See llms.txt for all machine-readable content.

Back to Templates

Qualify, route, and follow up on inbound leads with Sheets, Slack, Telegram, and GPT-4o

Inbound leads pile up in a form inbox and the good ones wait as long as the bad ones. This workflow scores every lead with GPT-6.1-sol, routes it by tier, alerts sales on Slack and Telegram for the hot ones and keeps a clean pipeline sheet. It also catches repeat inquiries, parks unreliable AI scores for a human, sends cold leads a polite reply and chases hot leads nobody has touched after a day.

Last updated: November 2026.

Quick Overview

A webhook receives the lead, a Set node normalises the fields and an IF node rejects leads without an email, company or message. Valid leads are matched against the Leads tab so a repeat inquiry from the same email in the last 30 days is flagged on Telegram instead of being scored twice. An AI agent scores the rest from 1 to 10 with a category, summary, recommended action and follow up question. A Code node checks that reply, and malformed or out of range scores go to a Review Queue tab with a Slack ping. Trusted scores are appended to the pipeline and a Switch routes them: hot leads alert Slack and Telegram and get logged as alerted, warm leads get a nurture note, and cold leads get a short honest reply by Gmail. Twice a day an SLA lane reminds sales about hot leads that are still untouched after 24 hours, and every Monday a digest goes out on Telegram and email.

How it works

  1. Lead Form Submitted, Normalize Lead Fields and Validate Lead Data take the POST. Invalid leads go to Flag Invalid Lead, Notify Team About Invalid Lead and Log Rejected Lead to Sheet.
  2. Read Existing Leads and Check Duplicate Lead compare the email with rows from the last 30 days. Duplicate Lead? posts repeats to Telegram and stops.
  3. Score Lead with AI Agent runs with retries, and Extract AI Score pulls the JSON fields.
  4. Check Score Quality needs a score from 1 to 10, a known category and action, and a summary. Score Trustworthy? sends failures to Queue Lead For Manual Review and Ask Sales To Review Lead.
  5. Append Lead to Pipeline Sheet stores the lead, then Route By Lead Tier splits hot (7 and up), warm (4 to 6) and cold.
  6. Hot leads go to Notify Sales on Slack and Telegram, then Log Hot Lead Alert sets the status to HOT_ALERTED with the alert time and Telegram message id. Warm leads go through Mark Lead for Nurture. Cold leads get Send Cold Lead Reply.
  7. Hot Lead SLA Check runs at 09:00 and 15:00 on weekdays. Find Unworked Hot Leads picks alerted leads with no contacted_at after 24 hours, Remind Sales On Slack nudges the team and Mark SLA Reminder Sent stops repeat nags.
  8. Weekly Pipeline Digest computes hot, warm, cold and invalid counts plus the average score and sends them on Telegram and email.

Setup

  1. Connect your OpenAI credential on Lead Scoring Model and pick GPT-6.1-sol or the model you prefer.
  2. Connect Google Sheets and use one spreadsheet with a Leads tab and a Review Queue tab. Leads needs the columns from the append node plus alerted_at, telegram_message_id, contacted_at and reminder_sent_at.
  3. Connect Slack and pick the sales channel. Connect Telegram and replace 123456789 with your chat id.
  4. Connect Gmail and replace [email protected] in the digest node.
  5. Point your form at the Lead Form Submitted production URL and send a test lead.

Quick Answers

What happens when the same person submits the form twice?
Check Duplicate Lead finds their email in the last 30 days, the team gets a Telegram note with the existing lead id and the lead is not scored again.

What if the AI returns broken JSON or a strange score?
Check Score Quality marks it NEEDS_REVIEW, the lead lands in the Review Queue tab and Slack asks someone to score it by hand.

How do hot leads avoid being forgotten?
They are logged as HOT_ALERTED, and the SLA lane reminds Slack once if contacted_at is still empty 24 hours later.

Do cold leads hear anything back?
Yes. Send Cold Lead Reply sends one short, honest email and the lead stays in the sheet.

Additional info

Built with n8n. Need an assessment on your business? Feel free to reach out at https://khmuhtadin.com/consultation/