Make.com AI Agent Workflow for Local Business: Complete Blueprint (2026)

What This Workflow Does

New form submission (Typeform / Gravity Forms / any webhook)
      ↓
  Extract: name, service interest, budget, location
      ↓
  AI qualification (is this a good fit?)
      ↓
  Good fit → Generate personalized email → Send via Resend → Log to Google Sheets
  Poor fit → Send polite "not right for us" email → Log reason
      ↓
  Notify owner via Slack/email only if lead is high-value or needs human input

Module-by-Module Setup

Module 1: Webhook Trigger

{
  "name": "Sarah Johnson",
  "email": "sarah@coffeeshop.com",
  "phone": "+15551234567",
  "business_type": "Coffee shop",
  "service_interest": "Google Ads + Local SEO",
  "monthly_budget": "$500-1000",
  "city": "Austin, TX",
  "message": "We opened 3 months ago and need help getting found on Google Maps"
}

Module 2: OpenAI / Groq — Lead Qualification

Authorization: Bearer {{GROQ_API_KEY}}
Content-Type: application/json

{
  "model": "llama-3.3-70b-versatile",
  "messages": [
    {
      "role": "system",
      "content": "You are a lead qualification assistant for a local marketing agency. Classify leads as GOOD_FIT, MAYBE, or NOT_FIT. Reply with only a JSON object: {\"classification\": \"GOOD_FIT\", \"reason\": \"one sentence\", \"priority\": \"high/medium/low\"}"
    },
    {
      "role": "user",
      "content": "Lead details:\nName: {{1.name}}\nBusiness: {{1.business_type}}\nService interest: {{1.service_interest}}\nBudget: {{1.monthly_budget}}\nMessage: {{1.message}}\n\nGOOD_FIT: local business, budget $300+/month, clear need\nNOT_FIT: wants free work, huge enterprise, completely wrong service, budget under $200\nMAYBE: everything else"
    }
  ],
  "max_tokens": 100,
  "temperature": 0.2
}

Module 3: Router — Branch by Classification

• Path A: classification = GOOD_FIT
• Path B: classification = MAYBE
• Path C: classification = NOT_FIT

Module 4A: AI Email Generation (Good Fit)

{
  "model": "llama-3.3-70b-versatile",
  "messages": [
    {
      "role": "system",
      "content": "You write personalized, conversational business emails. Sound like a real person, not a marketing robot. No buzzwords. Under 200 words. Sign off as Nataliia from DataLatte."
    },
    {
      "role": "user", 
      "content": "Write a reply to this inquiry:\n\nFrom: {{1.name}} ({{1.business_type}} in {{1.city}})\nInterested in: {{1.service_interest}}\nBudget: {{1.monthly_budget}}\nTheir message: {{1.message}}\n\nMake it:\n- Acknowledge their specific situation (mention their business type and city)\n- Reference 1 specific thing from their message\n- Suggest a quick 20-min call to see if we're the right fit\n- Include a Calendly link: https://calendly.com/datalatte/discovery\n- Warm but concise"
    }
  ],
  "max_tokens": 300,
  "temperature": 0.7
}

Module 4B: Template Email (Maybe / Not Fit)

Hi {{1.name}},

Thanks for reaching out to DataLatte! After reviewing your inquiry, I don't think we're the best fit for what you need right now — our services are focused on [specific gap in their request].

I'd recommend checking out [alternative resource].

Wishing you the best with {{1.business_type}}!

Nataliia
DataLatte.pro

Module 5: Send Email via Resend

Authorization: Bearer {{RESEND_API_KEY}}
Content-Type: application/json

{
  "from": "Nataliia at DataLatte <hello@datalatte.pro>",
  "to": ["{{1.email}}"],
  "subject": "Re: Your inquiry about {{1.service_interest}}",
  "html": "<p>{{4A.choices[0].message.content}}</p>"
}

Module 6: Log to Google Sheets

• Timestamp
• Name, email, business type, city
• Service interest, budget
• Classification + reason
• Email sent (yes/no)
• Priority level

Module 7: Slack / Email Notification (High Value Only)

Full Workflow JSON (Import into Make.com)

{
  "name": "Lead Follow-Up AI Agent",
  "flow": [
    {
      "id": 1,
      "module": "gateway:CustomWebHook",
      "version": 1,
      "metadata": {
        "designer": {"x": 0, "y": 0}
      }
    },
    {
      "id": 2,
      "module": "http:ActionSendData",
      "version": 3,
      "parameters": {
        "handleErrors": false,
        "useNewZLibDecompression": true
      },
      "mapper": {
        "url": "https://api.groq.com/openai/v1/chat/completions",
        "method": "post",
        "headers": [
          {"name": "Authorization", "value": "Bearer {{GROQ_API_KEY}}"},
          {"name": "Content-Type", "value": "application/json"}
        ],
        "bodyType": "raw",
        "contentType": "application/json",
        "body": "{\"model\":\"llama-3.3-70b-versatile\",\"messages\":[{\"role\":\"user\",\"content\":\"Classify this lead as GOOD_FIT, MAYBE, or NOT_FIT. Reply with JSON only: {classification, reason, priority}. Lead: {{1.name}}, {{1.business_type}}, Budget: {{1.monthly_budget}}\"}],\"max_tokens\":100}"
      }
    },
    {
      "id": 3,
      "module": "json:ParseJSON",
      "mapper": {
        "string": "{{2.body.choices[].message.content}}"
      }
    }
  ],
  "metadata": {
    "version": 1,
    "scenario": {"roundtrips": 1, "maxErrors": 3, "autoCommit": true}
  }
}

Real Results: What This Workflow Replaces

• 30–60 min/day replying to inbound inquiries
• 10–15 min qualifying each lead manually
• Inconsistent follow-up timing (some leads wait 4+ hours)

Frequently Asked Questions

Related Articles

• AI Appointment Reminder Agent: Python Script That Cuts No-Shows
• Make.com vs Zapier for Local Business: Which Platform Is Better?
• AI Agent for Lead Generation: How to Automate Your Follow-Ups
• How to Build an AI Agent for Local Business (No-Code Guide)
• AI Agents for Local Business: What They Are and How They Work