In Uzbekistan Telegram is not "another channel", it is where people spend the day. So "maybe a bot is enough" is a reasonable question, and often the answer really is yes.
This piece is not about which messenger to pick — there is a separate one for that. Nor about Mini Apps — also covered. This is about one decision: a bot instead of a website, or a bot alongside one.
What a bot does better than a site
-
It launches faster and costs less. At the start that decides it.
-
It asks nothing of the customer. Telegram is already open on their phone.
-
It lets you write first. Someone who once messaged the bot stays reachable — a website cannot do that.
-
Notifications arrive. Not to spam, not to a promotions tab.
For a small range, repeat orders and business by word of mouth, a bot can be enough for years. That is not a stopgap.
Where the ceiling starts
First and most important: a bot is not in search results. Somebody typing "buy X in Tashkent" will never reach your bot from Google. Only the people you hand a link to ever arrive. Which means every new customer costs you advertising money — a free stream will never appear.
Then, in increasing order:
-
Comparing things in a chat is awkward. Past a couple of dozen items, scrolling messages is worse than looking at a catalogue.
-
There is nothing to show a serious client. A link to a bot in a proposal does not read like a link to a site with registration details and case studies.
-
You are standing on somebody else's ground. The rules, the availability and the behaviour of the bot are not yours to set.
-
There is almost no analytics. Where an order came from and what the person looked at first is far harder to see in a bot.
How they work together
The arrangement that works is nearly always the same, and it is not either-or:
-
The site collects the people who are searching — from search, from ads, from a link in a conversation. It also answers "who are you" and "can you be trusted".
-
The bot takes the order and carries the conversation where the person finds it easy to reply.
-
Enquiries from the site and orders from the bot land in one place — otherwise half of them go missing.
How to tell it is time
Not by turnover, but by these signs:
-
You are paying for advertising and want a stream that does not stop when the budget does.
-
Customers ask you to "send a link, I'll show management".
-
You are tired of answering the same questions a page could have closed.
-
You are moving into B2B or into export.
If none of those is you, a bot is enough, and that is a perfectly good answer.
If a bot is already running and hitting its ceiling, building a site behind it goes faster than usual: you already know the ordering flow from practice.
