Make.com + Gmail: Automate Email Outreach and Response Workflows
Make.com connects directly to Gmail and lets you build visual workflows that watch for incoming emails, parse their content with AI, route them to the right team, and send replies — without writing a single line of code.
If your team manually triages support tickets, drafts acknowledgments, or forwards customer emails to different inboxes, you're doing work a machine can handle. Make.com's Gmail integration covers every operation you need: watching for new threads, reading message bodies, applying labels, creating drafts, and sending replies — all orchestrated through a drag-and-drop scenario builder.
This guide walks through four production-ready workflows:
- Watch and label — auto-categorize incoming emails by keyword or sender
- AI intent classification — use Claude or GPT-4o to classify customer inquiries
- Auto-respond — send templated acknowledgments within seconds of receipt
- Full triage pipeline — classify → route → acknowledge in a single scenario
What can Make.com actually do with Gmail?
Make.com treats Gmail as a first-class integration. The Gmail module set covers:
| Module | What it does |
|---|---|
| Watch Emails | Triggers a scenario when a new email matches a filter |
| Watch Email Labels | Fires when a label is added or removed |
| Get an Email | Fetches full message content by ID |
| Search Emails | Queries your inbox using Gmail search syntax |
| Send an Email | Sends a new message as you or as an alias |
| Create a Draft | Saves a draft without sending |
| Reply to an Email | Replies in-thread, preserving conversation history |
| Add a Label | Applies one or more labels to a message |
| Move an Email | Moves a message to a different mailbox folder |
| Mark as Read/Unread | Sets the read status |
| Delete an Email | Moves to Trash |
That is enough surface area to handle the entire lifecycle of an inbound email — from detection through resolution — without touching the Gmail interface.
Scenario 1: Watch emails and auto-label by keyword
This is the simplest starting point. Make watches your Gmail inbox and applies a label based on keywords in the subject line or body.
Use case: every email containing "invoice" gets tagged "Finance"; every email with "bug report" gets tagged "Support".
How to build it
- Open Make.com and create a new scenario.
- Add the Gmail > Watch Emails trigger. Set it to watch your inbox.
- Add a Router module to create parallel branches.
- In each branch, add a Filter — use the condition
Subject contains "invoice"for the Finance branch. - Connect each branch to a Gmail > Add a Label action with the appropriate label.
- Activate the scenario and set the polling interval (every 15 minutes on the free tier; near real-time on paid plans).
Tip: Gmail search operators work inside the Watch Emails filter field. You can use from:@company.com to target emails from a specific domain, or has:attachment to trigger only on emails with attachments.
Scenario 2: Parse email content with AI
Keyword matching breaks down with natural language. A support email might say "it stopped working" — no obvious keyword, but high-priority intent.
This scenario pipes every incoming email body through an AI model (Claude, GPT-4o, or Gemini) and extracts structured intent labels.
The scenario structure
Gmail: Watch Emails
→ OpenAI: Create a Completion (or Anthropic: Claude)
→ Router: branch by extracted intent
→ [Branch A] Gmail: Add Label "Urgent"
→ [Branch B] Gmail: Add Label "General Inquiry"
→ [Branch C] Gmail: Add Label "Billing"
AI classification prompt
In the OpenAI or Anthropic module, use this system prompt:
You are an email triage classifier. Given the email body below, return a single JSON object:
{
"intent": "urgent_support" | "billing" | "general_inquiry" | "sales" | "spam",
"priority": "high" | "medium" | "low",
"summary": "<one sentence summary>"
}
Return only valid JSON — no explanation, no markdown fencing.
Map the intent field from the JSON response to downstream router branches. Make's built-in JSON parsing modules extract the field values automatically.
Which AI to use? Claude Haiku 4.5 is the cost-effective choice for high-volume classification ($1/MTok input). GPT-4o mini is comparable. For complex email threads with attachments, Claude Sonnet 4.6 gives higher accuracy.
Scenario 3: Auto-respond to emails
Once you've classified an email, you can send a templated reply within seconds. This is the "acknowledgment" pattern common in support and sales workflows.
Building the auto-responder
- After the AI classification module, add a Gmail > Reply to an Email action.
- In the reply body, mix static template text with dynamic Make variables:
Subject: Re: {{1.subject}}
Hi {{1.from.name}},
Thank you for reaching out. We've received your message regarding "{{summary}}" and
classified it as {{intent}}. Our team will get back to you within 24 hours.
In the meantime, you can check our help center at https://help.yoursite.com.
Best regards,
Support Team
The {{1.subject}}, {{1.from.name}} tokens pull directly from the Gmail trigger module output. The {{summary}} and {{intent}} tokens come from the parsed AI JSON.
- Set the Reply Type to "Reply" (not "Reply All") to avoid sending auto-responses to entire distribution lists.
- Add a Gmail > Add a Label action after the reply to mark the thread as "Auto-Replied" — this prevents the Watch Emails trigger from picking it up again on the next poll.
Scenario 4: Full triage pipeline — classify, route, and acknowledge
This is the production-grade scenario the writing prompt describes: customer inquiry arrives → AI classifies intent → route to the correct team inbox → send acknowledgment.
Architecture
Gmail: Watch Emails (Inbox)
↓
Anthropic: Claude — classify intent + extract urgency
↓
Router (4 branches)
├─ [Billing] → Gmail: Forward to billing@yourcompany.com
│ → Gmail: Reply with billing template
├─ [Support] → Gmail: Forward to support@yourcompany.com
│ → Gmail: Reply with support template
├─ [Sales] → Gmail: Forward to sales@yourcompany.com
│ → Gmail: Reply with sales template
└─ [Spam] → Gmail: Move to Trash
↓ (all branches)
Gmail: Add Label "Triaged"
Step-by-step build
Step 1 — Connect Gmail
In Make, go to Connections > Add a Connection > Gmail. Authenticate with the Google account you want to automate. Make requests only the OAuth scopes it needs (gmail.modify, gmail.compose).
Step 2 — Configure the Watch Emails trigger
- Folder: INBOX
- Criteria: All email (or filter to specific senders if needed)
- Maximum number of emails: 5 per cycle (prevents burst processing on a busy inbox)
- Mark as read: Leave unchecked unless you want Make to mark emails read as it processes them
Step 3 — Add the Claude classification module
Use the HTTP > Make a Request module to call the Anthropic API directly, or connect through Make's native OpenAI integration and swap to the Anthropic base URL. The prompt:
Classify this customer email into one of: billing, support, sales, spam.
Also extract: urgency (high/medium/low) and a 1-sentence summary.
Email subject: {{1.subject}}
Email body: {{1.snippet}}
Return JSON only:
{"intent": "...", "urgency": "...", "summary": "..."}
Step 4 — Parse the JSON response
Add a Tools > Parse JSON module to convert the raw AI text response into structured Make variables.
Step 5 — Add the Router and branches
Each branch filter checks: intent equals "billing" (or "support", "sales", "spam"). Connect the appropriate forwarding and reply modules to each branch.
Step 6 — Test with sample emails
Use Make's Run once button to process a single email end-to-end. Inspect each module's output in the execution history to verify intent extraction and routing.
Sending proactive outreach campaigns
The Gmail integration isn't limited to reactive workflows. You can drive outbound sequences from Make scenarios.
Pattern: Google Sheets → Make → Gmail
- Maintain a list of prospects or customers in Google Sheets with columns:
email,name,company,personalization_note. - In Make, use Google Sheets > Watch Rows to trigger when a new row is added.
- Pipe the row data through an OpenAI module to generate a personalized email body.
- Send with Gmail > Send an Email, setting
Toas{{sheets.email}}.
This is a legitimate, consent-based outreach flow — not a spam blast. Make does not have built-in throttling controls for cold email sequences; if you need rate limiting, use a Sleep module between sends or trigger the scenario on a schedule with batch limits.
Important: Google's Gmail API has a sending quota of 500 emails per day for consumer accounts and 2,000 per day for Google Workspace accounts. For higher-volume outreach, use a dedicated email service provider (Mailgun, SendGrid, Postmark) connected to Make instead of the Gmail module directly.
Make.com pricing for email automation
Make's pricing is based on operations — each module execution in a scenario counts as one operation.
| Plan | Price | Operations/month | Scenario interval |
|---|---|---|---|
| Free | $0 | 1,000 | 15 minutes |
| Core | $10.59/mo | 10,000 | 1 minute |
| Pro | $18.82/mo | 10,000 | 1 minute |
| Teams | $34.12/mo | 10,000 | 1 minute |
| Enterprise | Custom | Custom | Real-time webhooks |
Source: Make pricing page
A typical email triage scenario (Watch → AI classify → Route → Reply) runs 4–6 operations per email. At the Core tier, you can process ~1,500–2,500 emails per month before needing to upgrade or add operation bundles.
Free tier limitation: The 15-minute polling interval means emails won't be processed immediately on the free plan. For support workflows where response time matters, the Core tier at $10.59/month is the practical minimum.
Make.com vs Zapier for Gmail automation
Both platforms support Gmail, but their architectural approaches differ:
| Feature | Make.com | Zapier |
|---|---|---|
| Visual scenario builder | ✅ Canvas-based | ❌ Linear steps only |
| Gmail modules | Full read/write/label/reply | Limited (send + trigger) |
| Built-in JSON parsing | ✅ Native | Requires formatter step |
| AI integration | OpenAI, Anthropic, Gemini native | OpenAI only (native) |
| Free tier ops/month | 1,000 | 100 tasks |
| Minimum paid plan | $10.59/mo | $19.99/mo |
| Best for | Complex routing, multi-branch | Simple linear workflows |
Best for simple Gmail → Slack/Sheets workflows: Zapier is faster to set up for single-step automations.
Best for AI-powered email triage and multi-branch routing: Make.com handles conditional logic and parallel branches more naturally.
FAQ
How to automate Gmail with Make?
Create a Make.com account, add a Gmail connection via OAuth, and build a scenario starting with the Gmail > Watch Emails trigger. Connect downstream modules to label, reply, forward, or route emails based on content. Activate the scenario to run on a schedule (free: every 15 min; paid: every 1 min).
Can Make.com read emails?
Yes. The Gmail > Get an Email module retrieves the full message including subject, body, sender, attachments, and metadata. The Gmail > Watch Emails trigger fires on new emails matching a filter and surfaces the same fields for use in downstream modules.
How do I auto-respond to emails without code?
Use the Gmail > Watch Emails trigger plus a Gmail > Reply to an Email or Gmail > Send an Email module. Set the reply body with a static template, or pipe the original email through an AI module to generate a personalized response. No code required.
Does Make.com support Gmail labels?
Yes. The Gmail > Add a Label, Gmail > Watch Email Labels, and Gmail > Remove a Label modules give full programmatic access to Gmail's label system. You can create labels in Gmail first and reference them by name in Make.
What is the Gmail API daily sending limit?
Standard Gmail accounts: 500 emails/day. Google Workspace accounts: 2,000 emails/day. These are Google-imposed limits, not Make limits. For higher volumes, use Make with a dedicated email service (Mailgun, SendGrid) instead of the Gmail module.
Can Make.com handle Gmail attachments?
Yes. The Watch Emails trigger exposes attachment metadata (filename, MIME type, size). You can download attachments using the Gmail > Get an Email module and pass them to other modules — for example, saving PDFs to Google Drive or running them through a document parser.
Next steps
The Gmail + AI triage scenario described here is one pattern. The broader use case for Make.com in email workflows extends to:
- CRM hydration: extract contact data from inbound emails and create or update HubSpot/Salesforce records
- Ticket creation: parse support emails and create Jira or Linear issues with AI-generated summaries
- Invoice processing: watch for emails with PDF attachments, extract line items with AI, and push to accounting software
- Newsletter monitoring: watch for emails from specific senders and route them to a Notion database or Slack channel
Start building with Make.com for free — the free tier includes 1,000 operations/month and full access to the Gmail integration.
For data extraction workflows that feed into these email pipelines — scraping product prices, monitoring competitor content, or aggregating web data — Apify connects directly to Make.com via webhook and API, letting you build end-to-end pipelines from web data to email delivery without additional infrastructure.