Introduction
Every lead that lands on your website is a potential customer — but not every lead is ready to buy, a good fit, or worth your time right now. Manually sifting through contact form submissions is slow, inconsistent, and pulls you away from actually running your business.
This tutorial walks you through building an AI-powered lead qualifier that automatically scores and categorizes every new inquiry you receive. When someone fills out your contact form, the system uses AI to read their message, decide whether they’re a hot, warm, or cold lead, and send you a summary with a recommended next action — all without you lifting a finger.
What you’ll end up with: A working automation that reads new form submissions, sends them to an AI for analysis, and delivers a pre-qualified summary to your inbox or Slack channel.
Time to complete: About 45–60 minutes, spread across four phases.
What You’ll Need
Before you start, set up accounts on these three free-or-low-cost tools:
- Tally (tally.so) — a free form builder. You’ll use this to collect leads. If you already use Typeform or Google Forms, those work too.
- Make (make.com) — the automation platform that connects everything. The free tier covers this project comfortably.
- OpenAI (platform.openai.com) — the AI that reads and qualifies each lead. You’ll need to add $5–10 in API credits (more than enough for thousands of leads).
- A Gmail or Slack account to receive your qualified lead summaries.
Step-by-Step Setup
Phase 1: Build Your Lead Form (10 minutes)
Step 1: Log into Tally and click New form from your dashboard. Give it a name like “Website Contact Form.”
Step 2: Add these four fields to your form using the + button after each block:
- Name (Short text)
- Email (Email field)
- Company size (Dropdown — add options: Just me, 2–10 employees, 11–50, 50+)
- What are you looking for help with? (Long text — this is the field the AI will analyze)
Step 3: Click Share in the top right, then copy the form link. Paste it somewhere you can find it — you’ll embed this on your website later.
Step 4: Click Integrations on the left sidebar, then select Webhooks. Toggle it on. You’ll come back to fill in the webhook URL in Phase 2.
Phase 2: Set Up Your Make Automation (20 minutes)
Step 5: Log into Make and click Create a new scenario from your dashboard.
Step 6: Click the large + circle in the center of the canvas. Search for Webhooks and select it. Choose Custom webhook as your trigger. Make will generate a unique webhook URL — copy it.
Step 7: Go back to Tally, paste that URL into the Webhooks field you opened in Step 4, and click Save. Tally will now ping Make every time someone submits your form.
Step 8: Back in Make, click OK on the webhook module. Then click the small + to the right of that module to add the next step. Search for OpenAI and select it. Choose Create a Completion (GPT).
Step 9: In the OpenAI module settings, you’ll need to connect your account. Click Add next to the Connection field, then paste your OpenAI API key (find this at platform.openai.com under API keys → Create new secret key).
Step 10: Set the Model to gpt-4o-mini (it’s fast and inexpensive). In the Prompt field, paste this exactly:
You are a lead qualification assistant. A potential customer submitted this contact form:
Name: {{1.name}}
Company size: {{1.company_size}}
Message: {{1.what_are_you_looking_for_help_with}}
Classify this lead as HOT, WARM, or COLD based on:
- HOT: clear need, decision-maker signals, urgency
- WARM: genuine interest but vague or early stage
- COLD: off-topic, spam, or not a good fit
Return a 3-sentence summary and your classification. End with: "Recommended action: [one specific next step]"
The {{1.name}} tags pull data directly from your form submission. Make will auto-suggest them as you type.
Step 11: Click OK to save the OpenAI module.
Phase 3: Deliver the Summary to Your Inbox (10 minutes)
Step 12: Click + to add one more module. Search for Gmail (or Slack if you prefer). Select Send an Email.
Step 13: Connect your Gmail account when prompted and allow the requested permissions.
Step 14: Fill in the email fields:
- To: your business email address
- Subject:
New Lead: {{1.name}} — [AI will fill classification here] - Content: Click into the body field and select the choices[].message.content output from your OpenAI module. This inserts the AI’s full summary.
Step 15: Click OK, then click Save (floppy disk icon, top left). Finally, toggle the scenario from Off to On using the switch in the bottom left.
Testing It Works
Step 16: Open your Tally form link in a new browser tab. Fill it out as if you were a customer — use a realistic message like “I run a 12-person accounting firm and I’m looking for help automating client onboarding. We need something up and running before tax season.”
Step 17: Hit submit, then switch back to Make. In your scenario view, click Run history on the left panel. You should see a new entry appear within 30 seconds. Click it to confirm all three modules show a green checkmark.
Step 18: Check your inbox. You should have an email with the AI’s lead summary and classification. If the lead above was processed correctly, it should come back as HOT with a clear recommended action.
If all three steps are green and the email arrived, you’re live.
Troubleshooting
The webhook isn’t triggering
Go back to Tally’s webhook settings and confirm the URL matches exactly what Make generated. Even a missing character breaks it. In Make, click your webhook module and hit Re-generate to get a fresh URL, then update Tally.
The OpenAI module is returning an error
The most common cause is an invalid or missing API key. In Make, open the OpenAI module, click the connection name, and select Verify. If it fails, delete the connection and re-add it with a freshly copied key from OpenAI. Also confirm your OpenAI account has available credit.
The email is arriving but the AI summary is blank
This usually means the variable mapping is wrong. Open your OpenAI module, scroll to the output section, and confirm that choices[].message.content is what’s being passed to Gmail. If the field is empty, re-run the scenario manually using Make’s Run once button while a test submission is active.
Next Steps
Once this is running smoothly, add a Google Sheets module between OpenAI and Gmail. Route every lead summary into a spreadsheet — one row per submission — with columns for name, email, classification, and the AI summary. Within a week you’ll have a clean pipeline view of every inquiry that came in, automatically sorted by lead quality, without ever opening a CRM.