How a Telegram Bot Replaced Two Sales Managers

This case is about a window installation company. Not the most tech-forward business, but it's exactly in these "ordinary" businesses that automation delivers the most visible results.

The Initial Situation

Before working with us, the company received about 40 applications per day from various sources: their website, VKontakte, Avito, and phone calls. Two sales managers processed these applications manually: calling each person back, clarifying parameters (what windows, which floor, installation needed, budget), entering data into the CRM, and handing off to the measuring team.

The problems were systemic:

• Response time to an application averaged 2-4 hours. Competitors responded faster.
• Evening and night applications waited until morning.
• Managers spent 60% of their time on non-target leads — people just curious about prices with no intention to buy.
• CRM data was filled in incompletely — each manager had their own style.

The owner calculated: if the application volume grew to 80 per day (and they were actively investing in advertising), they'd need to hire two more managers. That's about 150,000 rubles per month in additional payroll затраты.

What We Did

We developed a Telegram bot that connected to AmoCRM via API and became the first point of contact for every new application.

How it works:

• An application arrives — from any source. Via webhook, it enters our system.
• The bot messages the client within 30 seconds: "Hello! I'll help you quickly find the right windows. Tell me, what kind of room do you need windows for?"
• Qualification dialogue. The bot asks 5-7 questions: room type, number and size of openings, profile preferences, installation needed, budget. Questions adapt based on answers — not a decision tree, but live LLM-based conversation.
• Classification. Based on the dialogue, the system assigns the lead a status: cold, warm, or hot.
• Handoff to manager. Hot leads go immediately to the manager with a complete card: what the client wants, their budget, convenient time for measuring. The manager sees everything and calls prepared.
• Cold leads receive automatic messages with useful information — catalog, work examples — and stay in a nurture sequence until they're ready to buy.

Results After Three Months

The data is real; numbers are rounded at the client's request.

• First response time: from 2-4 hours to 30 seconds.
• 80% of applications are qualified without a live manager's involvement.
• Conversion from application to measuring visit grew by 22% — managers now work only with warm leads.
• Application volume grew to 65 per day (advertising was boosted), but the same two managers handle it.
• CRM is filled automatically and standardized — all fields are in place.

The two additional managers that were planned — they weren't needed. The automation investment paid off in 2.5 months.

What to Understand

The bot doesn't replace humans in sales — it makes sure humans only do sales work. Live communication, handling objections, complex cases — that stays with the manager. Routine work — qualification, data collection, first contact — that's the system's job.

This approach scales. Application volume can double — the system handles it, manager workload stays the same. That's the main value of automation: not reducing headcount, but the ability to grow without linear cost increases.

Applicability to Other Businesses

This approach works anywhere there's an incoming application flow and repeating qualification questions: construction and renovation, legal services, medical clinics, educational centers, auto repair shops. If your managers ask clients the same questions every day — it can be automated.