WebApp in Telegram: when it helps and when it hurts
A WebApp is an ordinary web page that opens inside Telegram on top of the chat. To the user it looks like an app: a storefront with pictures, filters, a cart, a calendar, a dashboard. To the developer it is a website that knows how to talk to a bot. Nothing has to be installed and nobody has to register, since Telegram itself tells the app who opened it.
When a WebApp is needed and when it gets in the way
Whether you need one comes down to a single test: does the choice fit into buttons. Five services fit, and that is an ordinary menu. Two hundred products with sizes, colours and stock levels do not, because paging through them with buttons is painful and searching among them is impossible. As soon as a person needs to compare options, see several things at once or drag something around, buttons run out and a WebApp begins.
There is a flip side. If the scenario is linear, meaning you need to ask four questions and take an enquiry, a WebApp only gets in the way. It takes longer to open, needs a better connection, behaves worse on older phones and adds a step where there was none. A button-based bot turns out faster, cheaper and more reliable. And here is a small thing people forget: an ordinary bot can be answered with a voice message or a photo straight from the chat, and a mini app offers no such option.
Sign-in without registration, and the main mistake
What a WebApp gives you almost for free is the user's identity. Telegram tells the app who opened it and signs that data. No registration, no passwords, no email confirmation, so the most expensive step of the funnel simply disappears. Sometimes a WebApp is taken on for that reason alone.
The main mistake lives in the same place. The signature has to be verified on the server with a key derived from the bot token. The data arrives from the client, and the client is a browser where anything can be substituted. Without the check, anyone can open your API and claim somebody else's identifier along with their orders and their balance. It is easy to skip, because without it everything works right up until the day someone tries.
Payments inside a mini app
Payments inside a WebApp come with a fork in the road that is better known before you start. Digital goods and subscriptions inside Telegram are sold for Stars. This is the built-in route, it opens access immediately after payment and requires nothing from the owner beyond deciding the prices. However, when a user buys stars in a mobile app, the Apple and Google stores keep around a third. Physical goods and services are taken by ordinary payment providers directly and without that loss, but connecting one requires a registered business entity.
Fitting the design to Telegram, and its limits
A separate piece of work that is regularly underestimated is making the app not look like a stranger inside the messenger. Telegram hands over its own theme, meaning the background, text and button colours, and those change when the user switches from light to dark. An app that ignores them looks like somebody else's site pasted in. Add to that your own bottom button, your own header and your own back-navigation logic. All of it needs not just building but genuinely fitting to the shell.
There are limits too. A mini app lives inside the messenger, so some of what an ordinary website can do is unavailable or available with caveats. A link to such a page will not appear in search, since it only opens through the bot. On weaker phones the difference between buttons and a full interface is felt immediately.
Combining a bot and a WebApp, and what it costs
In practice it is more often the combination that wins rather than the choice. Enquiries, questions, notifications and statuses are best handled with ordinary messages, because that is fast and works everywhere. The catalogue, the cart and checkout belong in a WebApp. Both scenarios live in one bot and one database, so the user never notices the boundary and you do not pay for an interface where a button would do.
The cost difference is noticeable and quite explicable. A WebApp means developing a full interface on top of the bot, that is, a site and a bot in one project. This is why a Telegram store takes 3 to 5 weeks while a lead capture bot fits inside a week, and the two differ severalfold in price. If you are unsure whether you need a WebApp, the cheapest way to find out is to launch on buttons and watch where exactly people stumble. Usually that makes it clear whether an interface is needed at all, or whether catalogue search would have been enough.