Supported by

Is this what most of your day looks like: You spend 30 minutes pulling performance data from your ESP to figure out what worked last month. Then you copy your campaign brief from Monday.com and manually recreate it. You ask ChatGPT for subject line ideas, but it gives you generic advice because it has no idea what actually works for your audience. You know you should be analyzing competitor emails, but you’d need more hours in the day to do it properly. By the time you’ve actually set up the campaign, you’ve burned two hours before even thinking about strategy.

You’re not doing email marketing. You’re being the integration layer between tools that refuse to talk to each other.

That changes with Model Context Protocol (MCP).

MCP lets AI connect to your tools and do things. Not just give you suggestions, but actually pull your data, analyze your campaigns, create emails in your ESP, update your CRM, and log everything to your project management system. All in one conversation.

This isn’t coming soon. It’s here. Klaviyo has it. MailerLite has it. HubSpot has it. And email marketers are already using it to cut hours of busywork out of their week.

Here’s what you need to know.

What is MCP?

Model Context Protocol is a bridge that lets AI assistants talk directly to your tools.

Before MCP, you’d copy information out of one tool, explain it to your AI, get a response, then manually implement it in another tool. Every question required context. Every action required manual work.

With MCP, your AI reads your tools directly. It sees your campaign briefs in Monday.com or Asana. It knows your past campaign performance in Klaviyo or HubSpot. It can pull contacts from your CRM, analyze competitor emails in Gmail, and log findings to Notion. All without you copying and pasting anything.

But here’s what makes it different: MCP doesn’t just pull data. It takes actions. It can create campaigns, update records, set up automations, and schedule sends. You ask it to do something, and it does it.

MCP launched in November 2024 from Anthropic as an open standard. That means it’s not locked to one AI platform. ChatGPT can use it. Claude can use it. Cursor and other AI code editors can use it. Right now, most email marketers are using it through Claude Desktop since that was the first to support it fully, but the protocol itself works with any AI that connects to it.

Why This Matters for Email Marketers

Email production involves so many tools! You brief campaigns in one place, design in another, write copy somewhere else, and manage everything in your ESP. Every handoff is a chance for delays or mistakes.

MCP closes those gaps.

Faster campaign setup: Your AI reads your campaign brief, analyzes what’s worked before, and can even create a campaign for you. I’m not saying that you would necessarily use that campaign, but it might give you a good starting point!

Better context: When you ask for recommendations, your AI already knows your brand guidelines, past campaign performance, audience segments, and what your competitors are doing. No explaining the same context every time.

Fewer errors: When systems talk directly to each other, there’s less room for mistakes. You could use MCP to help you scan for mistakes or identify issues before you send.

More time for strategy: Less time managing files and formats means more time thinking about what actually moves metrics.

The goal isn’t to replace people. It’s to remove busywork so you can focus on the strategic decisions that matter.

Real Workflows Email Marketers Are Using Today

Campaign Creation & Performance Analysis in Klaviyo

Klaviyo was one of the first ESPs to build MCP support, and they’ve made it available to all customers. A few use cases they support include:

Creating campaigns with conversational prompts: “Create a fall sale campaign targeting engaged subscribers. Use my best-performing subject line style. Include preview text and set it to draft.”

The AI pulls your past campaigns to understand brand voice. It configures the right segments automatically. It writes copy that matches your tone. Everything goes into draft mode ready for your review.

Time from prompt to reviewable campaign: under 2 minutes. Time saved per campaign: 30-45 minutes.

Performance analysis that’s actually useful: “Which of my new product campaigns is driving the most revenue?” or “Compare flow performance between my two accounts.”

Instead of pulling reports manually, the AI analyzes your data and gives you answers. Real insights like: “Your welcome series converts 3x better than promotional flows. Tuesday sends at 10am outperform Thursday afternoons by 40%. Subject lines under 40 characters get 25% more opens.”

This isn’t generic best practices. It’s your data telling you what works for your audience.

Custom reporting without the setup: “Show me campaign performance for the last 90 days with open rates, click rates, and revenue by segment.”

The AI builds the report for you. No dashboard navigation. No export-to-spreadsheet workflow. Brian Whalley, Co-founder of Wonderment, describes it: “We can now instruct Klaviyo directly through Cursor to develop custom reporting dashboards. It’s such an elevated, easier experience compared to before.”

Profile and segment management: “Create a profile for [email protected] and add her to the VIP segment.” or “Show me all segments with declining engagement over the last 30 days.”

Done in seconds. No clicking through menus or searching for the right settings.

Wiehan Britz, CEO of Magnet Monster, sums it up: “The Klaviyo MCP server is like chatting directly with Klaviyo. I’m using conversational language to create and update profiles, organize segments, build campaigns, and analyze performance all in one place.”

Automation Setup in MailerLite

Setting up an abandoned cart workflow used to mean clicking through multiple screens, configuring triggers, and testing the logic.

Now it’s: “Create an automation for abandoned carts. Trigger after 3 hours. First email is a reminder. If no purchase in 24 hours, send a second email with a 10% discount code.”

The AI creates the complete automation and marks it ready to enable. You review it and turn it on.

Other quick wins with MailerLite’s MCP:

• Subject line suggestions based on your past performance, not generic advice
• Bulk cleanup of old campaign drafts with one prompt
• Schedule sends with natural language: “Send this next Tuesday at 9:30am”

Competitive Intelligence Workflow

Here’s where it gets interesting. One agency built a complete competitive intelligence system using MCP:

The workflow:

1. Dedicated Gmail account collects competitor emails, organized by brand labels
2. Weekly prompt: “Using your connection to my inbox, compare the last 5 campaigns from my competitors. What subject line patterns and offer strategies are consistently working?”
3. AI analyzes all competitor emails, identifies patterns, compares against your campaigns
4. Based on findings, it recommends new campaign ideas
5. Creates the campaign directly in Klaviyo
6. Logs all analysis to a Notion database for future reference

What used to take 2-3 hours of manual screenshot-and-analyze work now takes 10 minutes. And because it’s logged to Notion, the intelligence compounds over time. You’re building a competitive analysis database without the manual work.

CRM + Research Workflows in HubSpot

A sales team shared this example: “Find me the executives at Anthropic and do research on them.”

The AI searches the web, identifies the executives, pulls relevant information about each person, then creates contact records in HubSpot with research notes attached. All in one conversation.

Or this: “Show me deals that closed in the last 60 days, then show me the contacts associated with those deals.”

It pulls the deals, analyzes the close dates, finds associated contacts, and presents everything in a readable format. Then you can say: “Create Sally Jane as a contact at Acme Inc and associate her with the Boston property deal.”

Done.

You’re having a conversation with your CRM instead of clicking through dropdown menus and relationship fields.

Prompts to Get You Started

The best way to understand MCP is to try it. Here are prompts you can use with Klaviyo, MailerLite, HubSpot, or other connected platforms.

Performance Analysis & Reporting

“Give me a performance summary of campaigns sent in the last 90 days. Show me open rates, click rates, and revenue by campaign type.”

“Which of my campaigns in the last 60 days drove the most revenue? Break down by segment.”

“Compare my welcome flow performance to my abandoned cart flow. What’s working better and why?”

“Show me subject line patterns from my top 10 performing campaigns. What’s the common thread?”

“Analyze my Tuesday sends vs Thursday sends over the last quarter. Which day performs better for my audience?”

Campaign Creation

“Create a 20% off campaign for [product/service]. Target engaged subscribers. Write copy based on my previous campaigns. Include subject line and preview text.”

“Build a campaign announcing our fall sale. Use the same tone as my last promotional email. Target customers who purchased in the last 90 days.”

“Create a newsletter highlighting our three newest blog posts. Include a brief summary of each and a CTA to read more. Use casual, friendly copy.”

Automation Setup

“Create an automation for abandoned carts. Trigger after 3 hours. First email is a reminder. If no purchase in 24 hours, send a second email with a 10% discount code.”

“Set up a post-purchase follow-up sequence. First email after 3 days asking for feedback. Second email after 14 days with related product recommendations.”

“Build a re-engagement flow for subscribers who haven’t opened in 60 days. Include a win-back offer in the second email.”

Competitive Intelligence

“Using your connection to my Gmail, analyze the last 5 campaigns from [competitor name]. What subject line patterns are they using?”

“Compare my recent campaigns to what [competitor 1] and [competitor 2] sent in the last 30 days. What are they doing that I’m not?”

“Pull all competitor emails from the last week labeled ‘competitors.’ What offers and messaging themes are trending?”

Profile & Segment Management

“Create a profile for [[email protected]] and add them to the VIP segment.”

“Show me all segments with declining engagement over the last 30 days.”

“Create a new segment of subscribers who opened at least 3 campaigns in the last 14 days but haven’t purchased.”

“Find profiles who match [specific criteria] and add them to [segment name].”

Maintenance Tasks

“Give me a list of all campaign drafts. Delete any older than 60 days.”

“Show me all inactive automations. Which ones haven’t triggered in the last 90 days?”

“Pull a list of subscribers who hard bounced in the last 30 days.”

Custom Analysis

“Analyze my email content from the last quarter. What topics get the most engagement?”

“Compare revenue per email for promotional campaigns vs educational content.”

“Show me which day of the week and time of day generates the best engagement for my audience.”

“Identify my most engaged segment and tell me what makes them different from other subscribers.”

Start with one or two prompts that solve an immediate problem. See what works. Adjust from there.

Which Platforms Already Support MCP

MCP adoption is growing fast. Here’s where email marketers can use it today (That I know of):

Email Service Providers:

• Klaviyo
• HubSpot
• Brevo
• Braze
• ActiveCampaign
• Customer.io
• MailerLite

Project Management:

• Monday.com
• Asana
• Notion

Other:

• Zapier

What this means: You can build workflows that span multiple tools without being the one who manually transfers data between them.

Current limitation: Most MCP implementations work through desktop apps rather than web interfaces. The specifics vary by platform, but expect to set up local connections initially.

How Email Love Is Integrating MCP Into the Figma Plugin

The Email Love Figma Plugin already cuts email production time in half. You design in Figma, export clean HTML.

But there’s still a gap: you have to manually move your campaign brief into the design phase. You read the brief in Monday.com, Asana, or Notion, remember the details, then start designing in Figma.

MCP closes that gap.

Here’s what we’re building right now:

Step 1: Create your campaign brief in Monday.com, Asana, or Notion, like you normally do. Include goals, target audience, key messaging, offer details, and any asset links.

Step 2: Designer opens the Email Love Figma Plugin and connects to the brief through MCP.

Step 3: AI reads the brief, pulls relevant brand components from your design system, and generates a starting design directly in Figma. Images, copy blocks, CTAs, product information – all pulled automatically based on what’s in the brief.

The goal is to get designers 70-80% there. They focus on refinement—adjusting layouts, tweaking visual hierarchy, making sure it feels right for the audience. Not on setup work.

Time saved: 30-45 minutes per email. Multiply that across a week of Black Friday campaigns and you’ve bought back hours.

We’re rolling out MCP support over the next few months. If you’re already using the plugin, you’ll see updates soon!

What Happens Next

Look, it’s too early to say if MCP becomes the standard way everyone works or just another thing that sounded good in theory.

What I can tell you: it’s relatively easy to connect Claude or ChatGPT to the platforms mentioned above. The setup takes 10-15 minutes. And it’s worth exploring what’s possible.

The data analytics piece is what interests me most. Being able to ask questions about campaign performance and actually get answers—without building reports or exporting CSVs, that’s immediately useful. “Show me which subject line styles work best for my audience” or “Compare segment performance over the last quarter” and getting real answers in seconds? That saves actual time.

I’m less convinced about the content creation side. Personally, I don’t see myself generating email copy through MCP. Email content needs a human touch, and the risk of something feeling generic or off-brand isn’t worth the time saved. But that’s me.

Where this gets interesting: workflows that span multiple tools. Pulling competitive intelligence from Gmail, analyzing it, creating a brief in Notion, then using that context in your ESP—that’s where the real efficiency gains might be.

We’re building MCP support into the Email Love Figma Plugin because the use case makes sense: reading a campaign brief and pulling it into design automatically. That removes a manual step without trying to replace creative decisions.

If you’re curious, try connecting one of your tools and see what works for your workflow. Start with data analysis. That’s the lowest-risk, highest-value place to experiment.

Want updates on how we’re integrating MCP into the Figma Plugin? Subscribe to the Email Love newsletter.

Much love,
Andy

Email: [email protected]
Twitter: @emaillove