
- Which delivery platforms and payment terminals are supported in my country?
- What are the support hours, and who answers on a Friday night?
Rolling out without disrupting service
| When | What to do |
|---|---|
| Two weeks before | Enter the menu, modifiers, taxes and staff roles. Print test tickets in the real kitchen. |
| One week before | Train staff on real orders, in small groups, on a quiet shift. |
| Launch week | Start in one branch during a quiet period. Keep the old process as a fallback for the first days. |
| Two weeks after | Review reports, fix menu mistakes, adjust roles and discount limits. |
| One month after | Roll out to the next branch using what you learned. |
Common mistakes
- Choosing on price per terminal without counting add-ons and payment fees.
- Skipping the offline test because the vendor says it works.
- Training only managers, not the staff who use the till every day.
- Loading an unfinished menu on launch day.
- Leaving permissions open, then discovering voids and discounts later.
If you are comparing options or considering a bespoke system, see our POS and restaurant systems service or send us your brief.
Frequently asked questions
Does a restaurant POS need internet?
A modern POS should work offline for orders, printing and cash payments, then synchronise when the connection returns. Card payments depend on your terminal and provider, so test that too.
Can one POS manage multiple branches?
Yes. Look for central menu, price and role management with per-branch overrides and consolidated reporting.
How much does a restaurant POS cost?
Prices vary by terminals, branches and features. Compare monthly fees, hardware, payment fees and add-ons over three years, not just the first month.
When should a restaurant build a custom POS?
When you run several concepts or branches, need integrations or loyalty logic that products do not offer, or want to own your data and avoid per-terminal fees.
How long does it take to roll out a new POS?
A single branch can go live within a couple of weeks of menu setup and training. Plan one branch first, then expand once the menu, roles and reports are right.
What hardware do I need for a restaurant POS?
Typically terminals or tablets, a receipt printer, kitchen printers or display screens, a cash drawer, card terminals and a reliable router with a backup connection. Check which models the system supports before you buy hardware.

