Digital Loyalty Cards for UAE Shops: How Wallet Cards Work and What They Cost to Run (2026)

Most repeat-visit shops in the UAE already have some kind of loyalty card. A paper punch card in a drawer, a stamp app the owner signed up for once, or nothing at all. This post sorts out which is which, how a wallet card works at the counter, and what it costs to run, with every price read off the vendor's own page on October 5, 2026.

We build connected systems for shops, so we have a view, and we say below when we would tell you not to build anything. We have built one wallet card ourselves, for a repeat-visit shop in the UAE. It is not live yet, so we describe how it works and what we tested, and we quote no results from it.

What is the difference between a paper punch card, a stamp app and a wallet loyalty card?

The paper punch card. Buy ten washes, the eleventh is free. It costs almost nothing, needs no explaining, and works well when the owner is at the counter and knows the faces. It breaks when a card is lost, when someone stamps it without a sale, and on the one question that matters: who stopped coming? A paper stamp card cannot tell you, because you never knew who held one.

The stamp app. A stamp app, or loyalty app, puts the card on the customer's phone, usually inside the vendor's own app. Stamp Me, for example, says on its site that customers first download its app to join your program, and can then save a member code to Apple Wallet or Google Wallet. That download is the weak point for a shop a customer visits twice a month. The strong points are a low price and a quick start.

The wallet loyalty card. This is a pass saved in the phone's own wallet: Apple Wallet on an iPhone, Google Wallet on Android. The customer adds it from a link, a QR code or an NFC tag, with no app. Because the shop can change the pass after it is saved, the points and level on the card update themselves, and a message can show on the lock screen. Google lists the UAE among the countries where Google Wallet is available.

If you run a loyalty card for a Dubai shop with one counter and one trusted person at the till, any of the three can work. The differences show up when more than one person takes payment, when one customer pays for several people, or when the points should follow the money. We come back to that below.

How does a digital loyalty card work in a shop, from the first tap to the next level?

Here is how a card connected to the till works, based on the one we built. Everything below is design plus our own tests with made-up customers. It is not a customer result.

1. The customer joins on their own phone

An NFC tag sits at the counter, with a QR code beside it for phones that do not read tags. Tapping or scanning opens one page. It asks for the phone number first. A new number gets the short registration form and then an Add to Wallet button.

A number already on file gets a masked check before the card is shown, so someone who types another person's number cannot walk off with that customer's card. This matters more than it sounds: a card tied to a typed phone number is a way into someone else's points.

Keep the form short. On an earlier project, replacing a paper registration with one page took the time at the counter from 7 min 15 sec to 1 min 10 sec, 84% faster.

2. The card lives in the customer's wallet

It shows their name, points and level. A welcome gift is credited when the card is added, once per phone number. Every time the card changes, the system draws the whole card again from the shop's own record, so the card cannot drift away from the real balance.

3. The till adds points on each paid visit

No staff tap, no code to remember. When a payment is closed at the till, the system adds the points to whoever paid. In a shop where one person often pays for a group or two people split a bill, that matters: the points follow the payment, not the person standing at the counter.

Points a customer spends come off the invoice line, and the invoice says so, so the books still match the money taken. Notifications are routed by the same rule. We tested who hears what over 160 generated visits (groups, split payments, repeated writes) so a message never reaches the wrong customer.

4. The card updates, and says something readable

A bare number change is a poor lock-screen message. The card carries a message line, so the customer sees a sentence about the points earned today instead of a new total. In our tests on an iPhone, the update arrived in under five seconds from the moment the till answered.

Android is different, and you should know before you promise anything. In our test on one older Android phone the card was always current in Google Wallet, but the notification was generic ("updated"), and one arrived much later than the iPhone's. The timing is Google's, so do not sell Android customers an instant ping.

5. Levels unlock

Levels are counted in visit days over a rolling year, so five purchases in one afternoon do not buy a level. Each level has its own card design. When a payment also moves someone up a level, the system sends the points change first and the level change second, so each shows its own message instead of a generic "pass changed" banner on the iPhone.

If the pass service is unreachable for a while, the points are already in the system's record and the card catches up on the next sweep. What happens when your own internet drops is a different problem, which we cover in our post on POS offline mode.

How do you bring back customers who stopped coming?

Most loyalty programs reward the people who already visit. The shop's real leak is the customer who came three times and vanished, and a punch card cannot see that person. A card connected to the till can, because the till knows the date of every paid visit.

The design is simple. After a set gap with no paid visit, the card holder gets a time-limited return offer as a message on the card. The gap should match how often your customers normally come, and a month is a sensible starting point for a shop people visit often. The rules we chose, and why:

• A discount on the next visit, never bonus points. A points reward for being away can be collected by timing a visit to land just as the gap ends.
• Time-limited. It gives a reason to come this week, not someday.
• Applied at the till and used once. The pay screen shows the offer, and the invoice line and payment record name it, so the owner can count exactly what win-backs cost and how many used them.
• Repeat rules are the owner's call. Whether the same customer can get the offer again after another gap is a setting, because it is the owner's money.

There are two ways to deliver it. A wallet notification only reaches people who added the card, carries no WhatsApp fee, and can be removed by the customer whenever they like. Two of the services we checked state unlimited push messages on their paid plans (Loopy Loyalty and Boomerangme).

WhatsApp can also reach customers who never added the card, but it needs an approved marketing template and the customer's consent, and it costs money per message. Meta's UAE marketing rate is about AED 0.18 per message (USD 0.0499), so 100 win-back messages come to roughly AED 18, before your provider's own handling fee. Our posts on what WhatsApp automation costs and on consent and the rules cover both in detail. In our build we used wallet notifications only, with no WhatsApp at the start.

The same trigger can send a quiet-hours offer: a short-lived message to card holders for a slow window, to fill a free slot or an empty chair rather than discount people who would have come anyway. We have not run one yet, and the right window should come from your own till records, not from a guess.

One honest limit: our build is not live yet, so we have no return-rate figure to show, and we will not guess one.

What does a digital loyalty card cost to run in the UAE?

Four cost lines, whichever route you take: the pass or stamp service, Apple's developer fee if the service needs it, messaging, and the system that connects the card to your till. Here are the services we could verify, each read from its own pricing page on October 5, 2026.

Read what the meter counts

The monthly price tells you less than the thing it is metered on. Stamp Me and Loopy Loyalty charge a flat plan by features and locations. PassKit charges by team members and by cards held. The service we chose for our build charges on card creations and changed updates, not on cards held, so a lapsed customer's card costs nothing to keep and we never need to delete one to save money. Ask which number your bill grows with.

Check whether you need your own Apple Developer account

To sign Apple Wallet passes, some services need you to hold an Apple Developer account. PassKit's help page says there is no alternative. Apple lists the program at 99 USD per membership year (about AED 364), shown in local currency at enrolment. Other services sign passes with their own certificate, so you need no account. The one we chose does, and its support confirmed that to us in writing.

This choice is hard to undo. On the service we used, the certificate is fixed when a card is created, so switching to your own later means issuing every customer's card again. Decide before the first card goes out.

Messaging

Wallet notifications carry no Meta fee. WhatsApp templates do: about AED 0.18 for a marketing message and about AED 0.06 for a utility message in the UAE, per the rates in our WhatsApp cost post. The send fees are small. The hours spent getting templates approved and consent recorded are not.

The system that connects the card to your till

A stamp app has none of this cost, because the vendor hosts everything. A connected card needs a small system between the till and the card service. Hosting for one shop usually runs under 100 AED a month. The real cost is building it: our projects start at 7,000 AED, with most landing between 10,000 and 35,000 AED, one time.

Put together, a single-counter shop on a flat wallet plan pays roughly AED 92 to 250 a month, plus about AED 18 for every 100 WhatsApp offers it chooses to send. That is the number to beat before any custom build is worth discussing.

Is an off-the-shelf stamp app enough?

For many single-counter shops, yes. Buy a flat plan and do not pay us. A stamp app or a wallet pass service is enough when all of these hold:

• There is one counter, and one or two people you trust take every payment.
• The rule is simple, such as one stamp per visit, rather than points per dirham paid.
• Nobody pays for a group, and bills are rarely split.
• The card does not need to agree with your invoices or your VAT records.
• You have a few hundred active customers, not several thousand.

In that case the staff tap is a small risk and a connected build is money spent for little gain. Our custom versus off-the-shelf post makes the same argument for software in general: buy first, build when the gaps between your tools are the real cost.

When does a connected loyalty card earn its place?

When the points should come from the till automatically, not from a staff tap. That sentence hides five situations:

• Points depend on the amount paid or on who paid. A tap cannot get this right on a busy Saturday.
• More than one person takes payment. Every manual step is a place for a forgotten stamp or a stamp on the wrong customer.
• Points are spent against the bill. The discount then has to show on the invoice, with VAT worked out on what was actually paid, so the card and the books agree.
• You want win-back offers applied at the till. The offer, its use and its cost are recorded on the same payment.
• You want one set of numbers. Money taken, points issued and offers used should reconcile, which they only do when one system records all three.

This is where we fit. We do not sell a loyalty app. We connect the till, the invoices, the card and the messages you already use, so the card is one more output of a payment instead of a second job for your staff.

The honest cost is upkeep: a connection has to change when the till does. If you are still choosing a till, see our comparison of POS software sold in the UAE and ask each vendor whether it can send each payment out to another system. Where the tools keep their own islands of data, that gap is the usual cause.