DevPilot

DTC brand

DMs that sell themselves

From 500 DMs/month converting at 4% to 11% conversion and a 6× revenue lift.

Result
000×
Timeline
ROAS ×6.2
Category
DTC brand
What was happening

500 DMs/month handled manually, 4% conversion, managers can't keep up with peak hours.

A DTC clothing brand received 500 Instagram DMs per month: sizing, price, stock and shipping questions.

The manager couldn't physically keep up at peak hours. Purchase conversion was 4%, and some leads bled to competitors.

Their CRM (Bitrix24) wasn't integrated with Instagram — requests got lost, no follow-ups.

What we did

TG bot qualifies, books calls or closes with a promo code. Conversion rose to 11%.

  1. 01

    Built a Telegram bot with Instagram integration through a ManyChat-style pipeline on n8n.

  2. 02

    The bot qualifies on size / budget / timing. Based on customer type, it picks a path: book a consult, send a promo code, or hand off to a manager.

  3. 03

    GPT-4 answers off-script questions from the store knowledge base (200 SKUs + return policy).

  4. 04

    Each lead lands in Bitrix24 with tags and pipeline stage. The manager sees the full context and a draft reply.

What changed

Result

Purchase conversion: 4% → 11%

ROAS up from 1.8× to 6.2× in 2 months

Bot paid for itself on day 9 after launch

The bot now handles 1,200 DMs/month without growing the team

Stack
AiogramGPT-4Bitrix24MakePython

Next case

Support that doesn't burn out

Logistics

Want a project like this?

Want a project like this?

Talk to us