How to Automate HubSpot Lead Follow-up Emails with ChatGPT and Make.com

Why Automated Follow-Up Emails Matter

What You’ll Build

1. Triggers when a new contact is created (or a deal stage changes) in HubSpot

2. Calls ChatGPT with lead data to generate a personalized follow-up email

3. Sends the email via HubSpot’s email API

Prerequisites

• A HubSpot account (Starter or higher for API access)

• An OpenAI API key for ChatGPT access

• A Make.com account (free tier works for low volume)

Step 1: Set Up the HubSpot Trigger in Make

1. Create a new scenario in Make.com

2. Add a HubSpot — Watch Contacts or Watch Deals module as the trigger

3. Connect your HubSpot account via OAuth

4. Set the trigger to fire when a new contact is created, or when a deal moves into your target stage (e.g., “New Lead”)

5. Map the fields you want to pass to ChatGPT: first name, company name, industry, lead source, and any notes

Step 2: Build the ChatGPT Prompt

1. Add an OpenAI — Create a Completion module

2. Connect your OpenAI API key

3. Write a prompt template that uses the lead data from Step 1. Example:

You are a helpful sales assistant. Write a short, personalized follow-up email (3–4 sentences max) for a new inbound lead.

Lead details:
- Name: first_name last_name
- Company: company
- Industry: industry
- How they found us: lead_source

The email should be warm, specific to their industry, and end with a soft CTA to book a 20-minute call. Do not use generic filler phrases.

1. Set the model to gpt-4o for best results

2. Map the lead data fields from the HubSpot trigger into the prompt

Step 3: Send the Email via HubSpot

1. Add a HubSpot — Create or Update a Contact module (or use the Send Email action if your plan supports transactional email)

2. Map the ChatGPT output as the email body

3. Set the “From” address to your sales rep’s email

4. Set the subject line — you can also have ChatGPT generate the subject in a second API call, or use a template like “Quick question for you, first_name”

Tips for Better Results

Write Better Prompts

• Tone guidance (friendly but professional, not pushy)

• Length constraints (3–4 sentences is a good ceiling for follow-ups)

• What NOT to do (avoid phrases like “I hope this email finds you well”)

• Company context (add a brief description of your business in the system prompt)

Use Conditional Logic for Different Lead Types

Track Email Performance Back in HubSpot

Common Issues and Fixes

• ChatGPT returns too long a response: Add a token limit (max_tokens: 200) and tighten the prompt constraint

• HubSpot rate limits: Add a delay module in Make if you’re processing bulk imports

• Emails landing in spam: Make sure your HubSpot sending domain is properly authenticated (DKIM + SPF)

Ready to Set This Up?