Biometric Payments: How Face Pay Works at Checkout

With biometric payments, customers pay with their face, palm or fingerprint instead of a card or phone. Face pay links a verified face to a payment account, then matches the customer at the counter and authorises the charge. It shines in closed environments such as canteens and campuses, provided liveness checks stop spoofs.

What Biometric Payments Are: Face, Fingerprint and Palm

In short, a biometric payment uses a physical trait to confirm who is paying. The trait either replaces the card or phone entirely, or it unlocks a payment credential stored elsewhere. Today, three methods dominate, and our comparison of fingerprint vs face recognition covers two of them in depth.

Method	Where it runs	What the customer carries	Main trade-off
Fingerprint on phone or card	The customer’s own device or a biometric card	A phone or card	Still needs the device, so it speeds up payment rather than removing the wallet
Palm	A dedicated palm reader at the counter	Nothing	However, it needs special hardware at every point of sale
Face	A camera at the counter, kiosk or gate	Nothing	Still, it needs strong liveness and careful matching to avoid errors

The idea, however, is older than most people think. For example, Alipay launched its “Smile to Pay” service in a Hangzhou restaurant in 2017. In 2022, Mastercard announced a Biometric Checkout Program, with a first pilot in Brazil. Since then, large payment processors have also brought face and palm checkout to more merchants.

How Face Pay Works: Enrolment, Matching and Authorisation

Face pay looks instant at the counter, but in fact it rests on three steps. Each one carries its own risks, so it helps to look at them separately.

Step 1: Enrolment

First, the customer signs up, usually in an app. Then they take a selfie, pass a liveness check and link a payment method. Strong programmes then compare that selfie with a verified ID, so the face belongs to a real, checked person. Otherwise, a fraudster can enrol a stolen card under their own face.

Step 2: Matching at the Counter

Next, the terminal captures the customer’s face and searches for a match. Here, design choices matter most. A one-to-many search compares the face against every enrolled customer, which is fast but riskier as the gallery grows. A one-to-one check, by contrast, first narrows the search with a phone number, a member ID or a short PIN, then confirms the face.

Step 3: Authorisation

Finally, once the face matches, the system charges the linked account. Also, many deployments show a confirmation screen with the amount and the customer’s first name, and they ask for a tap to approve. That last step adds a second or two, but it prevents accidental charges when someone simply walks past the camera.

Above all, the matching step depends on solid face verification. Otherwise, the convenience disappears in false matches and failed payments.

Face Pay vs Card vs QR at Checkout

Of course, each checkout method suits a different setting. The comparison below focuses on what operators actually trade off.

	Card	QR code	Face pay
What the customer needs	A physical card	A phone with a banking or wallet app	Only their face, after enrolment
Hardware at the counter	A card terminal	A printed or displayed code	A camera terminal or tablet
Steps for the customer	First hand over or tap the card, then sign or enter a PIN	First open the app, then scan, enter the amount and confirm	Simply look at the camera, then confirm
Typical fraud risk	Stolen or skimmed cards	Fake QR codes and phishing apps	Spoofed faces and false matches
Works without a phone	Yes	No	Yes
Best fit	Open retail, any customer	Open retail with high wallet use	Repeat customers, especially in closed environments

The pattern, then, is clear. Face pay wins where the same people pay often, such as staff, students or members. In contrast, it adds little for one-off customers who would need to enrol first.

Benefits and Limits of Face Pay for Operators

For the right operator, face pay changes the economics of a busy counter. Still, it is worth weighing the gains against the work involved.

What Operators Gain

• Shorter queues. Customers skip the wallet and the phone, so lunchtime lines move faster.
• Fewer lost cards and disputes. There is no campus or staff card to lose, lend or replace.
• Built-in loyalty. Because the system recognises each customer, points and offers apply automatically.
• Better data. Operators see who buys what and when, within the limits of the customer’s consent.

What Operators Must Invest In

• Enrolment effort. Every customer must sign up once, which needs a simple app flow and clear communication.
• Camera placement. Poor lighting at the counter causes failed matches, so plan the hardware with care.
• Privacy work. Consent screens, retention rules and deletion requests all need owners and processes.
• A fallback. Guests and customers who opt out still need a card or QR option.

Fraud and Privacy Risks to Plan for in Biometric Payments

A face at the counter is convenient for customers and, unfortunately, for fraudsters too. So plan for four risks before launch.

• Spoofing at the terminal. Someone holds up a photo or plays a video of an enrolled customer. Therefore, liveness detection at the terminal is the main defence.
• False matches. In a large gallery, two similar faces can match by mistake. However, narrowing the search with a phone number or member ID turns a risky one-to-many search into a safer one-to-one check.
• Enrolment fraud. A fraudster links a stolen card to their own face. Fortunately, matching the enrolment selfie to a verified ID closes that gap.
• Privacy concerns. Face data is sensitive in most privacy laws. So customers should opt in, see a clear notice, and be able to delete their enrolment at any time.

Notably, two of these risks depend on the same control. A liveness check at both enrolment and payment stops most photo and video spoofs, which are the cheapest attacks to try at a counter.

Where Face Pay Fits: Canteens, Campuses and Closed-Loop Retail

Generally, face pay works best in closed loops, where the operator controls both the enrolment and the payment, and where the same people return every day.

• Staff canteens. For example, employees pay with a face linked to payroll deduction or a stored balance, which removes cards and queues at lunchtime.
• Universities and schools. Students pay at cafeterias and shops with a face linked to a campus wallet or parent-funded account.
• Vending machines and lockers. Customers buy snacks or collect parcels without a phone or a code.
• Membership retail and gyms. Members pay and earn loyalty points in one step.
• Events and theme parks. Similarly, guests pay for food and rides after enrolling at entry, so they need no wallet.

In each case, the settings share a known population and a reason to come back. As a bonus, that keeps galleries small, enrolment simple and the time saved meaningful.

Rollout Checklist for a Face Pay Pilot

1. Pick a closed site. Start with one canteen or campus where you can enrol most users quickly.
2. Verify at enrolment. Run liveness and match the selfie to an ID, then link the payment method.
3. Narrow the match. Use a member ID or phone number with the face, especially as the user base grows.
4. Add a confirm step. Show the amount and a first name, and require a tap to approve.
5. Light the counter well. For instance, place cameras away from strong backlight and at a natural head height.
6. Offer a fallback. Keep card or QR payment available for guests and anyone who opts out.
7. Measure and adjust. Track match rate, average checkout time, false matches and opt-out requests every week.

How Verihubs Powers Face Pay

Verihubs Facepay applies face recognition to checkout, so customers can pay at counters, vending machines and lockers with a quick face scan. Under the hood, Verihubs face recognition handles enrolment matching and the search at the terminal, while liveness detection guards against photos and videos. Verihubs also supports linking loyalty programmes to the same scan.

For a wider view of face-based checks beyond payments, see our overview of AI biometrics in identity verification.

Frequently Asked Questions About Biometric Payments

What are biometric payments?

With biometric payments, customers pay by face, fingerprint or palm instead of handing over a card or scanning a QR code. The trait either replaces the card or unlocks a stored payment credential.

How does face pay work?

First, customers enrol a selfie and link a payment method. Then, at checkout, a camera captures their face, the system finds the matching account, and the customer confirms the charge on screen.

Is paying with your face safe?

Yes, as long as the system uses liveness detection to stop photos and videos, verifies identity at enrolment, narrows matching with a member ID or phone number, and stores face data securely with the customer’s consent.

Where do businesses use face pay?

Face pay works best in closed environments with repeat customers, such as staff canteens, campuses, vending machines, lockers, gyms and membership stores.

How does face pay compare with QR payment?

QR payment needs a phone and several taps in an app. By contrast, face pay needs only a quick look at the camera and a confirmation, so it suits places where customers pay often and may not carry a phone.

Biometric Payments Work Best Where Customers Come Back

Admittedly, face pay is not a universal replacement for cards or QR codes. Its strength is narrow but real: repeat customers in places the operator controls, where seconds saved at every visit add up and enrolment happens once.

Meanwhile, the security case rests on a few deliberate choices. Verify identity at enrolment, run liveness at every payment, narrow the match where galleries grow, and ask for a quick confirmation. Operators that make those choices get faster queues without trading away safety, which is the only version of face pay worth deploying.

Thinking about face pay for a canteen, campus or vending network? Talk to Verihubs about a Facepay pilot.