A founder I spoke to last quarter had been paying $4,200 a month for an SDR for six months. Total meetings booked: 11. Total cost per meeting: $2,290.
We rebuilt his outbound in n8n using Apollo and Instantly. Monthly cost: $245. Meetings in the first 60 days: 14.
The workflow isn’t complicated. But the order matters — and most people get the order wrong, which is why their automations produce lists instead of conversations.
Here’s the exact setup, step by step.
n8n Apollo Instantly Workflow: Complete Guide
What is the n8n Apollo Instantly Workflow?
Think of this workflow as your outbound operating system, where n8n is the central orchestrator connecting every tool via APIs.
n8n is the brain. It pulls data from Apollo. It verifies emails through ZeroBounce. It enriches leads using AI (Perplexity/Claude/ChatGPT). It pushes sequences to Instantly.ai, monitors replies, and routes everything to Slack and your CRM.
Here’s what makes this different from other automation setups: n8n connects tools that don’t natively integrate. Apollo doesn’t talk to ZeroBounce. ZeroBounce doesn’t talk to Perplexity. Perplexity doesn’t talk to Instantly.ai.
n8n makes them all work together in one automated flow.
Here’s the complete workflow architecture:
┌────────────────────────────────────────────────────────────┐
│ n8n │
│ (Central Orchestrator) │
│ │
│ Step 1: Pull leads from Apollo API │
│ Step 2: Verify emails via ZeroBounce API │
│ Step 3: Enrich leads via Perplexity/Claude API │
│ Step 4: Personalize sequences using AI responses │
│ Step 5: Push to Instantly.ai API │
│ Step 6: Monitor Instantly for replies │
│ Step 7: Send Slack webhook alerts │
│ Step 8: Sync to HubSpot/GoHighLevel API │
└────────────────────────────────────────────────────────────┘What Each Component Does
Apollo = Your lead database (275M contacts with filters)
ZeroBounce = Email verification service (removes invalid/risky emails)
n8n = Workflow orchestration platform (connects everything via APIs)
Perplexity/Claude/ChatGPT = AI enrichment engine (researches leads, generates personalization)
Instantly.ai = Email sending platform (manages warmup, sequences, deliverability)
Slack = Real-time notification system (alerts your team on replies)
GoHighLevel = CRM (stores all lead data and interactions)
Why Manual Outbound Is Breaking Down
The average cold email reply rate in 2026 sits at 5-8.5%. That means 92 out of every 100 emails get deleted, ignored, or marked as spam.
But the bigger problem: 30% of positive replies never get followed up on. Not because teams are lazy. Because replies arrive across multiple inboxes, CRMs don’t sync in real-time, and nobody’s monitoring dashboards at 11 PM.
Meanwhile, you’re paying $4,000/month per SDR for manual research, copy-pasted sequences, and inconsistent personalization.
The n8n Apollo Instantly workflow eliminates every friction point. It verifies emails before sending. It uses AI to personalize at scale. It routes every reply to your team in under 5 minutes via Slack. And it costs $245/month instead of $4,000.
The Complete Tech Stack
n8n: The Central Orchestrator
n8n is an open-source workflow automation platform that connects APIs from different tools. Think of it as the conductor — each tool (Apollo, ZeroBounce, Instantly, Slack, GoHighLevel) is an instrument, and n8n makes them all play together.
Why n8n instead of Zapier or Make.com? Free tier supports 5,000 workflow executions/month, unlimited API connections, open-source (self-hostable), and handles complex conditional logic other tools can’t.
Apollo: Lead Sourcing
Apollo is your lead database — 275M contacts filterable by job title, company revenue, employee count, hiring signals, tech stack, and funding rounds. n8n connects to Apollo via API to automatically pull leads matching your saved search criteria. Critical tip: export 100-500 leads max per campaign. Small, hyper-targeted campaigns consistently outperform massive blasts.
ZeroBounce: Email Verification
Apollo’s email accuracy is roughly 85-90%. That means 50-75 emails per 500-lead list will bounce. Three hard bounces can trigger most ESPs to flag your domain. ZeroBounce verifies every email against SMTP validation, catch-all detection, and spam trap removal. Cost: $0.01 per email. For a 500-lead list, that’s $5 to protect your domain reputation.
Perplexity/Claude: AI Enrichment Engine
This is where the magic happens. n8n takes each verified lead and sends it to an LLM with a structured enrichment prompt. The LLM scrapes the web, analyzes LinkedIn activity, reads company news, and returns structured data that n8n uses to personalize outreach. This level of personalization is why the workflow hits 15-25% reply rates vs. 5-8% industry average.
Instantly.ai: Email Sending Platform
n8n pushes enriched leads with personalized sequences to Instantly via API. Instantly handles email warmup (4.2M+ warmup network), sequence sending, deliverability optimization, and AI Reply Agent triage. Why Instantly specifically: unlimited email accounts, built-in warmup network, and a robust API that n8n can push leads to and monitor replies from programmatically.
Slack: Reply Alert System
n8n monitors Instantly API every 5 minutes for new replies. When a positive reply arrives, n8n triggers a Slack webhook with the lead’s name, company, reply text, enrichment data, and CRM link. Your team gets notified instantly and can respond in under 5 minutes — critical for conversion.
How to Build the Workflow: 7-Day Setup
Days 1-2: Account Setup + ICP Definition
Create accounts for: n8n.io (free tier), Apollo, ZeroBounce, Perplexity API or Claude API, Instantly.ai, Slack, and HubSpot or GoHighLevel.
Then define your ICP with specifics: exact titles, revenue range, employee count, location, tech stack signals, and hiring signals. Write it down — these filters drive your Apollo search.
Days 3-4: Apollo Lead Extraction + API Setup
Create a saved Apollo search matching your ICP. Then connect n8n to Apollo via API: go to Apollo Settings → API, copy your API key, create an HTTP Request node in n8n, and configure it to POST to https://api.apollo.io/v1/mixed_people/search with your saved search parameters.
Day 5: Build the n8n Workflow (8 Nodes)
Node 1 — Schedule Trigger: Runs daily at 9 AM automatically.
Node 2 — Apollo API: Pulls 50 leads matching your saved search. Outputs: Name, Email, Title, Company, LinkedIn URL.
Node 3 — ZeroBounce API: Verifies each email. Filters out invalid, catch-all, spam traps. Outputs only verified emails.
Node 4 — Perplexity/Claude API (Enrichment): Sends enrichment prompt with lead data. Receives: recent LinkedIn activity, company news, pain points, tech stack. Outputs: enriched JSON per lead.
Node 5 — Claude/ChatGPT API (Sequence Generation): Generates a 3-email personalized sequence using enrichment data. Outputs custom sequences per lead.
Node 6 — Instantly API (Push Campaign): Creates new lead in Instantly campaign with personalized sequence. Campaign activates automatically.
Node 7 — Instantly API (Monitor Replies): Separate workflow, runs every 5 minutes. Checks for new replies, filters for positive intent. Outputs: new positive replies.
Node 8 — Slack Webhook: Triggered when Node 7 finds a positive reply. Sends formatted message with lead name, company, reply text, enrichment context, and CRM link.
Node 9 — HubSpot/GoHighLevel API: Creates or updates contact. Logs all enrichment data and email activity. Full audit trail in CRM.
Day 6: Configure LLM Prompts
The quality of your enrichment determines your reply rates. Use structured JSON-output prompts that ask the LLM for: recent LinkedIn post summary, company news, pain points, tech stack, specific trigger event, and a personalized one-sentence hook. If data isn’t available, return “none” — don’t hallucinate.
For sequence generation: pass the enrichment JSON to Claude or ChatGPT with instructions to write a 3-email sequence using the trigger event and pain points. Specify founder-to-founder tone, 6-8 sentences per email, and return as JSON with subject/body pairs.
Day 7: Test With 10 Leads, Then Launch
Run the workflow manually with 10 test leads first. Check each step completes, verify enrichment quality, review generated sequences, confirm leads appear in Instantly, and test the Slack webhook. Once all steps pass, activate the daily schedule trigger and let the warmup run 7-14 days before ramping volume.
n8n Apollo Instantly vs. Manual SDR: The Real Comparison
| Factor | Manual SDR Process | n8n Workflow |
|---|---|---|
| Monthly Cost | $4,000+ (salary + tools) | $245 (all tools + API costs) |
| Setup Time | 60-90 days (hiring + ramp) | 7 days |
| Operating Hours | 8 hours (business hours only) | 24/7 automated |
| Lead Research | 10-15 min per lead | 30 seconds per lead |
| Email Verification | Rarely done | 100% verified before sending |
| Reply Monitoring | 2-3x daily inbox checks | Every 5 minutes |
| Reply Leakage | ~30% | 0% (Slack alerts) |
| Turnover Risk | High (9-12 month average) | None |
Common Mistakes to Avoid
Skipping email verification — Three hard bounces flag your domain. Always include the ZeroBounce node. It costs $5 per 500 leads.
Generic LLM prompts — Vague prompts return vague data. Use structured prompts that request specific JSON fields.
No error handling — APIs fail. Add n8n Error Trigger nodes that log failures, send Slack alerts, and continue with remaining leads.
Pulling 500 leads at once — Start with 50/day. Test. Optimize. Then scale.
Not monitoring LLM output quality — Add a Code node that validates LLM output before pushing to Instantly. If data is malformed, flag for review.
FAQ
Do I need to be technical? You need basic API understanding — copying keys, making HTTP requests. n8n is visual drag-and-drop. If you can follow documentation, you can build this.
Total monthly cost? Apollo ($49) + ZeroBounce (~$5) + Perplexity API (~$20) + Claude API (~$30) + Instantly ($97) + n8n (free) = $245/month.
Which LLM for enrichment? Perplexity for web research and recent data. Claude for analysis and writing quality sequences. Use both: Perplexity to enrich, Claude to write.
How many meetings to expect? Well-targeted campaigns with good enrichment typically generate 10-15 meetings per 500 leads sent (2-3% conversion). Depends heavily on ICP, offer, and messaging.
Want this built for you instead of doing it yourself? Six Figure Consulting builds custom n8n outbound workflows for B2B service companies. We configure everything — Apollo filters, n8n nodes, LLM prompts, Instantly setup, Slack routing, CRM sync — and test with 50 leads before handing it over.
📧 Weekly GTM Newsletter: Real workflow breakdowns, tools we’re testing, campaign data. Subscribe here →
Related: Clay vs n8n for AI Sales Automation
Leave a comment