How to Build an AI Lead Qualification System with Make (Integromat) in 2026
The Biomog Team
May 31, 2026
26 min read
Make Tutorials
![AI lead qualification workflow diagram showing Make Integromat scenario with GPT-4o integration and CRM routing](https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhYDwiPu7_n2Bqe1S7ZwE2Ffo9kxyiLOLRaCJMfLjKm8uv5GlCeuKJ-Tkf-c4-TwkTZfZ35mB5tfehyphenhyphenwAGA4wOBrvopRVgtdrEsWYITv9sdzAXSevhzzmULa3q6MZovBM3Cxtifj5y-nmVui41d6RhhLNHnR1L21Sq9FkA-ET92WKsoeNfsVU_3oy1vYaAJ/s473/How%20to%20Build%20an%20AI%20Lead%20Qualification%20System%20with%20Make%20%28Integromat%29%20in%202026.PNG"" "AI Lead Qualification with Make and GPT-4o 2026")
Build an AI-powered lead qualification system with Make (Integromat) that automatically scores, routes, and nurtures leads using GPT-4o. Complete step-by-step tutorial with copy-paste scenarios, no coding required. Perfect for small businesses looking to automate sales in 2026.
73% of leads never get followed up on. Not because sales teams are lazy — because they are drowning. Inbox zero is a myth. CRMs are graveyards of unqualified contacts. And every minute spent manually scoring a lead is a minute not spent closing one. In this tutorial, you will build an AI-powered lead qualification system using Make (Integromat) and GPT-4o that captures, scores, routes, and nurtures leads automatically — turning your sales pipeline into a self-driving machine.
1 Why Manual Lead Qualification Fails
Before we automate, let us diagnose the disease. Here is what happens in a typical small business when a new lead arrives:
- Lead fills out a form on your website
- Notification email sits unread for 4 hours
- Sales rep opens CRM, reads the submission
- Rep googles the company, checks LinkedIn, estimates budget
- Rep scores the lead mentally (Hot/Warm/Cold)
- Hot leads get a call … eventually
- Warm leads get added to a spreadsheet … forgotten
- Cold leads get deleted or ignored
- Lead goes cold, competitor responds first
- Repeat 50 times per week
Time per lead: 8–15 minutes. At 100 leads per month, that is 13–25 hours of manual triage. Worse, human scoring is inconsistent. Monday morning scores are optimistic. Friday afternoon scores are brutal. Your best rep and your newest rep score the same lead differently.
GPT-4o scores leads in under 3 seconds with 85–92% accuracy when trained on your ideal customer profile. It never has a bad day. It never forgets a criteria. And it works 24/7 while your sales team sleeps.
2 What We Will Build: System Overview
Here is the complete system we are building today:
| Stage | Action | Tool/Module | Time Saved |
|---|---|---|---|
| Capture | Ingest lead from any source | Make Webhook / Form | 2 min/lead |
| Score | AI analysis with GPT-4o | OpenAI Module | 5 min/lead |
| Route | Route by score (Hot/Warm/Cold) | Router + Filters | 3 min/lead |
| Nurture | Auto-email sequence for Warm/Cold | Email / ActiveCampaign | 4 min/lead |
| Notify | Alert sales team for Hot leads | Slack + SMS | 1 min/lead |
| Sync | Update CRM with full context | HubSpot / Pipedrive / Airtable | 2 min/lead |
Total time saved per lead: 17 minutes. At 100 leads per month, that is 28 hours reclaimed — nearly a full work week every month.
A Make account (free tier works for testing), an OpenAI API key (GPT-4o access), a CRM (HubSpot, Pipedrive, or Airtable), and a Slack workspace. Total setup time: 35–45 minutes.
3 Prerequisites and Setup
3.1 Make Account
Sign up at make.com. Choose your plan:
| Plan | Operations | Cost | Best For |
|---|---|---|---|
| Free | 1,000 ops/month | $0 | Testing, <50 leads/month |
| Core | 10,000 ops/month | $9/month | Small business, 100–200 leads |
| Pro | 40,000 ops/month | $16/month | Growing business, 500+ leads |
| Teams | 80,000 ops/month | $29/month | Agencies, multiple scenarios |
Recommendation: Start with the Free tier to test. One lead qualification uses approximately 8–12 operations. 100 leads = 800–1,200 ops.
3.2 OpenAI API Key
- Go to platform.openai.com
- Create an account or log in
- Navigate to API Keys → Create new secret key
- Copy the key (starts with
sk-) — you cannot see it again - Add a payment method (GPT-4o costs ~$0.005–$0.015 per lead scored)
3.3 Required Connections
- Make → OpenAI (pre-built connector)
- Make → Your CRM (HubSpot, Pipedrive, Zoho, or Airtable)
- Make → Slack (for team notifications)
- Make → Email/SMS (for nurture sequences)
Store your OpenAI API key in Make Connections, never in plain text within modules. Make encrypts credentials at rest. Never share scenario exports containing real API keys.
4 Step 1: Capture Leads from Any Source
Make can ingest leads from virtually any source. For this tutorial, we will use a Webhook — the most flexible method. Any form, landing page, or tool can POST data to this URL.
Create a New Scenario
Log into Make, click Create a new scenario (top right). Name it
AI Lead Qualification — GPT-4o Scoring.
Add the Webhooks Module
Search for Webhooks and select Custom webhook. Click Add to create a new webhook. Give it a name and click Save. Make will generate a unique URL like:
https://hook.make.com/abc123def456...
Test the Webhook
Click Run once. Send a test POST request using curl, Postman, or your form. Make will capture the data structure automatically.
curl -X POST https://hook.make.com/YOUR_WEBHOOK_URL \
-H "Content-Type: application/json" \
-d '{
"name": "Sarah Johnson",
"email": "sarah@techcorp.com",
"company": "TechCorp Inc",
"job_title": "VP of Operations",
"company_size": "50-200",
"budget_range": "$5,000-$10,000",
"message": "Looking for automation solutions to streamline our invoice processing",
"source": "website_contact_form"
}'Connect Typeform, JotForm, Gravity Forms, or your website’s native form to this webhook. Even Facebook Lead Ads, LinkedIn Lead Gen Forms, and Calendly can POST here. One webhook, infinite sources.
5 Step 2: AI Scoring with GPT-4o
This is the brain of the system. We will send lead data to GPT-4o and receive a structured score, reasoning, and recommended action — all in under 3 seconds.
Add the OpenAI Module
Click the + icon after your webhook. Search for OpenAI and select Create a Completion (or Create a Chat Completion for GPT-4o).
Configure the AI Prompt
Set the following parameters:
- Model:
gpt-4o - Messages: System + User (see prompt below)
- Temperature:
0.3(low = more consistent scoring) - Max Tokens:
500
You are an expert B2B lead qualification analyst. Score each lead on a scale of 0-100 based on the following criteria:
SCORING RUBRIC:
- Job Title (20 pts): C-level = 20, VP/Director = 15, Manager = 10, Other = 5
- Company Size (20 pts): 200+ = 20, 50-200 = 15, 10-50 = 10, <10 = 5
- Budget Fit (25 pts): Matches our range = 25, Close = 15, Unknown = 10, Low = 5
- Message Intent (25 pts): Clear need + timeline = 25, Vague interest = 15, Just browsing = 5
- Source Quality (10 pts): Demo request = 10, Contact form = 7, Newsletter = 5, Unknown = 3
OUTPUT FORMAT (JSON only):
{
"score": 0-100,
"category": "HOT" | "WARM" | "COLD",
"confidence": 0-100,
"reasoning": "Brief explanation",
"recommended_action": "Immediate call" | "Email nurture" | "Add to newsletter",
"estimated_deal_size": "$X,XXX",
"urgency": "High" | "Medium" | "Low"
}
Rules:
- HOT = score 75-100 (contact within 15 minutes)
- WARM = score 50-74 (nurture sequence)
- COLD = score 0-49 (newsletter, long-term nurture)
- Be conservative. A lead needs clear signals to score above 80.
- Always return valid JSON. No markdown, no explanations outside JSON.Lead Data:
Name: {{1.name}}
Email: {{1.email}}
Company: {{1.company}}
Job Title: {{1.job_title}}
Company Size: {{1.company_size}}
Budget Range: {{1.budget_range}}
Message: {{1.message}}
Source: {{1.source}}
Analyze this lead and return JSON scoring only.GPT-4o is 2× faster and 50% cheaper than GPT-4 Turbo for this use case. It handles structured JSON output reliably at $0.005 per 1K input tokens. Scoring one lead costs approximately $0.003–$0.008. At 1,000 leads/month, your AI bill is under $8.
5.1 Parse the JSON Response
Add a JSON module after OpenAI to parse the response. Use
Parse JSON with the OpenAI output as source. This creates structured variables
you can use in subsequent modules: score, category,
recommended_action, etc.
6 Step 3: Route Leads by Score
Now we route leads based on their AI score. Make’s Router module (the flow-control icon) lets you create conditional paths.
Add a Router
Click the wrench/flow-control icon and select Router. This splits your scenario into multiple paths.
Set Up Filters
Click the wrench between router paths to set filters:
- Path 1 (HOT):
score ≥ 75 - Path 2 (WARM):
score ≥ 50 AND score < 75 - Path 3 (COLD):
score < 50
| Category | Score | Response Time | Action |
|---|---|---|---|
| 🔥 HOT | 75–100 | < 15 minutes | Instant Slack alert + SMS + CRM assignment |
| 🌡️ WARM | 50–74 | < 4 hours | Email nurture sequence + CRM tagging |
| ❄️ COLD | 0–49 | 24–48 hours | Newsletter subscription + long-term drip |
Responding to leads within 5 minutes increases contact rates by 900% versus 30 minutes (MIT study). Responding within 15 minutes increases qualification rates by 400%. Your AI system responds in 3 seconds.
7 Step 4: Auto-Nurture Warm and Cold Leads
Not every lead is ready to buy today. But ignoring them means losing future revenue. We will build automated nurture sequences that keep your brand top-of-mind.
7.1 Warm Lead Nurture (Score 50–74)
Add Email Module
On the WARM path, add an Email module (Gmail, Outlook, or SMTP). Configure a personalized welcome email.
Subject: {{1.name}}, I noticed your interest in [solution]
Hi {{1.name}},
Thanks for reaching out from {{1.company}}! I noticed you're exploring
automation solutions for {{1.message}}.
I put together a 3-minute video showing exactly how teams like yours at
{{1.company_size}} companies are saving 15+ hours per week with automated
workflows.
[Video Link]
Would a quick 15-minute call next Tuesday or Wednesday work to discuss
your specific needs?
Best,
[Your Name]
P.S. Based on your profile, I'd estimate a potential ROI of ${{3.estimated_deal_size}}
in your first quarter.7.2 Cold Lead Long-Term Nurture
For COLD leads, add them to a newsletter or low-touch drip sequence. Use ActiveCampaign, Mailchimp, or ConvertKit modules in Make to add contacts to specific lists.
Industry data shows 80% of sales require 5+ touchpoints. A lead that scores COLD today may become HOT in 3–6 months. Automated nurture ensures you are there when they are ready — without manual effort.
8 Step 5: Notify Sales Team
For HOT leads, speed is everything. We will send instant notifications to Slack and optionally via SMS.
Add Slack Module
On the HOT path, add a Slack → Create a Message module.
Select your workspace and target channel (e.g., #hot-leads).
🔥 *HOT LEAD ALERT — Score: {{3.score}}/100*
*Name:* {{1.name}}
*Company:* {{1.company}}
*Title:* {{1.job_title}}
*Email:* {{1.email}}
*Budget:* {{1.budget_range}}
*Urgency:* {{3.urgency}}
*Confidence:* {{3.confidence}}%
*AI Reasoning:* {{3.reasoning}}
*Recommended Action:* {{3.recommended_action}}
*Est. Deal Size:* {{3.estimated_deal_size}}
⏰ *Response target: 15 minutes*
🔗 *Lead source:* {{1.source}}
_Automated by Biomog AI Lead System_ 🤖Optional: Add SMS Alert
For ultra-hot leads (score ≥ 90), add a Twilio or ClickSend module to send an SMS to your sales manager. This ensures no HOT lead goes unnoticed.
Create a dedicated #hot-leads channel in Slack. Pin a message explaining
response time expectations. Use Slack’s reminders feature to follow up
on unclaimed leads after 30 minutes.
9 Step 6: Sync to CRM with Full Context
Every lead — HOT, WARM, or COLD — gets synced to your CRM with complete AI scoring data. This gives your sales team full context before they ever pick up the phone.
Add Your CRM Module
Make supports 3,000+ apps. Search for your CRM:
- HubSpot: Create/Update Contact + Create Deal
- Pipedrive: Create Person + Create Deal
- Airtable: Create Record
- Zoho CRM: Create Lead
- Google Sheets: Add Row (budget-friendly option)
// Map these fields from your Make variables:
Email: {{1.email}}
First Name: {{1.name}} (use split function if needed)
Company: {{1.company}}
Job Title: {{1.job_title}}
Lead Source: {{1.source}}
AI Score: {{3.score}}
AI Category: {{3.category}}
AI Confidence: {{3.confidence}}%
AI Reasoning: {{3.reasoning}}
Recommended Action: {{3.recommended_action}}
Estimated Deal Size: {{3.estimated_deal_size}}
Urgency: {{3.urgency}}
Last AI Scored: {{now}}Create a custom Lead Score property in HubSpot/Pipedrive. Use Make to update this score dynamically. Then build lists and workflows in your CRM based on AI score thresholds — completely automated.
10 Complete Scenario Blueprint
Here is the complete Make scenario structure. In Make, click More → Import Blueprint to paste JSON directly.
After importing, you MUST configure your own connections (OpenAI, CRM, Slack) and replace placeholder values. Never share blueprints containing real API keys.
{
"name": "AI Lead Qualification — GPT-4o Scoring",
"flow": [
{
"id": 1,
"module": "webhooks:CustomWebhook",
"metadata": {
"designer": { "x": 100, "y": 100 }
}
},
{
"id": 2,
"module": "openai:CreateAChatCompletion",
"metadata": {
"designer": { "x": 300, "y": 100 }
},
"parameters": {
"model": "gpt-4o",
"messages": [
{
"role": "system",
"content": "You are an expert B2B lead qualification analyst..."
},
{
"role": "user",
"content": "Lead Data:\nName: {{1.name}}\nEmail: {{1.email}}\nCompany: {{1.company}}\nJob Title: {{1.job_title}}\nCompany Size: {{1.company_size}}\nBudget Range: {{1.budget_range}}\nMessage: {{1.message}}\nSource: {{1.source}}"
}
],
"temperature": 0.3,
"max_tokens": 500
}
},
{
"id": 3,
"module": "json:ParseJSON",
"metadata": {
"designer": { "x": 500, "y": 100 }
},
"parameters": {
"data": "{{2.choices[0].message.content}}"
}
},
{
"id": 4,
"module": "tools:FlowControl",
"type": "Router",
"metadata": {
"designer": { "x": 700, "y": 100 }
}
}
]
}Note: This blueprint is abbreviated for readability. The complete production-ready JSON with all routes (HOT → Slack + CRM, WARM → Email + CRM, COLD → Newsletter + CRM) is available in the downloadable template pack below.
11 Testing and Calibration
11.1 Test with Sample Leads
Before going live, test with 10–20 sample leads covering all score ranges:
| Test Lead | Expected Score | Expected Category |
|---|---|---|
| CEO, 500+ employees, $50K budget, urgent need | 90–100 | HOT |
| Manager, 50 employees, $5K budget, exploring | 55–70 | WARM |
| Intern, 5 employees, no budget, just curious | 10–30 | COLD |
11.2 Calibrate Your Prompt
If scores feel off, refine your rubric. Common adjustments:
- Too many HOTs? Raise score thresholds or tighten criteria
- Too many COLDs? Lower thresholds or weight budget less heavily
- Inconsistent? Add more specific examples to your system prompt
- Wrong deal sizes? Add industry-specific pricing context to prompt
11.3 Monitor and Iterate
Review AI scores weekly against actual conversion outcomes. Track:
- What % of HOT leads actually converted?
- What % of WARM leads became HOT after nurture?
- Any COLD leads that should have been WARM?
Use this data to refine your prompt monthly. Your AI gets smarter as you train it.
AI scoring is a tool, not a replacement for judgment. Always have a human review edge cases. Set up a weekly report of all scored leads for your sales manager to audit.
12 Advanced AI Variations
Once your basic system is running, level up with these upgrades:
12.1 Multi-Source Lead Enrichment
Before scoring, enrich lead data with Clearbit or Apollo.io. Get company revenue, employee count, tech stack, and funding status — automatically. More data = more accurate scoring.
12.2 Competitor Detection
Add a prompt instruction: “Flag if the lead mentions a competitor (Zapier, n8n, Workato). If competitor mentioned, reduce score by 15 points and add tag ‘competitor-aware’.”
12.3 Sentiment Analysis
Use GPT-4o to analyze the tone of the lead’s message. Urgent language (“ASAP”, “need this week”) boosts score. Passive language (“just looking”, “maybe someday”) reduces it.
12.4 Predictive Deal Timeline
Ask GPT-4o to estimate time-to-close based on lead signals. Route “closes in 30 days” to your top closer. Route “6+ months” to long-term nurture.
12.5 A/B Test Your Prompts
Create two parallel scoring paths with different prompts. Route 50% of leads to each. After 30 days, compare conversion rates and keep the winner. This is prompt engineering as a science.
In late 2026, AI agents like Relevance AI and 11x.ai will handle entire sales conversations — not just scoring. The system you build today is the foundation for fully autonomous sales pipelines tomorrow.
13 ROI and Time Saved
Let us talk numbers. Here is the ROI of AI lead qualification for a typical small business:
| Metric | Before AI | After AI | Impact |
|---|---|---|---|
| Time per lead scored | 12 minutes | 0 minutes (automated) | 12 min saved |
| Monthly time (100 leads) | 20 hours | 0 hours | 20 hours saved |
| Annual time | 240 hours | 0 hours | 240 hours saved |
| Cost at $50/hour | $12,000/year | $108/year (Make Core) | $11,892 saved |
| Lead response time | 4–24 hours | 3 seconds | 99.9% faster |
| Conversion rate | 8–12% | 15–20% | +50–75% lift |
| Scoring consistency | Variable (human bias) | 85–92% accurate | Standardized |
It is not just time saved — it is revenue gained. A 50% lift in conversion rate on 100 leads/month at $2,000 average deal value = $100,000+ in additional annual revenue. For a $108/year Make plan + ~$100/year in OpenAI costs. That is not an ROI. That is a business transformation.
Get the Complete AI Lead Pack
This tutorial covers the foundation. The Complete AI Lead Qualification Pack includes the full Make blueprint JSON, 5 optimized GPT-4o prompts, email nurture templates, Slack message templates, CRM field mappings for HubSpot/Pipedrive/Airtable, and a 20-minute video walkthrough.
Download Free AI Lead Pack — No Credit Card14 Frequently Asked Questions
15 Final Thoughts: The AI Sales Revolution Starts Now
You now have a complete, AI-powered lead qualification system built with Make and GPT-4o. Let us recap what you accomplished:
- Captured leads from any source via webhook
- Scored leads automatically with GPT-4o in under 3 seconds
- Routed HOT leads to sales instantly via Slack + SMS
- Nurtured WARM leads with personalized email sequences
- Added COLD leads to long-term nurture campaigns
- Synced everything to your CRM with full AI context
This is not just automation — it is augmented intelligence. Your sales team now spends 100% of their time on leads worth talking to. No more manual triage. No more “I forgot to follow up.” No more guessing.
The businesses that adopt AI lead qualification in 2026 will have an insurmountable competitive advantage by 2027. The ones that don’t will be left manually scoring leads while their competitors scale infinitely.
Your Next Steps
Download the Complete AI Lead Qualification Pack for the full Make blueprint and templates, read our guide on n8n vs Make vs Zapier for sales automation, and subscribe to the Biomog Weekly newsletter for new AI automation tutorials every week.
Start Make Free
