Telegram Mini App for Restaurants: How to Launch a Loyalty Program Without App Store Tax

Telegram now counts 1 billion monthly active users as of March 2025 (Pavel Durov; aggregated by Backlinko/Semrush) and 500 million people opening it every single day as of October 2025 (Backlinko/Semrush, citing Telegram's own channel) — the most recent official figures available as of this writing, with no newer milestone announced since. Numbers like that make a headline promising to "launch a loyalty program without App Store tax" sound like a shortcut every restaurant marketer should take: build inside Telegram, skip Apple's cut, pocket the difference.

Here's the catch: Apple has never taken a commission on restaurant food orders. Guideline 3.1.3(e) of the App Store Review Guidelines has long required physical goods and real-world services to be paid for through methods other than in-app purchase — that's not a 2026 change, it's how the rule has always worked. So what does a Telegram Mini App loyalty program actually save a restaurant, and does it make sense for your guests at all? We'll work through both, with real numbers from two dev.family loyalty builds and one Telegram Mini App we shipped ourselves — and this isn't a pitch for every restaurant to build in Telegram, just a way to spot whether yours is one where it genuinely fits.

A Telegram Mini App is a full web application that runs inside Telegram through a bot — no separate listing in the App Store or Google Play, and no install beyond the Telegram app your guest already has open.

Technically, it's a WebView loaded through a bot registered in BotFather, Telegram's bot-management tool. The Telegram Web App API handles authorization through initData instead of a login form — Telegram already knows who your guest is. Because it's a web app under the hood, one codebase can run across iOS, Android, desktop, and web without separate native builds. That's the key difference from a bot: a bot works through text commands, while a Mini App gives you a real interface — menus, buttons, forms — the way a website or app would.

Restaurants tend to test loyalty first, not ordering or delivery, because points and a digital card are the simplest feature to validate before committing to anything bigger. We've built Telegram Mini Apps ourselves — including Clicker, a tap-to-earn game rewarding users with points, giveaways, and promotions — so we've seen how fast a simple, gamified mechanic inside Telegram generates real usage data.

For more on why businesses outside foodtech are moving into Telegram in the first place, see our earlier piece on Telegram Web App opportunities for businesses, and for the build itself, our walkthrough of Telegram Mini App development and testing.

Book a short session with our dev team to launch your tax-free Mini App

Anna S., Business Development Manager

Write in LiveChat

Write in LinkedIn

Apple never charged a commission on food. The 30% (or 15% for smaller developers) in-app purchase fee has only ever applied to digital goods and subscriptions — physical goods and services consumed outside the app, including restaurant orders, ride-hailing, and event tickets, are carved out under guideline 3.1.3(e), which requires those purchases to run through a method other than in-app purchase, like Apple Pay or a card processor. A restaurant app has always been free to charge for food via Stripe, PayPal, or Adyen without Apple touching it.

So the real cost of the native path isn't a commission — it's the development and distribution overhead:

• Parallel builds. iOS and Android are two separate codebases (unless you're already cross-platform), which usually means double the engineering time and QA cycle per release.
• Store fees. The Apple Developer Program runs $99 a year, plus a comparable one-time registration fee to publish on Google Play.
• Review cycles. We typically plan for a one-to-three-week Apple review window per update — a Mini App update goes live the moment you deploy it, no queue at all.
• Install conversion. Most guests won't download a dedicated app for one restaurant. A loyalty program that requires a fresh install starts with a conversion penalty a Mini App doesn't have, since guests are already inside Telegram.

None of this argues against native apps — chains that already justify the investment by scale get real value from owning that channel (see how we approached a 50-location chain in Franchise App Development). It's an argument for being precise about what you're actually saving, instead of framing the project around a commission food was never subject to. For where loyalty fits into the bigger stack, see Restaurant Tech Stack 2026.

Only about 9% of US respondents said they use Telegram regularly, per Statista Consumer Insights survey data fielded between July 2023 and June 2024 and published in Backlinko's country breakdown — the most recent US-specific figure available, since no fresher independent survey exists yet. BankMyCell separately puts US all-time downloads at roughly 26.85 million, 4th globally behind India, Russia, and Indonesia, based on 2022–2023 data (page last updated April 2024). For comparison, Pew Research Center's 2025 survey of 5,022 US adults found 32% now use WhatsApp, up from 23% in 2021 — over three times as common in the US as Telegram, by the best available data for each.

That doesn't make Telegram irrelevant — it makes it a channel for specific segments, not a mass-market default:

1. Restaurants serving Telegram-active communities — Russian-, Ukrainian-, and Central Asian-diaspora audiences, parts of several Asian communities in large US metros, and crypto-adjacent younger diners all skew toward regular Telegram use well above the national average.
2. Chains testing a low-cost pilot before native investment — if you're not ready to commit a native-app budget to an unproven loyalty hypothesis, a Mini App is a cheap way to test the mechanic first.
3. Brands with an existing Telegram channel or community — if you already have guests engaging with a Telegram channel or group, a Mini App adds a transactional layer on top of an audience you've already earned, rather than trying to build one from zero.

Where it doesn't make sense: a mass-market restaurant with no Telegram-active audience, where the effort spent explaining "what Telegram is and how to use it" outweighs whatever you save on app-store overhead. Our piece on why loyalty programs fail in restaurant franchises covers the more common ways loyalty rollouts go wrong regardless of channel, and if you're weighing a specialist foodtech partner against a generalist shop for this kind of decision, we wrote about that trade-off directly in dev.family vs. Generic Agency.

The mechanics are the same ones you'd expect from any loyalty program — points per purchase, tiers, coupons — the difference is that they live inside a chat app your guest already has open, with no plastic card and no separate install.

In practice, that usually means:

• A digital loyalty card as a QR code inside the Mini App, functioning like an Apple Wallet or Google Wallet pass without requiring either.
• Points and tiers that update in real time once the Mini App is synced to your POS.
• Bot messages instead of native push notifications — Telegram Mini Apps don't get iOS- or Android-style push, so the bot itself becomes your notification channel for promotions and reminders.
• Referral loops built into Telegram's own sharing — a guest forwards the Mini App link into a group chat, and a new user is one tap away from opening it, no App Store search required.

A word on where this comes from. None of what follows is a Telegram Mini App built specifically for restaurant loyalty, and we want to be upfront about that. What we have is two relevant halves: gamified loyalty mechanics proven in a native retail app, and Telegram Mini App experience from a different industry.

Beerpoint, a beer retail chain, is where we tested the gamification mechanic that translates directly into a Mini App: a Wheel-of-Fortune-style reward spin layered onto an existing points program. After launch, items per order rose from about 2.5 to roughly 3, monthly active users grew by about 40%, and average bill size increased by about 10%. The mechanic itself — spin, points, instant reward — works the same whether it's native or inside a Telegram WebView.

John Dory, a seafood retail chain, shows the other half: replacing a plastic punch card with a digital one tied to real purchase data. The app has driven more than 8,000 new users a month and reached #2 in the App Store's Food & Drink category — proof that ditching the physical card alone moves the needle on enrollment.

And Clicker, the Telegram Mini App mentioned earlier, confirms the tap-to-earn, points-and-rewards mechanic works natively inside Telegram's WebView, with no native app underneath it at all.

For more mechanics worth adapting into a Telegram-native format, see The Perfect Retailer Loyalty Program App and 9 ways retail apps improve engagement and retention.

The Perfect Retailer Loyalty Program App

Personalized rewards, gamification, and digital wallet features — a full playbook for retail loyalty apps.

Read the article

The right channel depends on three things: how much of your guest base is already on Telegram, whether you want native development now or want to validate first, and whether you need an in-store, point-of-purchase moment mobile options don't cover.

If your audience is already on Telegram, start there — it's the cheapest way to test the mechanic before a native build. If they're mostly on iOS/Android with little Telegram presence, a PWA or a native app sized to your scale fits better. If loyalty needs to be captured at the register, pair either with a self-ordering kiosk instead of expecting a phone-based channel to cover that moment.

A basic MVP — points, a digital card, and POS integration — typically takes 4 weeks to a few months, depending on complexity and the number of integrations. Here's the sequence:

1. Register a bot in BotFather. This is your entry point — the Mini App opens from inside this bot, not from a store listing.
2. Scope the MVP tightly. Points plus a digital card is enough for version one. Save gamification (spins, tiers, streaks) for after you've validated basic usage.
3. Pick your stack and follow Telegram's design guidelines. React works well — the same front-end approach we use for cross-platform builds, so transactional logic (points, wallet, history) ports cleanly.
4. Integrate with your POS so points are credited automatically at the moment of purchase, not manually or on a delay.
5. Test inside Telegram's own sandbox environment before any guest sees it.
6. Apply for Telegram Mini Apps Verify — a trust badge — and launch.

For deployment details, see Deploying a Telegram Mini App. And before locking in what "points" means for your margins, read Smart Restaurant Loyalty Programs That Boost Profit Without Killing Margins — the channel is only half the equation; reward economics are the other half.

You can't sell food through Telegram Stars — that's Telegram's currency for digital goods, not physical ones. For a restaurant, payment runs through an external processor (Stripe is the common choice) integrated into the Mini App, at standard card-processing rates.

Telegram Wallet Pay and crypto payments exist as options, but they're not the primary rail for a restaurant selling food — Stripe or an equivalent processor handles it the same way it would on your website or in a native app. That's also why a Mini App has no platform "tax" on food to begin with: the only cost is the standard 2.5–3% (plus a small fixed fee) any restaurant already pays for card processing elsewhere — not a Telegram fee, and not an Apple one either, per the same guideline 3.1.3(e) covered above. For the fundamentals, see Web Payments: What Do You Need to Know? — most of it applies directly to a Mini App's payment flow. If you're evaluating who should build that integration, our comparison of specialist vs. generalist foodtech developers covers what tends to go wrong without foodtech-specific payment experience.

Most failed Telegram loyalty launches are a channel-selection mistake made before development even starts, not a technical one.

• Skipping the audience check. Launching in Telegram without first confirming what share of actual guests use it is the single most common mistake — see the reality check above before you commit a budget.
• No POS integration. If points are credited manually or with a delay, guests notice immediately, and trust in the program erodes fast.
• Expecting native-style push notifications. They don't exist in Telegram Mini Apps. Plan your bot-messaging strategy as a first-class part of the build, not an afterthought.
• Over-building gamification in the MVP. Points and a digital card are enough to validate the idea. Spin wheels and streak mechanics can come once you know people are actually using the basic version.

Beerpoint and John Dory both make the same point from different angles: even a simple mechanic — points, tiers, one gamified spin — produces a measurable result when it's wired into real purchase data. The technology underneath matters less than getting that connection right. For a wider view of AI and personalization reshaping loyalty and local commerce, see GEO vs. SEO for FoodTech.

GEO vs SEO: How FoodTech Companies Can Get Recommended by ChatGPT, Claude, and Perplexity

How FoodTech brands actually get recommended by ChatGPT, Claude, and Perplexity.

Read the article

• Apple has never charged a commission on restaurant food orders — physical goods and real-world services have always been exempt from in-app purchase rules, so there's no "tax" to escape in the first place.
• The real savings from a Telegram Mini App are development and distribution costs: no parallel iOS/Android builds, no store fees, no review queue, and a lower barrier to first use since guests are already in Telegram.
• Telegram is a niche channel in the US — about 9% of adults use it regularly, versus 32% for WhatsApp — so this makes the most sense for specific audiences, not as a universal replacement for your core loyalty system.
• The loyalty mechanics themselves (points, tiers, digital cards) don't change based on the channel; what changes is install friction and notification strategy.
• Food and physical goods still get paid for through a standard processor like Stripe, at normal card-processing rates — Telegram Stars is for digital goods only.
• Start with a tight MVP (points plus a digital card, synced to POS) before adding gamification, and confirm your audience is actually on Telegram before you build anything.