CRM bot for guide: automation of booking and client return system
I have developed a comprehensive solution for a private guide that transforms Telegram into a full-fledged order management system. The main goal of the project is to ensure that no client gets "lost" during the excursion selection or booking stage.
What I implemented (Technical part)
In this bot, I focused on tracking user actions and automatic data collection:
Interactive excursion showcase: Cards with photos and descriptions, where the selection of each item is recorded as a specific interest of the user.
Deep tracking: I set up a system of tags and variables that mark each step of the client. If a person clicks "Sign up" but does not select a time, the bot sees it.
Automation through Make: I configured complex logic for data transfer. All clicks, selected routes, and contacts are instantly sent to Google Sheets.
Smart "nudge" scenarios: I implemented reminder logic. The bot automatically messages those who showed interest but stopped halfway, helping to lead them to booking.
Analytics of frequently asked questions: The system records which questions are asked most often, helping the project owner optimize content and address objections in advance.
Review gallery with a trigger: A scrollable review section that at the right moment (at the peak of trust) offers a sign-up button.
Project architecture
The bot is built on the ManyChat platform using Make as the "brain" for connecting with external tables. I attached screenshots of the funnel to show how the transition conditions and data recording are set at each stage.
Result for the business
This solution addresses the main problem of private specialists — the loss of applications due to slow responses.
Value of the solution:
Transparent funnel: It is clear at which stage people "drop off," allowing for sales improvement.
Client retention: Thanks to tags, the bot reminds clients of itself, increasing the number of bookings without spending on advertising.
Time savings: The guide does not need to manually transfer data to the table — the system does it for them.
Database for mailings: All interests are segmented — promotions can be sent only to those who showed interest.
What I implemented (Technical part)
In this bot, I focused on tracking user actions and automatic data collection:
Interactive excursion showcase: Cards with photos and descriptions, where the selection of each item is recorded as a specific interest of the user.
Deep tracking: I set up a system of tags and variables that mark each step of the client. If a person clicks "Sign up" but does not select a time, the bot sees it.
Automation through Make: I configured complex logic for data transfer. All clicks, selected routes, and contacts are instantly sent to Google Sheets.
Smart "nudge" scenarios: I implemented reminder logic. The bot automatically messages those who showed interest but stopped halfway, helping to lead them to booking.
Analytics of frequently asked questions: The system records which questions are asked most often, helping the project owner optimize content and address objections in advance.
Review gallery with a trigger: A scrollable review section that at the right moment (at the peak of trust) offers a sign-up button.
Project architecture
The bot is built on the ManyChat platform using Make as the "brain" for connecting with external tables. I attached screenshots of the funnel to show how the transition conditions and data recording are set at each stage.
Result for the business
This solution addresses the main problem of private specialists — the loss of applications due to slow responses.
Value of the solution:
Transparent funnel: It is clear at which stage people "drop off," allowing for sales improvement.
Client retention: Thanks to tags, the bot reminds clients of itself, increasing the number of bookings without spending on advertising.
Time savings: The guide does not need to manually transfer data to the table — the system does it for them.
Database for mailings: All interests are segmented — promotions can be sent only to those who showed interest.