Telegram bot for collecting leads
The bot asks the questions that decide whether an enquiry is worth a call and writes the answers into your CRM as fields. A manager gets a record, not a greeting in a chat.
- Buildfrom $750
- Timeline1–3 weeks
- Supportfrom $420/mo
Who this suits: Ads and posts that point at a direct message, where a manager works out the budget, the timing and the city from scratch in every thread.
What is in the first version
- question script
- source tag
- a CRM record
- manager alert
The figures above are the package this sits on: Telegram bot. Anything past its edges the calculator adds before the work starts, not after.
How the path goes
- 01
The enquiry arrives already labelled
A link from an ad or a post opens the bot with a tag that travels into the CRM.
- 02
Only the questions that decide something
Task, timing, city, volume: buttons where the answers are finite, and the branch changes with each answer.
- 03
A record reaches the CRM
Each answer lands in its own field, the deal opens at the right stage, and the thread stays linked.
What decides whether this works or annoys people
The form outlasts the patience
Every extra question loses people, and the busy ones go first, so the first version asks only what routes an enquiry.
A no is still an answer
Silence on a request you cannot take becomes a review, so the bot declines in a sentence and saves the enquiry anyway.
The reply around the system
A reply from a personal account moves the deal into a private chat, so answers go through the bot or your screen.
What to measure once it is live
- dialogues that reach the end
- the question that loses people
- time to a human reply
- qualified enquiries per source
Why this comes out faster
The shape of this one is known: the states, the edge cases and the things that usually go wrong have been decided before. Nothing here is a template, and the saving is not in your half of the work. It goes into your process, your content and the systems this has to talk to, which is the part nobody can have solved in advance.
A likely stack for this
Picked against the task when we scope it, not decided in advance. This is the shelf it usually comes off.
- Bot and scriptPythonFastAPITelegram Bot API
- Data and the CRM linkPostgreSQLRedisn8nDocker
- Reading free text and watching it runOpenAIAnthropicGrafana
Price it yourself, right here
Five steps, and you can see the number without leaving a contact. The estimate accounts for the kind of work, what you already have and what it will need inside.
These bills do not come from us
- hosting or a VPS for the bot
- your CRM plan and its API limits
- AI model calls
- the ads that bring people in
Asked before the first call
- Do the enquiries reach our CRM?
- Yes: the bot creates the deal through the API, spreads the answers across fields and attaches the thread. Without a CRM the enquiries sit in their own table, and connecting one later swaps one adapter.
- Can we change the questions ourselves?
- Wording, answer options, their order and the decline change from an admin panel, without us. A new branch that depends on stock or team load is a code change, priced before we start.
- Is it a script, or does it understand free text?
- A script by default: predictable and it invents no terms. A model goes in where people write freely anyway, and what it is allowed to touch is settled before the start.