AI agents for email marketing sound great in theory. But what does it actually look like in practice? What do you type, and what happens?

This guide contains 10 real workflows you can use with Sequenzy's AI agent today. Each workflow includes the exact prompts, what the agent does behind the scenes, and the time it saves compared to doing it manually in a dashboard.

These workflows assume you have Sequenzy MCP set up in Claude Desktop, Cursor, or Claude Code.

1. New User Onboarding Sequence

The prompt:

"Generate a 5-email onboarding sequence for new users who just signed up for a free trial. Focus on getting them to complete their first integration (Stripe or Shopify). Space emails 2 days apart. Make the tone helpful but not pushy."

What the agent does:

Calls generate_sequence with your goal and parameters
Creates 5 emails with subject lines, body copy, and CTAs
Sets up 2-day delays between each step
Configures the trigger for new trial signups
Shows you the full sequence for review

What each email covers:

Email 1 (Day 0): Welcome + quick start guide
Email 2 (Day 2): Integration setup walkthrough (Stripe/Shopify)
Email 3 (Day 4): First campaign creation guide
Email 4 (Day 6): Advanced features overview
Email 5 (Day 8): Trial ending reminder + upgrade CTA

Time saved: ~45 minutes of writing, configuring, and setting up in a dashboard

Follow-up prompt:

"The third email feels too long. Shorten it to focus only on creating a first campaign. Also, add a P.S. with a link to book a demo call."

The agent updates just that email while keeping everything else intact.

2. Weekly Performance Report

The prompt:

"Give me a performance summary for the past 7 days. Include: total emails sent, average open rate, average click rate, best performing campaign, and any sequences that have low engagement."

What the agent does:

Calls get_stats for overall metrics
Calls list_campaigns to find recent campaigns
Calls get_campaign_stats for each campaign
Calls list_sequences and get_sequence_stats for active sequences
Compiles a summary with highlights and flags

Sample output:

The agent gives you something like:

Emails sent: 12,450
Average open rate: 38.2% (up from 35.1% last week)
Average click rate: 4.8%
Top campaign: "API v2 Launch" - 52% open rate, 8.3% click rate
Flag: "Re-engagement Sequence" has 12% open rate on email 3 (down from 28% on email 2). Consider rewriting the subject line.

Time saved: ~15 minutes of dashboard navigation and mental math

Follow-up prompt:

"The re-engagement sequence email 3 has bad stats. Generate 5 alternative subject lines and show me the current one for comparison."

3. Product Launch Campaign

The prompt:

"I'm launching a new feature called 'Smart Replies' that auto-detects and categorizes subscriber email replies. Create a launch campaign: draft an announcement email for all active subscribers, create a separate version for developers that includes the API changes, and schedule both for next Tuesday at 10am."

What the agent does:

Creates segment for all active subscribers
Creates segment for subscribers tagged 'developer'
Drafts announcement email (general audience version)
Drafts developer-focused version with API details
Creates both campaigns with the Tuesday schedule
Sends test emails to your address

Time saved: ~1 hour of writing two email versions, creating segments, configuring campaigns, and scheduling

Follow-up prompt:

"Send test emails for both campaigns to my email so I can preview them on mobile."

4. Churn Prevention Flow

The prompt:

"Set up a churn prevention workflow. Find all subscribers who haven't opened an email in the last 45 days, create a segment called 'at-risk-inactive', and generate a 3-email win-back sequence. First email should be 'we miss you', second should highlight new features they missed, third should offer a 20% discount."

What the agent does:

Calls create_segment with the inactivity filter
Calls get_segment_count to see how many subscribers match
Calls generate_sequence with the 3-email brief
Connects the sequence to the at-risk segment
Reports the subscriber count and shows the sequence for review

Time saved: ~30 minutes

Follow-up prompt:

"How many subscribers are in the at-risk segment? And what's their average MRR if that data is available?"

5. A/B Test Subject Lines

The prompt:

"I want to A/B test subject lines for my monthly newsletter. Generate 5 options. The newsletter is about our Q1 product updates including new analytics dashboard, Shopify improvements, and faster email delivery."

What the agent does:

Calls generate_subject_lines with the context
Returns 5 variations with reasoning for each approach

Sample output:

"What we shipped in Q1 (and what's coming next)" - Curiosity + forward-looking
"New analytics dashboard, faster delivery, and more" - Direct feature list
"Your emails now deliver 2x faster" - Lead with the most impactful improvement
"Q1 update: 3 features your competitors will hate" - Competitive angle
"We rebuilt the analytics dashboard from scratch" - Story-driven

Time saved: ~10 minutes of brainstorming

Follow-up prompt:

"I like options 1 and 3. Set up an A/B test with those two subject lines for the newsletter campaign."

6. Subscriber Cleanup and Segmentation

The prompt:

"Help me clean up my subscriber list. Find subscribers with these characteristics and create segments for each: (1) bounced emails in the last 30 days, (2) unsubscribed but still in the database, (3) no opens in 90 days, (4) high-value users with MRR over $200."

What the agent does:

Creates four segments with the specified filters
Reports subscriber counts for each
Suggests actions for each segment

Sample output:

Bounced (30d): 47 subscribers - "Recommend removing to protect deliverability"
Unsubscribed: 312 subscribers - "Already excluded from sends, safe to keep for records"
No opens (90d): 891 subscribers - "Consider a final re-engagement attempt before suppressing"
High-value ($200+ MRR): 156 subscribers - "These are your VIPs - consider a dedicated nurture sequence"

Time saved: ~20 minutes

Follow-up prompt:

"Remove the bounced subscribers and generate a short re-engagement email for the 90-day inactive group. If they don't open within 14 days, suppress them."

7. Stripe Event Response Emails

The prompt:

"Set up transactional emails for key Stripe events. I need: (1) a welcome email when a subscription is created, (2) a receipt email after successful payment, (3) a dunning email when payment fails, (4) a win-back email when subscription is cancelled. Make them concise and professional."

What the agent does:

Generates four email templates
Creates each as a template in Sequenzy
Provides configuration guidance for connecting to Stripe webhooks

Time saved: ~1 hour of writing and configuring four separate email flows

Follow-up prompt:

"The dunning email should be more urgent. Add a line about losing access to their data after 7 days if payment isn't updated."

8. Competitor Comparison Campaign

The prompt:

"We're seeing users churn to [competitor]. Draft a campaign targeting users who visited our pricing page in the last 7 days (tag them as 'pricing-page-visitors'). The email should address why they might be considering alternatives and highlight our advantages: better Stripe integration, MCP support, and simpler pricing."

What the agent does:

Creates or finds the 'pricing-page-visitors' segment
Drafts a retention-focused email that addresses objections
Creates the campaign and targets the segment
Shows preview for review

Time saved: ~25 minutes

9. Event-Driven Nurture Setup

The prompt:

"Create an automation for users who trigger the 'first_campaign_sent' event. Wait 1 day, then send a congratulations email with tips for improving their next campaign. Wait 3 more days, send an email about advanced features (A/B testing, send time optimization). Wait 5 more days, send a case study of a similar business achieving good results."

What the agent does:

Creates a 3-email sequence
Configures the trigger for 'first_campaign_sent' event
Sets up delays (1 day, 3 days, 5 days)
Generates contextual content for each email
Activates the sequence

Time saved: ~40 minutes

Follow-up prompt:

"Add a condition: only send this sequence to users on the free plan. Pro users should get a different sequence focused on API integration."

10. Monthly Newsletter Automation

The prompt:

"Help me set up a recurring monthly newsletter workflow. Here's what I need: (1) a template I can reuse each month with sections for product updates, a featured blog post, and a community highlight. (2) a segment of all active subscribers who haven't unsubscribed. (3) a draft for this month's newsletter - product update is the new AI sequence generator, featured post is our guide on onboarding emails, community highlight is a customer who grew their list from 500 to 10,000."

What the agent does:

Creates a reusable newsletter template with the three sections
Creates or verifies the active subscriber segment
Drafts this month's newsletter filling in all three sections
Creates the campaign and attaches the segment
Sends a test email

Time saved: ~35 minutes

Follow-up prompt:

"Looks great, but move the community highlight to the top - social proof first. Also shorten the product update section."

Tips for Better AI Agent Workflows

Be Specific About Tone

Bad: "Write an email"

Good: "Write a friendly, concise email. No corporate jargon. Use short paragraphs. Sound like a real person, not a marketing team."

Provide Context

Bad: "Create a welcome sequence"

Good: "Create a welcome sequence for B2B SaaS founders who signed up from our Product Hunt launch. They're technical, time-constrained, and care about fast setup."

Iterate, Don't Restart

The agent remembers your conversation. Instead of starting over:

"Make the second email shorter" "Add a code example to email 3" "Change all CTAs to point to the setup guide instead of the homepage"

Chain Operations

The most powerful use of AI agents is chaining multiple operations:

"Find my top 100 subscribers by engagement score, create a VIP segment, generate a special offer email, and send a test to me"

One prompt, four operations. This is where AI agents truly outperform dashboards.

Use Follow-Up for Analysis

After any campaign send:

"How did the campaign I sent yesterday perform? Compare it to my average metrics and suggest what to do differently next time."

Getting Started

These workflows work with any MCP-compatible AI tool. The fastest path:

Sign up for Sequenzy (free, no credit card)
Add the MCP server to Claude Desktop or Cursor
Start with Workflow #2 (performance report) to see your current data
Try Workflow #1 (onboarding sequence) to generate your first AI-created content
Build from there

The goal isn't to replace everything you do in the dashboard. It's to save time on the repetitive parts so you can focus on strategy and creative direction.

Start using the AI agent and see how much time you save in your first week.