Every kiosk vendor sells QSR operators the same two numbers: average order value goes up, labour cost comes down. Both can be true. Neither is automatic. After deploying selfordering across a large number of outlets, here’s the honest version.
What genuinely changes
- Order composition, more reliably than order count. A screen upsells consistently. It never forgets the add-on, never rushes because there’s a queue behind you, and never skips the combo prompt at 1pm rush. The lift shows up in attach rate — drinks, sides, upgrades — more than in footfall.
- Peak-hour throughput. The real constraint in most Indian QSR outlets isn’t the kitchen, it’s the ordering counter during a 40-minute lunch spike. Two kiosks running in parallel with one counter turns a single-file queue into three lanes. The kitchen becomes the bottleneck — which is a better problem, and a measurable one.
- Order accuracy. Customisation-heavy menus — no onion, extra cheese, spice level — are where verbal ordering breaks. Kiosk orders arrive at the KDS as structured data. Remakes drop.
- Staff redeployment, not staff reduction. This is where most projections go wrong. In practice the counter person doesn’t disappear; they move to expediting, kiosk assistance, and table delivery. The saving is in not adding headcount as volume grows, not in cutting existing headcount on day one. Model it that way and your business case survives contact with reality.
What doesn’t change
- Bad menu structure stays bad. If your menu has 140 SKUs and no clear hierarchy, putting it on a touchscreen makes it worse. Kiosk deployment is a good forcing function for menu rationalisation.
- Slow kitchens stay slow. Faster ordering into a constrained kitchen just moves the queue from the counter to the pickup area, where customers are angrier because they’ve already paid.
- First-time users need help. Expect an assist rate in the early weeks, particularly in tier2/3 locations and with older customers. Staff a kiosk host for the first fortnight. Outlets that skip this see adoption stall and then blame the hardware.
The five things that decide whether a QSR kiosk deployment works
Menu design for the screen. Categories a customer can scan in two seconds, hero items surfaced, images that match what’s served. Not a PDF menu ported to a touchscreen.
Payment reliability. UPI must be the primary path, and it must be fast. A payment failure at a kiosk is worse than at a counter, because there’s no human to reassure the customer. Insist on tested failure handling: what the screen shows on timeout, how refunds are triggered, who the customer talks to.
POS and KDS integration. The kiosk must write into your existing POS as a first-class order type, with correct tax handling, discounts, loyalty and reporting. If reconciliation at day-end needs manual work, the store manager will quietly stop using the kiosk.
Placement. Inside the door, before the counter, with clear sightlines and enough space that a kiosk queue doesn’t block the entrance. Sounds trivial. It determines adoption more than the UI does.
Uptime. A dark kiosk trains customers not to use kiosks. One extended outage in week two can cost you months of adoption at that outlet.
How to run the pilot properly
Don’t pilot in your best outlet. Pilot in three:
- Your highest-volume urban outlet (throughput case)
- A mid-volume mall outlet (typical case)
- A highway or tier-2 outlet (worst case for user familiarity and power quality)
Run for a full 8 weeks. Measure: kiosk order share, AOV kiosk vs counter, attach rate, average order time, remake rate, uptime, and assist rate. Compare against the same outlets’ prior-year same-period, not against each other.
If kiosk order share crosses roughly half of transactions by week six without staff pushing customers toward it, the format works for your brand. If it stalls under a quarter, the problem is almost always menu structure, placement, or payment speed — in that order — not the customer.
Rollout economics
Model the payback on three lines: incremental margin from attach rate, avoided incremental headcount at growth, and reduced remake waste. Be conservative on all three. If the payback only works when you assume aggressive AOV lift, the deployment isn’t ready — the menu is.
Restro360 by Digitos is a self-ordering platform built on kiosks we manufacture and service ourselves — hardware, software, payments and field support from one accountable party.