Every freelancer has lost a lead simply by replying too late. Someone fills out a contact form at 11 PM, gets excited about your service, and by the time you reply the next afternoon, they've already hired someone else who answered faster.
This lesson fixes that problem permanently. You're going to build a complete, three-part automation that captures a lead, stores it like a mini CRM, and sends back a personalized AI-written follow-up — all within seconds of someone submitting a form. No manual checking, no delayed replies, no lost business.
This is also the first lesson in Phase 4, which means it's the first time everything you've learned so far — webhooks, Google Sheets, filters, and the OpenAI module — comes together into one real, sellable system. This exact automation is something freelancers charge $150–$400 to build for small businesses.
What You'll Learn
- How to turn a simple form submission into a structured lead record
- How to use Google Sheets as a lightweight CRM (Customer Relationship Management) system
- How to generate a personalized, context-aware follow-up message using AI instead of a generic template
- How to send that follow-up automatically, mapped to the actual lead's email address
- How this single automation becomes a real freelance service offer
Core Concept: The Digital Sales Assistant
Think about what a good sales assistant actually does. The moment a new inquiry comes in, they don't wait. They read it, understand what the person is interested in, and send back a thoughtful, relevant response — fast enough that the lead still remembers why they reached out in the first place.
That's exactly what you're building here, except it runs automatically, at any hour, for as many leads as show up.
The system has three distinct jobs, each handled by a different part of your scenario:
1. Capture. The moment someone submits their details, that information needs to land somewhere permanent. A spreadsheet row, not an inbox you'll forget to check.
2. Think. This is where the automation stops feeling robotic. Instead of sending the same copy-pasted reply to everyone, the AI reads what this specific lead is interested in and writes something that actually responds to it.
3. Respond. The generated message goes out immediately, addressed to the real person who submitted the form — not to you, and not to a placeholder email left over from testing.
You've already built versions of each of these pieces separately. Lesson 8 taught you how a webhook captures incoming data. Lesson 4 taught you how to write that data into Google Sheets. Lesson 11 taught you the "Read → Think → Respond" pattern for AI-generated replies. This lesson is where those pieces finally work together as one connected system, which is the real skill clients are paying for — not knowing one trick, but being able to chain several together into something useful.
Tool Walkthrough: Step-by-Step Build
Step 1: Set up your lead source
You need a way for someone to submit their details. The simplest option, and the one we'll use here, is sending the lead's information straight to a Make webhook — the same approach from Lesson 8. If you'd rather use an actual form tool (like a Blogger contact form or Google Form), the webhook step still applies; you're just capturing the same three fields: Name, Email, and Interest (what service or product they're asking about).
For testing purposes, you'll simulate a form submission by sending this data directly through a browser URL, exactly as you did in Lesson 8.
Step 2: Create the trigger — Custom Webhook
In a new scenario, search for the Webhooks app and choose Custom webhook. Click Add, give it a name like "New Lead Capture," and save it. Make will generate a unique webhook URL — copy this, you'll need it to send test data.
Step 3: Send test lead data
Open your webhook URL in a browser and add a query string with sample lead data, like this:
name=Sarah&email=sarah@example.com&interest=Website Redesign
With your scenario open and listening, visit that URL. Make should detect three values and confirm "data structure recognized" — this is your "Redetermine data structure" moment from Lesson 8, so click that if the fields don't appear automatically.
Step 4: Add Google Sheets — "Add a Row" (the CRM)
Click the + after your webhook module and search for Google Sheets. This time, choose Add a Row — not Search Rows. This module's job is simple: every new lead becomes a brand-new line in your spreadsheet.
Connect your Google account, select (or create) a spreadsheet named something like "Leads CRM," and set up four columns: Name, Email, Interest, and Status. Map the Name, Email, and Interest fields from your webhook data into the matching columns, and manually type "New" into the Status field — this gives you a simple way to track, in future lessons, which leads have already been followed up with.
Step 5: Add OpenAI — "Generate a Response" (the Think step)
Add the OpenAI (ChatGPT, Whisper, DALL-E) module and choose Generate a response, the same module from Lessons 10–13. Select your existing connection and GPT-4o model.
In the Prompt field, write something that pulls in the lead's actual details rather than a flat instruction. For example:
"Write a short, friendly, professional follow-up email to a potential client named [Name], who is interested in [Interest]. Thank them for reaching out, briefly explain why we're a good fit for that specific need, and invite them to reply with any questions. Keep it under 120 words and sign it as 'Kartik.'"
Replace [Name] and [Interest] with the actual mapping chips from your webhook module. This single change — pulling in real lead data instead of writing a generic prompt — is what separates a mass-produced auto-reply from something that reads like it was written specifically for that person.
Step 6: Add Gmail — "Send an Email" (the Respond step)
Add the Gmail module and choose Send an Email. In the To field, map the lead's email address from your webhook module — never type a fixed address here, since every lead needs their own reply. In the Subject field, write something like "Thanks for reaching out, [Name]!" using the mapped name. In the Content field, map the OpenAI module's output (its Result or Raw Result field).
Step 7: Test the full chain
Click Run once, then send a fresh test submission through your webhook URL with a new name, email, and interest. Watch all three modules light up with green checkmarks. Check your Google Sheet for the new row, and check the test email inbox for the AI-written follow-up.
Step 8: Activate
Once the test email arrives looking right, toggle the scenario to Active. Your lead capture system is now live and listening for real submissions, around the clock.
Real Example: The Freelancer Who Never Misses a Lead
Imagine you're running a small web design business. Someone finds your site at 1 AM, fills out your contact form with "I need a redesign for my bakery's website," and closes the tab expecting to hear back in a day or two.
Instead, within seconds, they get a reply: a warm, specific message that mentions their bakery and redesign needs by name, thanks them for reaching out, and invites them to share more details. They didn't know it was automated. They just know someone responded fast — and that single impression is often the difference between winning the client and losing them to a competitor who replied first.
This is the exact system you just built, and it's one you can resell to real businesses as a done-for-you lead response service.
Common Mistakes
Writing a generic AI prompt. If your prompt doesn't reference the lead's actual name and interest, you've just built a slower, more expensive version of a form auto-responder. The personalization is the entire value of this system.
Leaving the "To" field as a fixed test email. This is the easiest mistake to make and the most damaging one to miss, since it means every real lead's follow-up would go to the wrong inbox. Always map it dynamically.
Skipping the Status column. It feels optional right now, but it becomes essential once you're tracking multiple leads and need to know which ones still need a human follow-up call.
Try This Yourself
Add a second AI step after the follow-up email: use OpenAI again to generate a one-line internal note ("Lead is interested in e-commerce, mentioned a tight timeline") and have Google Sheets write it into a fifth column. This gives you an instant summary of every lead without re-reading their original message.
Key Takeaways
- A complete lead-response system needs exactly three jobs: capture, think, and respond
- Personalizing the AI prompt with real lead data is what makes the output feel human instead of automated
- Dynamic field mapping (especially on the "To" field) is non-negotiable in any system handling real contacts
- This project chains together webhook, spreadsheet, and AI skills you've already learned — proving they work as a connected system, not just isolated tricks
FAQ
Can I use a real form tool instead of testing through a browser URL?
Yes — any form tool that can send data to a webhook (or that Make has a direct integration for) will work the same way. The browser URL method is just the fastest way to test without building a form first.
What if a lead doesn't include their email address?
Add a filter after the webhook module that only lets the scenario continue if the Email field isn't empty, preventing errors on incomplete submissions.
Is this legal to use for real client outreach?
Yes, since this responds to someone who directly contacted you — it isn't unsolicited marketing. Just make sure your message is honest about who it's from.
Next Lesson Preview
In Lesson 15, you'll build Project 2: AI Review & Feedback Reply System, where AI reads customer feedback, judges the mood, and replies automatically to happy customers while unhappy ones wait for a human.
[← Back to Course Hub] | [Lesson 13: No-Code AI Chatbot] | [Lesson 15: AI Reply System→]