[ GUIDE ]

A practical guide to self-order kiosks

Kiosks are sold as a labour saving and are usually better understood as a throughput increase. The distinction changes both the business case and how many you buy.

What a kiosk actually changes

A self-order kiosk is a guest-facing touchscreen where customers build and pay for their own order, which then goes straight to the kitchen. Its main effect is adding an ordering lane without adding a person — so it helps most where the constraint is order-taking capacity during a peak, and helps little where the constraint is the kitchen.

The honest framing: if your kitchen is the bottleneck, a kiosk moves the queue from the counter to the pass. Fix throughput first, then add ordering lanes.

Where kiosks work, and where they do not

Peaky counter service

Strong fit

A queue that peaks far above its average — the kiosk absorbs the peak without staffing for it.

Large or configurable menus

Strong fit

Guests take their time without holding up a line, and see the whole menu rather than the top of it.

Campus, food court, transit

Strong fit

Guests already expect self-service and are often in a hurry.

Full table service

Weak fit

The service model is the product. A kiosk competes with what guests are paying for.

Very short menus

Weak fit

When ordering takes eight seconds, a screen does not beat a person.

Kitchen-constrained sites

Weak fit

More orders arriving faster at a kitchen already at capacity makes the experience worse, not better.

How many, and where

Size to your peak, not your average. Count orders in your busiest fifteen minutes and work back from a realistic ordering time — usually 60–120 seconds for a menu guests know, longer for one they do not.

What people get wrong

  1. Treating accessibility as optional. Reach height, large touch targets and readable contrast are requirements. A kiosk only usable standing, with good eyesight, excludes paying guests — and in many jurisdictions, breaks the law.
  2. Porting the printed menu unchanged. A kiosk menu needs its own hierarchy. Guests scroll far less than you expect.
  3. Hiding the modifiers that make money. Upsell prompts work on kiosks — but only when they are relevant, and once.
  4. Forgetting cash. A card-only kiosk excludes cash customers; keep a staffed lane.
  5. Ignoring the offline case. A kiosk that cannot take an order during an outage is a very expensive poster.

Where ThaliPOS fits: the kiosk runs on the same hardware as the counter station with reach-mode and large-target layouts, takes card payment at the screen, and keeps taking orders during an internet outage because it talks to the in-store hub. The ThaliPOS kiosk →

[ FAQ ]

Questions, answered straight

Do self-order kiosks reduce labour costs?

Usually they increase throughput rather than reduce headcount — they add an ordering lane without adding a person. If your kitchen is already the bottleneck, a kiosk moves the queue rather than removing it.

How many kiosks does a restaurant need?

Size to your peak, not your average. Count orders in your busiest fifteen minutes and divide by a realistic ordering time — typically 60 to 120 seconds per guest.

Should I remove the staffed counter if I add kiosks?

No. Some guests will not use a kiosk, cash customers need a lane, and accessibility requires an alternative. Keep a staffed counter.

What happens to kiosk orders if the internet goes down?

It depends on the architecture. If the kiosk talks to a local in-store hub, ordering continues and only card payment is affected. If it talks to the cloud, the kiosk stops.

Tell us about your restaurant.

We're onboarding pilot restaurants now. Email us and we'll tell you honestly whether ThaliPOS is a fit for how you run service.

Email hello@thalipos.com