How Proface cut first-response time to 41 seconds with a chatbot

Introduction

Proface is a cosmetology clinic. Clients write on WhatsApp and Instagram, and records are maintained through Altegio and Message Help. Previously, messages were processed manually, so a response could take from 5 to 15 minutes. We have set up a chatbot that works in the same systems, quickly responds to clients and helps with recording

The clinic lost speed on the first touch

It was important for the clinic to respond to messages faster, reduce the risk of missed calls, and relieve some of the burden on administrators. At the same time, they didn’t want to switch to a new CRM - they needed to keep the familiar tools

Messages from WhatsApp and Instagram were sorted manually. Because of this, responses took 5–15 minutes, and administrators had to search for information on the client and their bookings and services themselves. When there were more requests, it was more difficult to work consistently

Illustration for the section “The clinic lost speed on the first touch”

How to prepare the bot for launch

The project was not limited to simply installing a bot. First, we collected and structured information about the clinic’s work: services, typical client questions, recording, transfer and consultation scenarios

After this, we prepared a knowledge base, collected the bot logic and developed integration with Message Help for real client processes

Before launching into combat channels, the solution was tested and fine-tuned many times so that the bot worked correctly in conversations, did not confuse scenarios, and gave the highest quality first answer. When the logic was debugged, the bot was connected to the clinic’s main communication channels

Illustration for the section “How the bot was prepared for launch”

Connected a chatbot without changing the system

It was important for the clinic to speed up responses to clients on WhatsApp and Instagram, but without switching to a new CRM and without disrupting current processes. Therefore, the bot was built into the already familiar combination of Altegio and Message Help. As a result, the team retained a convenient work system, and some requests were processed automatically

Now the bot not only responds to messages, but can work with client context. It finds a person by phone number, uses data from past records and helps to quickly navigate the situation. This is especially important for repeat clients, when you do not need to start a conversation from scratch, but continue communication taking into account previous visits

It answers routine questions and hands complex ones to a person.

Not all requests should remain on the bot. If a client wants to cancel or reschedule an appointment, such a request is immediately sent to the administrator.

This helps not to confuse the client with unnecessary scenarios and quickly transfer the conversation to where the person is already needed

Illustration for the section “Where necessary, he answers himself; where necessary, he passes it on to the person”

The bot can answer typical questions about services, costs and procedure options on its own. This reduces the burden on administrators and helps the client receive a response faster without waiting.

Due to this, the team spends less time on repetitive questions and more time on live work with clients

Results in numbers

64% of requests were taken by a botThe clinic closed most of the standard messages without the participation of the administrator

41 seconds until first reactionClients began to receive a response almost immediately after contacting them

The bot works 24/7 without days off or breaksThis helps not to lose requests in the evening, at night and during rush hours

Withstands any flow of requestsWhen there are more messages, the bot continues to respond without overload or delay