This question comes up constantly in the industry. Almost every time, the person asking is approaching it incorrectly.
The question being asked is "where can a kiosk be placed?" as if it were a piece of furniture that needs a spot. This approach is backwards. The right question is: where does a kiosk actually make sense-not as a supplementary feature, but as the only reasonable option?
Many people will not acknowledge: most kiosk deployments fail. There is no hard data on this, but those who have been in the industry long enough will confirm. In any mid-tier hotel or government building, there are expensive touchscreens gathering dust in corners, screens frozen on error messages, or completely powered off. Someone approved a six-figure purchase order for those.
Start With What Actually Works
Airport check-in is the textbook example. Passengers have to be at the airport anyway. The process is completely deterministic-pick the flight, choose a seat, confirm, print. Nobody is "just browsing" at a check-in kiosk. And boarding passes and luggage tags need printing. Phones cannot do that. Period.
Airlines have increasingly adopted self-service technologies (SSTs) to deliver more streamlined service models. In 2007, the International Air Transport Association launched the Fast Travel Program aimed at boosting operational efficiency for both airports and carriers. This initiative encompasses enhancements across multiple touchpoints, including passenger check-in, luggage processing, document authentication, reservation management, boarding procedures, and lost baggage retrieval. Quote-https://pmc.ncbi.nlm.nih.gov/articles/PMC11471516/
There is a pattern here worth pulling out. Physical presence is non-negotiable-users are already there; this cannot be done from a couch. The process is completely standardized. No judgment calls, no "let me check with my manager." And the user already knows what they want. Not browsing, not exploring-just executing.
Here is a counter-example that complicates this logic: exhibition guides. Museums used to favor kiosks-unlimited depth of content, far more than any physical plaque could hold. This hits all three criteria. Users are there, the process is simple, they know what artwork they are looking at.
Then QR codes and museum apps happened. Now visitors scan and get everything on their phones, which they carry with them as they walk. The kiosk lost.
So there is a fourth element that people forget: the kiosk needs to do something a phone cannot. Usually that means hardware-printing, scanning, reading cards, dispensing something physical. If the whole interaction is just tapping a screen to get information, there is no reason someone would not just use their phone.
When do museum kiosks still win? When the big screen IS the experience. High-resolution art details impossible to see on a phone. Immersive video that loses impact at 6 inches. Interactive games with touching and manipulating. For pure text though? The phone won.
The McDonald's Thing
Fast food ordering-McDonald's specifically-reveals something unexpected. One regional QSR chain spent three months trying to figure out why their kiosk rollout was not hitting the numbers McDonald's was reporting. The investigation went deep.
McDonald's has 40,000+ of these things globally, and the average ticket went up 15-30% after installation. This sounds like marketing fluff at first. Then the psychology becomes clear.
At a counter, there is subtle pressure. Someone is waiting. There might be a line forming. Orders happen fast, safe choices get made, extras do not get added because saying "actually, throw in a pie too" feels awkward while someone stares.
The kiosk removes all that. Browsing happens. New items get considered. "Would you like to add a combo?" gets a yes without anyone feeling judged. It turns out people actually want to spend more-they just needed the social pressure removed.
What that regional chain got wrong was placement. Kiosks went next to the counter, so customers still felt watched. McDonald's puts them in a separate zone. This sounds trivial. It is not. The physical space design matters as much as the software. Menu engineering too-the counter menu cannot simply be digitized. The upsell prompts, the item ordering, the default selections, all of it needs rethinking. Most chains do not want to hear that.
This flips the usual thinking. The best kiosk scenarios are not just where kiosks CAN work. They are where human interaction is actually getting in the way.
Real Money Gets Wasted Too
Supermarket self-checkout sounds perfect until actual use. Few items, nothing unusual? Great, faster than the cashier line. But produce that needs weighing, alcohol that needs ID verification, one scan failure-now someone is standing there waving at an attendant anyway. And there is an uncomfortable truth nobody wants to admit: cashiers are professionals at scanning. They are fast; they know where the barcodes are. Regular customers are amateurs. For a basket of 30 items, the human is probably faster.
Self-checkout works as a supplement. The moment it becomes a replacement, problems multiply.
When self-service checkout (SCO) was initially introduced in the United States in 1992, it was met with significant skepticism, along with widespread concerns that substantial losses would ensue. Although research findings remain mixed regarding their effect on shrinkage, particularly customer theft, consumer-oriented payment technologies have become an increasingly prevalent feature within the retail landscape. Quote-https://journals.sagepub.com/doi/abs/10.1177/1748895816643353
Banking kiosks for anything beyond checking balances and printing statements hit a wall that is not rational but very real: when it involves money, people want a human. They want someone to confirm the transfer went through. They want to see someone nod when large amounts are moving. The machines are perfectly capable. Customer trust is not there.
Back to What Works
Hospital registration works the same way as airports-long lines, simple transactions. Once registration kiosks exist, adding payment functions costs almost nothing. Same for printing test results, checking queue status, and similar functions. One device becomes a swiss army knife.
Hotels? Mostly useful for late check-ins when staff is thin. Privacy is a factor too-some guests do not want to announce their room rate to whoever is behind them.
Corporate visitor management is underrated. Register name, company, scan ID, take photo, print badge, notify host. All standardized, all requires peripherals. And it upgrades the receptionist's job from "data entry human" to "actually greeting and helping people."
Government Services
Here is a structural problem: service halls are open when citizens are at work. Closed on weekends. Lunch breaks happen. Numbers stop being taken at 4 PM. Kiosks break that completely. Put them in community centers, bank branches, shopping malls. Suddenly "get it done on the way home" becomes possible.
And there is something that rarely gets mentioned: government windows become emotional outlets. People show up frustrated, argue about requirements, take it out on staff. The burnout is real. Kiosks absorb the standardized tasks. Human windows handle genuine complexity. Everyone is happier.
Obviously Wrong Applications
Anything requiring emotional intelligence-counseling, delivering bad news, funeral services-is obviously out. This should not need explaining, but proposals for "grief assistance kiosks" have actually crossed desks, so apparently it does.
Highly personalized decisions-insurance planning, investment advice, legal consultations-cannot be reduced to decision trees. The whole point is human judgment adapting to unique situations.
Complex form-filling is torture on a touchscreen. Typing efficiency drops to maybe a third of a physical keyboard. Forcing users to input paragraphs of text on a kiosk is hostile design.
Pure information lookup-"what time does the mall close?"-is already solved. People search for it online. Nobody is walking to a kiosk to ask.
Evaluating New Scenarios
There is no formula for this. It is more about whether a situation has the same DNA as the successful cases.
First thing to identify: are people already required to be somewhere? If this could be handled from a couch, the battle is already lost. Phones and laptops win. Hardware gets installed for people who do not need to be there, and then nobody uses it.
Then the flexibility question. "It depends" or "sometimes exceptions are made"-that means a human stays in the loop somewhere. Maybe 80% gets handled by kiosk, but that 20% frustrates everyone. Staff and users both.
And this keeps coming up: what is the kiosk doing that requires it to be a kiosk? "Displaying information on a bigger screen" is not enough. The printer, the scanner, the card reader, the thing that dispenses a physical object-something the phone literally cannot do. That is what justifies the hardware.
All three elements align, projects usually work. One is uncertain, think harder. Two are missing, the question becomes why kiosks are even in the conversation.
The kiosk itself is never the answer anyway. It is one possible implementation of an answer. The actual answer is figuring out what problem gets solved, for whom, under what constraints. Get that right and the hardware question almost answers itself.
