Loyalty without the icon

A Customer Loyalty App Has to Beat a Piece of Cardboard First

The paper punch card has never needed an update, a password or a privacy policy. It also costs next to nothing and works during a power outage. Any loyalty app for a small business has to beat that record, and most never do.

November 16, 2026 · 11 min read

acmecoffee.com · loyalty app, one month
Where a loyalty app loses people
A busy counter, and very few installs
Customers at the counter1,000
Noticed the offer310(31%)
Installed and signed up44(14%)
Opened it on the next visit19(43%)
Earned a reward6(32%)

The line behind them is longer than their patience.

Illustrative: made-up numbers for a fictional one-location coffee shop

The short answer

Most small businesses should not build a customer loyalty app. A punch card, a loyalty program in the point-of-sale system or a Wallet pass rewards repeat visits without asking anyone to download anything. A custom app, from $3,750, makes sense only when rewards depend on something your register cannot see, and enough regulars will actually use it.

On this page
  1. What the owner wants is a second visit, not software
  2. Five chores a customer must finish before your app earns a dime
  3. Cheaper roads to the same free coffee
  4. The rare shop where a loyalty app has a real case
  5. Screens, a ledger, and the cashier at 7:40 in the morning
  6. Three loyalty features that can land you in trouble
  7. Test the idea with cardboard before anyone writes code

What the owner wants is a second visit, not software

Start with the sentence nobody writes on the sales page. A loyalty program exists to make a customer come back sooner than they otherwise would. That is the entire job. The app is a delivery mechanism, and a demanding one, because it asks a person holding a coffee to do homework for a free coffee.

The owner of a very small shop already has an advantage no software supplies. They know the regulars. They remember the order. A one-person bakery that says "the usual?" is running a loyalty program with no monthly fee, and a customer who gets remembered is unlikely to feel more valued because of a push notification with a confetti animation in it.

None of this means rewards are pointless. It means the reward and the app are two separate purchases. Pew Research Center, in a fact sheet published October 19, 2026, found 91% of U.S. adults own a smartphone. So nearly everyone at your counter could install your app. Owning the phone is not the same as wanting your icon on it, and that gap is where loyalty budgets go to die. The broader question of whether a small business needs an app at all gets its own page. This one assumes you want repeat visits and asks the cheapest way to get them.

Five chores a customer must finish before your app earns a dime

  1. 1

    Notice the offer

    A sign by the register, or you saying it out loud at the busiest moment of the morning. Every customer who does not hear it is gone from the count.

  2. 2

    Decide it is worth the trouble

    A free sandwich after ten sandwiches is a good deal to a daily customer and an abstraction to someone in town for a funeral.

  3. 3

    Find it in the store

    With a line behind them, on the store's search, among businesses with similar names. Your own sign needs a code, and the code needs a working camera and a patient customer.

  4. 4

    Create an account and answer the permissions

    Name, phone, password, notifications yes or no. This is the step where people look up, notice the line, and promise to do it later.

  5. 5

    Remember it next time

    The app has to be opened, at the counter, before paying. Forget once and the customer learns the program is optional. Forget twice and it is deleted during the next storage warning.

Dark green chalkboard with chalk arrows and doodled stars, a black-and-white halftone cutout of a worn paper punch card with a hole punch in front
The competition your loyalty app has to beat costs almost nothing and never asks for a password.

A punch card skips steps three, four and most of five, because it lives in a wallet and you hand it over. So does a program tied to the register, where a phone number at checkout does the remembering. That is why the question to ask an app salesperson is not how many features it has. It is how many of your customers finish that list.

Cheaper roads to the same free coffee

Loyalty options for a very small shop, from what each vendor published as of November 2026
What the customer doesHonest weaknessCost to you
Paper punch cardCarries it, hands it overLost cards, no data, easy to fake with a hole punch from any office drawerPrinting
Square LoyaltySigns up at the register, on your Square site or through an invoice; rewards can be set by visits, spend or itemsFits only if you run your sales through SquareSquare advertises a free 30-day trial and no locked-in contracts; plan prices are on its pricing page
Apple Wallet store cardTaps a link in an email or on the web and adds the pass; no app neededNeeds an Apple developer account and signing, and something to issue the passes and update the balance$99 a year for Apple's program, plus a small web build
Custom web app with a loginSaves a page, shows a code at the counterNeeds staff to scan or type something every timeFrom $575, depending on complexity
iPhone or Android loyalty appEvery chore on the list aboveUpkeep, store review and the question of who opens it next monthFrom $3,750, plus store fees

The Wallet row deserves a second look. Apple's Wallet pages, read November 6, 2026, list store cards for loyalty cards and points, say a pass can be added without installing any related app, and say passes can go out in an app, by email or on the web. For a small shop, that may be the nearest thing to a free lunch here. The customer gets something that lives on the phone. You skip the store entirely. It still has to be built and signed, which is a small web application job rather than an app project. We have not priced that as a package, so ask.

The rare shop where a loyalty app has a real case

Build only when the reward depends on something the register cannot see. Say a one-location self-serve car wash in Wichita, invented for this example, that sells monthly memberships, tokens, and a wash that starts from a phone at the bay. The app is not a coupon. It operates the product. People will open it because they cannot get a wash otherwise. Apple's test, guideline 4.2, asks for features and lasting usefulness beyond a repackaged website, and this has some.

  • Visit frequency is high. Weekly or more. Rewards that arrive twice a year cannot hold a slot on anyone's home screen.
  • The app does a job besides the reward. It orders, starts the machine, holds the membership or shows the class schedule.
  • You are not on a system with loyalty built in, or that system cannot express your rule.
  • Enough regulars exist to matter. Count them. A couple of hundred people who come weekly is a different business from a couple of hundred who came once.
  • You can name who answers the complaint when the points are wrong, and it is a real person with free time.

If only the first and last are true, you are probably still in punch card country, and the kindest thing we can do is say so. Our list of app ideas for service businesses includes reasons not to build each one, and this category is high on it.

Screens, a ledger, and the cashier at 7:40 in the morning

When the case holds, the build is plain. Read this as a scope we would quote against. It has three customer screens and one for staff.

The card

A code that changes every minute or so, the current balance and how far to the next reward. Opens straight to this. A static code can be screenshotted and shared with a whole office, which is funny exactly once.

Rewards

What you can claim now, plus a plain expiry date. Rules written in the language you would use at the counter.

History

Every point earned and spent, with dates. This one screen prevents most arguments, and answers the question before the customer has to ask you.

Staff scan

A phone or tablet at the register that reads the code and adds the purchase in about a second. If it takes longer, staff will quietly stop using it by the end of the first rush.

The data is a ledger, not a balance. Each visit writes a line: who, when, how many points, which staff phone, and why. The balance is the sum. That design lets the owner reverse a mistake on a Tuesday without guessing, and it shows a staff member who has been giving a friend free points. It also means you can change the reward next spring without losing anyone's history. Points that customers can spend may have an accounting or legal treatment, so ask your accountant before they become a large number.

The owner's work is small but constant: glance at redemptions weekly, end the offer that nobody claims, and fix disputes the same day. If that sounds like more of a job than you want, a card is still right there. And if your actual problem is that customers forget you exist, automated follow-ups or the three email sequences worth having may do more for less than any card or app.

Three loyalty features that can land you in trouble

The second trap is the text blast. A program that collects phone numbers tempts everyone to send specials from them. Telemarketing calls made with an autodialer to wireless numbers require prior express written consent, under 47 CFR 64.1200 as published by Cornell LII. The same rule says consumers can revoke consent with words like "stop," that the revocation rules cover texts as well as calls, and that a request must be honored within 10 business days.

Twilio's documentation, read November 6, 2026, says anyone sending through an application from a 10-digit number to the US must register, and a Sole Proprietor Brand gets one Campaign. A phone number given for points is not automatically permission to market. We are not lawyers, so ask your attorney what your signup wording needs to say.

The third is the store itself. Guideline 4.2.2 says apps, other than catalogs, should not primarily be marketing materials or advertisements. A loyalty app that is a coupon and a menu is leaning on that line. Nobody can promise App Store approval, and we certainly do not. There are also fake-review and fake-follower bans in the FTC's August 14, 2024 final rule, so no feature that rewards people for inflating your numbers belongs in any version.

Test the idea with cardboard before anyone writes code

  1. 1

    Pick one reward with a real margin

    Work out what the free item costs you. If giving it away hurts, the program is a discount in disguise.

  2. 2

    Print cards and run it for a season

    Long enough for the slow regulars to finish one. Write a date on each card when you hand it out.

  3. 3

    Count what comes back

    How many cards were finished, and how quickly. Compare it with the customers you already recognized. That count, not the app salesperson's slide, is your forecast.

  4. 4

    Then choose the rung

    Few cards back: stop, and be pleased you spent a little. Many cards back and you run on Square: try its program. Many back and a rule nothing else handles: ask about a build.

If the cards come back and the idea survives, read what an iPhone app costs before talking to anyone, us included. Mobile projects can be spread out through app financing: 0%, 25% down, up to 12 months. DataCram is a web studio in San Antonio that has been building for the web since 1999, and it built and runs its own software rather than only selling other people's. We will quote it, and we will also tell you when the cardboard won. If the part of the business that really hurts is that new customers cannot find you, a free site audit is a better first step than any reward program, and the section's starting page lists the rest.

Questions owner-operators ask

Almost never as a first step. Frequency is high, so the idea sounds right, but a punch card or a register-based program skips the download and the login. Run a card for a season and count how many come back finished. Consider an app only if the count is large and the app also handles ordering or memberships your register cannot.

Yes. Apple's Wallet pages, read November 6, 2026, list loyalty and points cards as store cards, and say a pass can be added without installing a related app and can be distributed by email or on the web. It must be signed with a certificate from an Apple developer account. Square also says its loyalty customers can add a pass to Apple Wallet.

No. Google's Maps policy says merchants may not offer incentives such as payment, discounts or free goods or services for a review. The FTC's rule on incentives is different, but Google's policy still governs reviews on Google. We are not lawyers; ask your attorney about rules that apply to you. Ask every customer for an honest review and offer nothing for it.

A store app starts at $3,750, depending on scope, with Apple's $99 a year and Google's one-time $25 as of November 2026. A web app version starts at $575, depending on complexity. Then count the running costs: hosting, updates when phones change, text fees if you send any, and the owner's time on disputes. A card has far fewer lines.

Not without checking consent first. FCC rules, as published by Cornell LII, require written consent for autodialed telemarketing calls to wireless numbers, and their opt-out rules cover texts, including replies like stop. Software-sent texts from a 10-digit number also require registration, per Twilio. A number collected for points is not automatically permission to market. Take your signup wording to your own attorney, because we are not lawyers.

It can be copied. A screenshot of a fixed code works for whoever holds it, so one member can quietly earn points for an entire office. A code that refreshes, tied to a signed-in account, closes most of that gap. On a paper card the same risk exists with a hole punch, which is one more reason to watch the numbers rather than the technology.

Keep reading

Free · No obligation

Tell us what would bring customers back

Describe your regulars, your register and the reward you have in mind. A person at DataCram replies with the cheapest thing likely to work, and that may be a printed card.

Which of these sounds like you? (pick any)

Rather talk now? Call (210) 346-0848.

CallGet a quote