LinkedIn Marketing & B2B Lead Generationn8nAirtableClaude AIPerplexity AIGoogle SheetsRapidAPI

Automated Lead Capture & Qualification System

Growlancer

Went from 20 hours/week of manual lead research to processing 5,000–10,000 leads per list automatically.

20h

Weekly Hours Saved

10,000

Leads Per List

5–6

Qualified Meetings/Week

Concurrent Lists

Overview

Growlancer is a LinkedIn marketing agency that needed to automate AI-powered lead qualification at scale. Their manual process and a broken Make.com workflow couldn't handle large lead lists — causing memory errors and throttling. A new system was built from scratch that could process thousands of leads concurrently.

The Problem

The team manually researched leads one by one using Perplexity AI. A previous Make.com automation failed due to no batching support, memory overflow on large lists, and uncontrolled AI token burn. The result: 20 hours/week of manual effort and no path to scale.

The Solution

A multi-stage pipeline built on n8n using Google Apps Script as a controlled batching gateway. The system pulls 30 leads at a time, processes them in sub-batches of 6 through Perplexity AI, Claude AI, and RapidAPI — then loops until the full list is complete.

Stage 1 — Trigger & Initialization

When a campaign row in Airtable is set to 'To Process', it fires a webhook to n8n. The workflow creates a dedicated Google Sheet per campaign from a Drive template and injects tracking columns via Google Apps Script.

Stage 2 — Team Notification

Before processing begins, a notification is sent with campaign details, qualification criteria, and the template sheet link — no manual monitoring needed.

Stage 3 — Batched Lead Qualification

A Google Apps Script acts as a controlled gateway, returning only 30 unprocessed leads at a time. Each batch of 30 runs in sub-batches of 6 through: Perplexity AI (research) → Claude AI (qualification) → RapidAPI (LinkedIn activity check).

Stage 4 — Loop & Completion

After each batch, a second Apps Script counts remaining unprocessed leads. If any remain, the loop continues. Once complete, Airtable is updated to 'Leads Processed' and totals are logged.

Results

MetricBeforeAfter

Manual hours/week20 hours~0 hours

Concurrent lists1 at a time5–6 simultaneously

Lead list capacityMemory errors5,000–10,000 leads

Lead researchManual, one by oneFully automated

Weekly qualified meetingsInconsistent5–6 per week

Tech Stack

Orchestrationn8n

Campaign ManagementAirtable

Data GatewayGoogle Apps Script

Lead StorageGoogle Sheets

Research AIPerplexity AI

Qualification AIClaude AI

Activity CheckRapidAPI (LinkedIn)

Key Takeaways

Using Google Apps Script as a batching gateway solved n8n memory overflow — a pattern applicable to any large-dataset workflow.

Sub-batching at 6 items balanced API rate limits across 3 different services without manual throttling.

Per-campaign Google Sheet templates kept client data isolated and shareable without cross-contamination.

Want something similar for your firm?

Let's talk about what a custom automation system could look like for your operations.

Book a Discovery Call