Selling Digital vs Physical Goods on Telegram

To sell digital products on Telegram or physical ones is the first fork in every Telegram shop plan, and the fork is priced differently than newcomers expect. Digital goods sell instantly to anyone on Earth — and pay out through a rail that takes roughly 35%. Physical goods keep full payment freedom — and add logistics, stock and returns to your week. This guide runs the honest math on a $10 product both ways, then helps you pick.

The 35% question: Stars economics

Telegram's rules are unambiguous: digital goods sold inside bots and Mini Apps go through Stars. The chain of costs from buyer to seller:

• The buyer pays in Stars; one Star runs about $0.013 at purchase.
• The platform's take on Stars transactions is ~35% of the buyer's price.
• Payouts convert to TON with a hold of up to 21 days before withdrawal.

On a $10 digital product: the buyer pays ~770 Stars; you net roughly $6.50; and the last of it lands in your TON wallet up to three weeks after the sale. To price for a $10 net, divide by 0.65 — the buyer then sees ~$15.40. The Stars calculator runs this arithmetic interactively; the Stars rules guide translates the official terms.

Is 35% fatal? Depends on margin. Software with near-zero marginal cost absorbs it. A reseller buying licenses at $8 and selling at $10 cannot — that business needs a different rail or a different product. The take is the price of: zero fraud surface, instant global delivery, and checkout inside the chat where the buyer already is.

Digital goods: instant delivery, thin margin

What counts as digital here: files (e-books, templates, presets, fonts), keys and licenses, paid access to channels and groups, subscriptions to content.

The honest operating model in this platform:

• The product card is created like any other — name, price, description, a cover image; the "delivery" is whatever you ship.
• Payment runs through Stars; the order lands in your cabinet with the buyer's contact.
• Delivery is yours: the file goes out by chat or cloud link, access is granted by invite. Instant automatic delivery is on the roadmap; today the seller is the delivery pipeline, which for most one-person digital shops is a copy-paste habit that scales fine into hundreds of orders.

Strengths: no stock, no shipping, no returns logistics, worldwide reach from day one, and the whole catalog fits in the free plan's fifteen products. Weaknesses: the 35% take, copyability (people share files), and near-zero switching costs for buyers — retention has to be earned with content, not logistics.

Physical goods: logistics, full rails

Physical products — apparel, crafts, food — run through Stripe or local acquirers, with ordinary acquiring fees (typically 1.5–3.5% + fixed for cards) instead of the 35% take, and money settling on the provider's normal schedule.

The $10 physical product nets roughly $9.60 after typical acquiring — $3.10 better than the digital rail. But the same week now contains: stock tracking (the platform counts units and alerts below three left), packaging, the post office or courier handoff, and returns. Variants exist for a reason — "Size: XL, +200 ₽" is how one card covers a size run — and the catalog tools page covers the structure.

The honest accounting is not per-transaction, it is per-hour-of-your-week: a physical shop with 40% margins and 5 hours of logistics per week competes with a digital shop with 65% margins and 20 minutes of fulfillment. Neither wins in the abstract; your costs decide.

Hybrid models that work

The fork is false. Proven hybrids:

• Physical with a digital companion — a craft store selling patterns alongside the finished items: the pattern is pure margin at 65%, the item is full price at 96% net, the same audience buys both.
• Digital product, physical funnel — a free physical sample (at cost) that pulls buyers into the channel where the digital catalog lives.
• Services as products — consultations and commissions sold as catalog items; delivery is a calendar slot, margins are digital-grade.

The platform's rails support all three today: each product simply carries its own payment module — Stars on the digital ones, a provider on the physical ones — configured in the same settings screen. The digital products industry page collects the category-specific rules; the flagship how to sell on Telegram guide places the choice inside the full launch sequence.

The decision tree

Four questions, answered in order, decide the fork for any product:

1. Does the buyer receive it inside Telegram? A file, a key, channel access, an in-chat unlock — yes means Stars, full stop; that is platform policy, not a pricing choice. No (a parcel, a chair, an in-person hour) — the provider rail.
2. Is the marginal cost near zero? Yes — the 35% take is survivable arithmetic: price by dividing the target net by 0.65 and the category's strengths (no stock, no shipping, worldwide reach) do the rest. No — you are reselling or producing at real cost, and the take eats the business: you need the provider rail, which almost always means the product must be physical.
3. Can your market buy Stars? If the target country cannot (Germany today; availability shifts with app-store policy), a digital product inside the bot is unsellable there — the honest moves are a physical companion product or routing that market to a web checkout outside the bot.
4. Can you stomach the hold? If cash flow needs the money this week, the TON settlement with up to 21 days of hold belongs in the plan — either price it in or lean physical until the float exists.

Every "which should I sell" conversation with a new seller ends somewhere on this tree, usually at question two. The Stars calculator turns the branch you land on into numbers in a few seconds.

The decision, in one paragraph

Sell digital when your marginal cost is near zero and your audience trusts you with a 35% take; sell physical when your product demands a body and your margins can pay for your logistics hours; sell both when the same audience wants the file and the object. Price digital goods by dividing your target net by 0.65 — never by copying the physical price — and rehearse the whole flow with one test order on the free plan before your first buyer does it for real.