Do kiosks raise check size? What studies actually found

Sometimes. Chains that report bigger kiosk checks compare kiosk orders with register orders, and their guests chose which one to use. The only randomized test we found, at a campus Burger King, measured spending about 11% lower per person at the kiosk than at the cashier. Neither kind of evidence tells you what a kiosk will do in your store.

This page lists each source we could read in full, how it measured, and what it cannot show. For prices, see the kiosk cost guide. For choosing a kiosk, see the kiosk buyer's guide.

What have studies and chains reported?

Source	How it measured	What it reported	What it cannot show
Master's thesis, University of North Texas (May 2019)	Guests at a Burger King in a student union were randomly sent to a kiosk or a cashier; 192 surveys	Sales per person $5.56 at the kiosk and $6.26 at the cashier; ordering took 48.9 seconds at the kiosk and 58.7 at the cashier	One store, mostly students, spend reported on a survey, and a kiosk screen with no photos
Journal of Hospitality and Tourism Technology article (July 2021)	A 2 × 2 field experiment in a quick-service outlet, kiosk or cashier, busy or quiet periods	Guests rated the kiosk experience better; cashiers generated more revenue overall, and more sales than kiosks in busier periods	One outlet; we read the abstract, which gives no dollar figures
Shake Shack shareholder letters (August and November 2023)	The company's own comparison of kiosk orders with other in-Shack orders	Kiosk average order values about a high single-digit percent above in-Shack levels in the second quarter of 2023, and more than that in the third, driven by more items per check and mix	Guests picked the channel; no method, store count or time window given
Shake Shack earnings call, reported by Nation's Restaurant News (May 2024)	Executives' comments on first-quarter 2024 results	Average checks up in the high teens, in percentage terms, compared with other channels	A reporter's summary of a call; "other channels" is not defined
BurgerFi pilot, reported by Restaurant Business (June 2022)	Chain-provided data from four restaurants that added kiosks from the previous November	Kiosk tickets 18.5% higher than register tickets; kiosks took about three of every four transactions; 52% of kiosk guests took an automated upsell	Four stores, no control group, and the chain supplied the numbers
McDonald's CEO on CNBC (June 2018)	An interview during the U.S. kiosk rollout	"When people dwell more, they select more. There's a little bit of an average check boost."	No figure and no method
Duke University summary of a pizza-chain study (August 2015)	More than 160,000 orders from one North Carolina chain, 2007 to 2011, online compared with phone	Online orders had 14% more special instructions and about 100 more calories	Online ordering on a phone or computer; it measured calories and order complexity, with no dollar figures

Sources: Torres, "Examining Self-Service Kiosks in Quick-Service Restaurant Settings", Leung, Torres and Fan, "Do kiosks outperform cashiers?", Shake Shack's second-quarter 2023 and third-quarter 2023 letters, Nation's Restaurant News on Shake Shack, Restaurant Business on BurgerFi, CNBC on McDonald's kiosks and Duke Fuqua on the pizza study.

For what McDonald's said about its remodels and sales over several years, see our McDonald's kiosk history.

What did the Burger King experiment find?

It is the one source above that assigned guests to a kiosk or a cashier at random, so the two groups started out alike. Kiosk guests reported $5.56 per person and cashier guests $6.26, a gap of 70 cents, or about 11% of the cashier figure. The thesis reports that gap as statistically significant (p = 0.033). At peak the gap was $1.06 ($5.43 against $6.49); off-peak it was 34 cents ($5.69 against $6.03), though the thesis found no significant difference between the two gaps.

The thesis offers its own reasons: the kiosk showed no pictures, showed prices before guests committed, and did little upselling. Students on meal plans may also have watched what they spent. Read it as one store with one screen design. It shows a kiosk can come out below the cashier when the screen does little selling.

Why do chain figures come out higher?

Three things push a kiosk-versus-register comparison up. Only the last is the kiosk itself changing what guests buy.

• Guests choose the channel. A group ordering for four may head to the kiosk while a regular buying one coffee walks to the register. That alone can make kiosk tickets bigger.
• Mix. Shake Shack's letters credit more items per check and mix. In May 2024, its finance chief told investors that new kiosk screens were designed to help guests trade up to a double or add bacon.
• Every guest sees the offer. A kiosk shows the add-on on each order. A busy cashier may skip it. That is a real effect, and you can copy part of it at the register with a script.

What a store owner needs is the whole store's average check and sales before and after the kiosk. None of the chain disclosures above gives that.

What do kiosk vendors claim?

Square's Square Kiosk page tells restaurants to "grow your average check size" by offering upgrades and add-ons every time, with no figure. Bite's home page says its recommendations "will increase your check averages by more than 20%". GRUBBRR's home page says "Avg 20% upsell lift with kiosks". Neither Bite's nor GRUBBRR's page shows how its figure was measured or what it was compared against.

Maple's own voice ordering kiosk, in early access with a small number of restaurants, offers an add-on the restaurant chooses on every order. Maple publishes no check-size result for it.

How should you read a check-size claim?

1. What was compared: kiosk orders against register orders, or the same stores before and after?
2. Did guests choose the kiosk, or were they assigned?
3. Were menu prices the same in both periods? A price rise lifts the average check on its own.
4. Did items per order rise, or the price per item?
5. Did total store sales rise, or did orders move from the register to the kiosk?
6. How many stores and weeks, and who collected the numbers?

How can you measure it in your own store?

Pull four weeks of POS data before the kiosk arrives: whole-store sales, orders, average check and items per order, by daypart. Run the kiosk for at least four weeks with the same menu prices. Compare whole-store numbers, because kiosk tickets set against register tickets mix in who chose the kiosk. Toast's kiosk overview shows how to filter the sales summary to kiosk orders, which tells you how many orders moved to the screen.

Set up the add-on prompts before the test starts; the kiosk menu design guide covers where they go. Then compare the result with what the kiosk costs you each month.

Published by Maple, which sells an early-access voice ordering kiosk. This AI-assisted guide combines a University of North Texas thesis, a journal abstract, Shake Shack's shareholder letters, reporting by Nation's Restaurant News, Restaurant Business and CNBC, and vendors' own pages, with an original summary table and reading checklist. It does not report a Maple test or predict what a kiosk will do to your sales.