How to Build an AI Lead Qualification System in Under an Hour

You run a solo service business or a small team. Leads trickle in from your website form — vague, half-finished messages. Someone types “I need help with my website” and you have no idea whether they want a logo refresh or a full platform rebuild. Their budget could be £200 or £20,000. You reply. They go quiet. Three days later, you’ve completely forgotten they exist.

Sound familiar? You’re not bad at sales. You just don’t have a system. Every qualified lead you forget to follow up is money you earned and didn’t collect.

This tutorial builds that system for you — entirely on free tools, without writing a single line of custom code.

“Every time someone reaches out, you’ll know immediately: who to call first — and what to say when you do.”

What We’re Building

By the end of this tutorial you’ll have an eight-module Make.com scenario that does the following every time a prospect fills in your intake form:

• Captures the enquiry in your Google Sheets CRM — instantly, no copy-pasting
• Sends the form data to an AI (OpenAI or Claude) for analysis
• Returns a plain-English summary, a score out of 100, and a suggested next action
• Generates a personalised follow-up email, ready to send
• Fires a Slack notification to your phone with everything you need to act

Tools You’ll Need

All five tools have free tiers that comfortably run this workflow. If you prefer Notion, Airtable, HubSpot, or Claude over OpenAI, every step has a direct equivalent.

Build Your Intake Form & Tracker Sheet

The Five Questions That Actually Qualify Leads

Add Your AI Output Columns

Once you have the linked sheet, add these six empty columns at the end — this is where the AI output will be written in later steps:

AI_Summary  |  Lead_Score  |  Next_Step  |  Followup_Subject  |  Followup_Email  |  Status

// Status values: New · Contacted · Qualified · Proposal Sent · Closed · Not a Fit

Set Up the Trigger in Make.com

Make.com (formerly Integromat) is a no-code automation platform where you build visual “scenarios” — sequences of steps that run automatically whenever a trigger fires. Sign up free at make.com — it takes about two minutes.

Write the Lead to Your Sheet (Add a Row)

Click + after the trigger, search for “Google Sheets,” and choose Add a Row. Select your Lead Tracker spreadsheet and map each form field to the matching column:

• Name column → map the ‘Name’ field from the trigger
• Email column → map the ‘Email’ field from the trigger
• Service Needed → map ‘What do you need help with?’
• Timeline → map ‘What is your timeline?’
• Budget → map ‘What budget range are you comfortable with?’
• Success Goal → map ‘What does success look like for you?’
• Next Step Preference → map ‘What would be the best next step for you?’

Leave the six AI columns empty for now. This module locks the raw lead into your CRM the second it arrives, before any AI processing begins.

Find the Row (Search Rows)

Add Google Sheets → Search Rows. Set the Filter Column to Email and map the email from the trigger as the Filter Value. This bookmarks the exact row number so Step 7 knows precisely where to write the AI output back — no guesswork.

Generate the AI Summary & Score

Add an OpenAI → Simple Text Prompt module (or Anthropic → Create a Message if you prefer Claude). You’ll need an API key from platform.openai.com or console.anthropic.com — both platforms offer free trial credits to get started.

In the Prompt field, paste the following exactly. The {{double-brace}} syntax pulls live data from the Google Forms trigger.

// Instructs the AI to return structured JSON — no markdown, no code fences

You are an assistant that qualifies inbound leads for a service business.
Using the form data below, return ONLY valid raw JSON (no markdown, no backticks).

FORM DATA
Name:            {{Name}}
Email:           {{Email}}
Service Needed:  {{What do you need help with?}}
Timeline:        {{What is your timeline?}}
Budget:          {{What budget range are you comfortable with?}}
Success Goal:    {{What does success look like for you?}}
Next Step Pref:  {{What would be the best next step for you?}}

Return JSON with exactly these keys:
  name, email, summary  (2–3 sentences describing the lead in plain English)
  score                  (integer 0–100)
  next_step              (one short sentence: what to do next)
  followup_subject       (short, specific email subject line)
  followup_email         (5–8 sentences, friendly, personalised, clear CTA)

Scoring rules:
  Score 75–100 if: budget >= £5k AND timeline is within 2 weeks or 1 month
  Score 40–74  if: budget or timeline is unclear or mid-range
  Score 0–39   if: message is vague, no budget given, or "Just exploring"
  If score < 50: set next_step to "Send two clarifying questions before booking a call."

How the Lead Scoring Works

Parse the AI Response (JSON Parser)

The AI returns its answer as a block of JSON text. The JSON Parser module breaks that text into individual usable values — summary, score, next_step, etc. — that later steps can reference separately.

Write the AI Results Back to the Sheet

Add Google Sheets → Update a Row. For the Row Number field, map the row number returned by the Search Rows step from Step 4. Now map the parsed AI fields to the correct columns:

• AI_Summary column ← summary from JSON parser
• Lead_Score column ← score from JSON parser
• Next_Step column ← next_step from JSON parser
• Followup_Subject column ← followup_subject from JSON parser
• Followup_Email column ← followup_email from JSON parser

Send a Slack Notification

Add Slack → Create a Message. Connect your Slack account, choose a channel (a private #leads channel works well), and paste the template below. Map each variable to the matching field from the JSON parser.

🆕 New lead: {{Name}} ({{Email}})

Score:      {{score}} / 100
Next step:  {{next_step}}
Summary:    {{summary}}

— Follow-up email is ready in your Lead Tracker sheet —

Running Your End-to-End Test

Click Run Once in your Make scenario, then submit a realistic test response through your Google Form. Use a real-looking name, email, and answers that resemble an actual enquiry.

Switch back to Make. All eight modules should light up green in sequence. If any module shows a red error icon, click it — Make shows you exactly what went wrong and usually suggests the fix.

Then check two places: your Google Sheet (all five AI columns should be filled) and your Slack channel (the notification should have arrived).

“Form submission to qualified lead in your pocket — fully automated. That’s the whole system.”

Activate Your Scenario

Right now the scenario only runs when you manually click Run Once. To make it fire automatically, find the toggle switch in the bottom-left corner of the Make scenario editor and flip it from OFF to ON. Make will now check for new responses every 15 minutes and run the full automation without you touching anything.

What If I Don’t Want to Use AI for Scoring?

Totally valid. Make has a built-in Tools module with conditional logic — write rules like “if budget is £5k or more AND timeline is within a month, set score to 80.” It takes a bit more setup but is fully transparent and completely free with zero API costs. Drop a comment below if you’d like a separate video on rule-based scoring.