AI Outbound Sales Stack
Automate prospecting, enrichment, and personalized outreach at scale
The complete AI-powered outbound sales workflow that finds prospects, enriches their data with AI, crafts personalized messages, and automates follow-ups. This stack replaces 3-4 SDRs worth of manual work and typically increases reply rates by 40-60%.
Your Progress
0 of 20 tasks completed
Tools You'll Need
Step-by-Step Setup Guide
Connect your lead sources to Clay for automatic enrichment. Clay can pull leads from LinkedIn Sales Navigator, Apollo, or CSV uploads.
Create a Clay account and new table
Go to clay.com → Sign up → Create new table → Name it "Outbound Prospects"
Connect your lead source
Click "Add People" → Choose LinkedIn Sales Navigator, Apollo, or Upload CSV
Add enrichment columns
Click "+" to add columns → Select "Company Info", "Email Finder", "LinkedIn Profile"
Enable AI Research column
Add "AI Research" column → Prompt: "Research this person's recent LinkedIn posts, company news, and potential pain points for [your product category]"
Set up webhook trigger
Go to Integrations → Webhooks → Create outbound webhook for new enriched rows
Use Clay's "Waterfall" enrichment to try multiple email finders (Hunter, Apollo, Clearbit) automatically until one succeeds
LinkedIn Sales Navigator has rate limits. Start with 50-100 leads/day to avoid account flags
Create a Zapier/Make workflow that triggers when Clay enriches a new lead, generates a personalized email, and adds it to your CRM.
Create new Zap with Webhook trigger
New Zap → Trigger: Webhooks by Zapier → Catch Hook → Copy webhook URL
Paste webhook URL in Clay
Go back to Clay → Paste the Zapier webhook URL → Test with sample data
Add ChatGPT action for email generation
Add Step → ChatGPT → "Conversation" action → Use this prompt structure (see next step)
Configure the AI email prompt
System: "You are a B2B sales email writer. Write short, personalized cold emails (under 100 words)." User: "Write a cold email to {{name}} at {{company}}. Their role: {{title}}. AI Research: {{clay_research}}. My product: [describe your product]. Pain point to address: [your value prop]"Add CRM action
Add Step → HubSpot/Salesforce → Create Contact → Map fields: Email, Name, Company, Generated Email (store in notes)
Add email sequence enrollment
Add Step → HubSpot → Enroll in Sequence → Select your outbound sequence
Test with 5-10 leads first before scaling. Check email quality and deliverability
Add a filter step to only process leads with verified emails (email_status = "valid")
Build a 4-5 touch email sequence in your CRM with the AI-generated first email and templated follow-ups.
Create new sequence
HubSpot → Automation → Sequences → Create Sequence → Name: "AI Outbound v1"
Add first email step
Add Step → Email → Use personalization token {{contact.notes}} to pull AI-generated emailAdd follow-up emails
Add 3-4 follow-up emails spaced 3-4 days apart. Templates: Day 3: Value-add content, Day 7: Case study, Day 10: Breakup email
Set unenrollment triggers
Settings → Unenroll when: Reply received, Meeting booked, Contact unsubscribes
Enable send windows
Settings → Send window → Business hours only (9am-5pm recipient's timezone)
Never send more than 50 new outbound emails per day per inbox to maintain deliverability
A/B test your AI prompts. Try different angles: pain-point focused vs. benefit-focused vs. curiosity-driven
Track key metrics and continuously improve your AI prompts and targeting.
Set up sequence analytics dashboard
HubSpot → Sequences → Analytics → Track: Open rate, Reply rate, Meeting rate
Review AI email quality weekly
Spot-check 10-20 generated emails. Flag any that feel generic or miss the mark
Iterate on AI prompts
If reply rate < 5%, adjust prompt. Add more specific pain points or change the tone
Expand winning segments
When you find a high-performing ICP (>10% reply rate), double down on similar leads in Clay
Industry benchmarks: 50%+ open rate, 5-10% reply rate, 1-2% meeting rate
Create a "best performing emails" library and feed them back to ChatGPT as examples