In partnership with
Look, I’m going to level with you.
While you’re over there manually qualifying leads, copying data between spreadsheets, and wondering why your sales team can’t keep up, there’s a quiet group of operators running circles around you.
They’re not smarter. They’re not working harder. They just stopped pretending that manual work builds character.
Here’s what they’re doing instead:
They built a three-tool automation stack that runs 24/7. No coffee breaks. No sick days. Just relentless lead processing, enrichment, and outreach while they sleep.
I’m going to walk you through exactly how this works, not with theory, but with the actual implementation that one of our clients, Marcus, used to 3x his pipeline in three weeks.
The Complete Stack Breakdown:
Tool #1: Make.com (The Orchestrator)
This is your workflow engine. Think of it as the conductor telling every other tool when to do its job. Cost: $9/month starter plan (enough for 10,000 operations). You’ll upgrade eventually, but start here.
Tool #2: Clay (The Intelligence Layer)
This pulls in data you’d never find manually. Company tech stack, recent funding rounds, employee count changes, news mentions, social activity. It’s basically having a research team working every lead. Cost: $149/month for 2,000 credits (each enrichment costs 1 to 3 credits).
Tool #3: ChatGPT API (The Writer)
This takes all that data and turns it into personalized outreach that doesn’t sound like a robot wrote it. Cost: About $20/month for 500 to 1,000 generated emails (at GPT-4 rates).
Total monthly cost: $178
Marcus used to pay a part-time VA $800/month to do what this stack does better and faster. You do the math.
Speak fuller prompts. Get better answers.
Stop losing nuance when you type prompts. Wispr Flow captures your spoken reasoning, removes filler, and formats it into a clear prompt that keeps examples, constraints, and tone intact. Drop that prompt into your AI tool and get fewer follow-up prompts and cleaner results. Works across your apps on Mac, Windows, and iPhone. Try Wispr Flow for AI to upgrade your inputs and save time.
How Marcus’s Workflow Actually Works:
Step 1: Lead Capture (The Trigger)
Marcus runs outbound to B2B SaaS companies. His team uses LinkedIn Sales Navigator to find prospects and exports them to a Google Sheet daily. Here’s where the automation kicks in:
In Make.com, he set up a scenario that watches that Google Sheet every 15 minutes. New row appears? Scenario fires immediately.
The exact Make.com modules:
1. Google Sheets > Watch Rows
2. Filter > Only proceed if “Status” column = “New”
3. Router (splits the workflow into three paths)
This takes about 5 minutes to configure. The Sheet needs columns for: Name, Company, LinkedIn URL, Status. That’s it.
Step 2: Enrichment (The Intelligence Gathering)
Now Clay enters the picture. Make.com sends the company name and LinkedIn URL to Clay’s API. Clay comes back with:
• Company size and growth rate
• Technologies they currently use
• Recent funding announcements
• Key decision makers and their roles
• Recent company news or product launches
• Industry category and sub-category
Marcus set up Clay to prioritize certain data points. For him, it’s tech stack (he wants to know if they’re using competitor tools) and recent funding (signals they have budget).
The exact Clay setup:
1. Create a Clay table
2. Add enrichment columns: “Find company info”, “Find technologies”, “Find recent news”
3. Use Clay’s Clearbit and BuiltWith integrations
4. Set up the API endpoint in Clay’s settings
5. Connect to Make.com using HTTP module
Clay returns this data in about 10 to 30 seconds per lead. Make.com waits for the response, then moves to step 3.
Step 3: Personalization (The Outreach Generator)
Here’s where ChatGPT takes over. Make.com sends all that enriched data to the ChatGPT API with a specific prompt Marcus refined over two dozen iterations.
Marcus’s Exact Prompt:
“You are a B2B sales expert writing on behalf of [YOUR COMPANY]. Using the information below, write a personalized 4-sentence outreach email.
Lead Information:
- Name: [NAME]
- Company: [COMPANY]
- Role: [ROLE]
- Company Size: [SIZE]
- Current Tech Stack: [TECH_STACK]
- Recent News: [NEWS]
Email Structure:
1. Opening: Reference their recent news or a specific pain point their industry faces
2. Value Prop: Connect how [YOUR PRODUCT] solves that specific problem
3. Social Proof: Mention a similar company you’ve helped
4. Low-friction CTA: Ask for 15-minute call, offer specific time slots
Tone: Professional but conversational. No buzzwords. No hype. Write like you’re a peer, not a vendor.
Do not use these phrases: ‘I hope this email finds you well’, ‘reaching out’, ‘circle back’, ‘touch base’, ‘quick question’.”
ChatGPT generates the email in 3 to 5 seconds. Make.com catches the response and moves to step 4.
Step 4: Delivery (The Follow-Through)
Make.com takes that generated email and does three things simultaneously:
1. Sends it to Marcus’s CRM (he uses Pipedrive) as a new lead with all the enriched data
2. Creates a draft in his email (connected via Gmail API) so he can review before sending
3. Updates the Google Sheet to mark the lead as “Processed”
Marcus doesn’t auto-send cold emails. He reviews each one, makes minor tweaks if needed, and hits send. This review process takes him about 15 seconds per email because the heavy lifting is done.
The Results (Real Numbers, Not Hype):
Before automation:
• 40 to 50 leads processed per week
• 6 hours spent on research and email writing
• 8% response rate (generic outreach)
• 2 to 3 qualified meetings booked weekly
After automation (Week 3):
• 500+ leads processed per week
• 2 hours spent reviewing and sending
• 24% response rate (personalized outreach)
• 12 to 15 qualified meetings booked weekly
Time saved: 18 hours per week
Cost: $178/month
Meetings increase: 5x
He closed two deals in the first month that paid for the tools for the next two years. Not a bad return.
Your 45-Minute Implementation Guide:
Phase 1: Setup (20 minutes)
1. Sign up for Make.com (use free tier to test)
2. Sign up for Clay (they have a 14-day trial)
3. Get ChatGPT API access at platform.openai.com
4. Create a Google Sheet with these columns: Name, Company, LinkedIn URL, Status, Date Added
Phase 2: Build the Workflow (15 minutes)
1. In Make.com, create new scenario
2. Add Google Sheets “Watch Rows” module, connect your sheet
3. Add Clay HTTP module, paste your Clay API endpoint
4. Map the company name from Google Sheets to Clay’s company field
5. Add OpenAI “Create Completion” module
6. Paste the prompt template above
7. Map Clay’s response data into the prompt variables
8. Add your CRM module (Pipedrive, HubSpot, Salesforce, whatever you use)
9. Map all fields from previous steps
Phase 3: Test (10 minutes)
1. Add a dummy lead to your Google Sheet
2. Run the scenario manually in Make.com
3. Check each step for errors
4. Verify the email output makes sense
5. Confirm the lead appears in your CRM
6. Adjust prompt if needed, test again
Once it works with dummy data, add 5 real leads and test again. Then turn on the automatic schedule (every 15 minutes) and let it run.
The 5 Problems Everyone Hits (And Exact Fixes):
Problem 1: Clay returns incomplete data
Why it happens: The company name isn’t exact or the company is too small to have public data
Fix: Add a filter in Make.com that checks if Clay found at least 3 data points. If not, send to manual research queue
Problem 2: ChatGPT emails sound robotic
Why it happens: Your prompt is too formal or you’re not giving enough context
Fix: Add examples of your best manual emails to the prompt as reference. Use the phrase “Write like [YOUR NAME] would write this” in the instructions
Problem 3: Scenario runs but doesn’t create CRM records
Why it happens: Field mapping is off or required fields are missing
Fix: Check your CRM’s required fields. Make sure every required field has a mapped value. Use “Set Variable” modules in Make.com to set defaults for fields you don’t have data for
Problem 4: Costs spike unexpectedly
Why it happens: You’re using GPT-4 for everything or Clay is enriching more than you need
Fix: Switch to GPT-3.5-turbo for first drafts (80% cheaper, 90% as good). Limit Clay enrichments to only what you actually use in outreach
Problem 5: Scenario times out or runs slowly
Why it happens: Too many operations in one scenario or API rate limits
Fix: Break into two scenarios. First one handles capture and enrichment, second handles email generation and CRM updates. Connect them with a data store in Make.com
What to Expect in Your First Month:
Week 1: Setup and testing. You’ll process maybe 50 leads while you iron out kinks. Your emails won’t be perfect yet. That’s normal. Focus on getting the workflow stable.
Week 2: Prompt refinement. You’ll notice patterns in which emails get responses. Adjust your ChatGPT prompt based on what’s working. You should be processing 150 to 200 leads by now.
Week 3: Scale mode. Workflow is humming. You’re processing 300+ leads. Response rates should be 15 to 20% if you’re targeting right. You’re spending more time on calls than on research.
Week 4: Optimization. Add conditional logic (if company is in X industry, use Y approach). Set up A/B testing on email angles. You’re at 400 to 500 leads processed with minimal hands-on time.
Success benchmark: If you’re not processing at least 200 leads per week by end of week 2, something’s wrong. Most likely it’s your lead source (not enough incoming leads) or your Clay setup (taking too long per lead). Fix those first before optimizing other parts.
The Advanced Move (For Month 2+):
Once you’ve got this running smoothly, add a follow-up sequence. Here’s how:
In Make.com, create a second scenario that triggers 3 days after the first email if there’s no response. It checks your CRM for “No Reply” status, then generates a follow-up using ChatGPT with this context:
“The recipient didn’t respond to this email: [ORIGINAL EMAIL]. Write a 2-sentence follow-up that adds new value. Reference a different pain point or share a relevant case study. Don’t say ‘just following up’ or ‘bumping this to the top’. Provide something new they care about.”
Marcus added this in week 5. His response rate jumped another 8% just from smart follow-ups that didn’t sound desperate.
Why This Works When Templates Don’t:
The difference between this and those “email template packs” you bought for $47? Context and timing.
Templates assume everyone’s the same. This system adapts to each lead. It references their actual situation, not a placeholder you forgot to replace.
And it happens immediately. No “I’ll send that tomorrow” pile. The workflow runs whether you’re in meetings, on vacation, or sleeping. Consistency wins.
Want the Full System?
What I just gave you is the core workflow. But Marcus’s complete setup includes:
• Multi-step nurture sequences based on engagement level
• Slack notifications for high-value leads (enterprise accounts, recent funding)
• Automatic calendar booking when prospects reply with interest
• Weekly performance reports (open rates, reply rates, meeting conversion)
• A/B testing framework for prompt variations
I’ve packaged the entire system, including cloneable Make.com scenarios, Clay enrichment templates, ChatGPT prompt library, and troubleshooting playbook inside the AI Workflow Blueprint ($47).
It’s the exact stack I use. The one Marcus uses. The one that processes thousands of leads without breaking a sweat. Comment ‘WORKFLOW’ and I will get you the Blueprint details.
You can keep doing this manually, or you can let the robots handle it. Your call.
Jordan Hale
The AI Newsroom
P.S. If you’re still managing your inbox manually, check out SuperHuman. They’re giving $80 credit to new users, and it’s the only email tool I’ve found that actually saves time instead of just reorganizing the chaos. Worth a look.