How to Use the Kit MCP to Build an Email Analytics Dashboard
I Let AI Audit My Entire Email List. Here's What It Found.
Most creators have no real visibility into their email list. They know the total subscriber count. Maybe they check open rates after each send. But the deeper picture — where subscribers come from, which email formats actually drive sales, why people leave, what subject line patterns your specific audience responds to — that usually lives buried in dashboards you don't have time to dig through.
I recently ran a full audit of my HobbyScool email list using the Kit MCP — a connection that gives Claude AI direct, live access to my Kit account. No exports. No CSV uploads. Just prompts, and within minutes, a complete picture of everything happening across 41,000 subscribers, 200+ broadcasts, 100+ sequences, and hundreds of tags.
Here's what I found. And here's the exact setup and prompts you can use to run the same analysis on your own list.
What Is the Kit MCP?
MCP stands for Model Context Protocol. It's a standard that lets AI tools like Claude connect directly to other platforms and pull real data in real time.
The Kit MCP is a connector specifically built for Kit (formerly ConvertKit). Once it's installed inside Claude, you can ask questions like "What are my open rates for the last 90 days?" or "Show me which tags have the most subscribers" — and Claude retrieves the actual live data from your account and analyzes it on the spot.
This is a significant shift from the old workflow. No more manually pulling reports, exporting data, pasting it into a spreadsheet, and trying to find patterns yourself. The AI does all of that in seconds.
Subscriber counts and 90-day growth stats · Tags and their subscriber counts · Forms and landing pages · Sequences with subscriber counts and email counts · Broadcast performance: open rates, click rates, unsubscribes, total clicks · Cancelled subscriber records with tags · Account info including sending addresses and plan details
How to Set It Up
There are two ways to get started depending on how you use Claude.
If you're using Claude's Cowork mode: Add the Kit MCP as a connector directly from your Cowork settings. It will appear in the connectors marketplace — search for "Kit" and connect it with your Kit API key.
If you're using the Claude API or Claude desktop app: The community-built Kit MCP server is available on GitHub. You'll need your Kit API key (found under Advanced settings in your Kit account) and a few minutes to configure it.
Kit also maintains their own official MCP documentation at developers.kit.com/mcp. That's the best place to start if you want the official setup guide straight from Kit.
The community GitHub connector is at github.com/aplaceforallmystuff/mcp-kit — it's actively maintained and covers all the endpoints used in this analysis.
The Dashboard Prompt — Start Here
This was the first prompt I ran. It pulls everything at once and compiles it into a visual dashboard.
Using the Kit MCP for [your brand name], build me a dashboard with all the information you can find in my Kit account. Pull subscriber counts, list growth, tags, forms, sequences, and broadcast performance. Then give me your top insights and anything that surprises you.
That single prompt pulled my account details, 90-day growth data, all tags (200+), all forms (100+), all sequences (100+), and 50 completed broadcasts — simultaneously. It then organized everything into a visual HTML dashboard and surfaced the insights I hadn't noticed on my own.
Here's what it found for HobbyScool.
What the Data Showed — Real Numbers From a 41K List
The open rate was the first number that stood out. A 54% average open rate on a list of nearly 41,000 people is not normal. The industry benchmark for hobby and craft newsletters is 25–35%. HobbyScool is running almost double that.
That's not an accident. It's a signal about what the audience responds to — which we'll come back to in the subject line section.
The Summit Machine at Work
HobbyScool runs a series of free online summits throughout the year. The Kit data made the impact of this strategy very visible.
In 90 days, 12,270 new subscribers came in — almost entirely through summit registrations. The account had 8 summits scheduled in 2026 alone. Each one comes with its own registration form, Facebook Lead Ad form, VIP pass purchase sequence, and abandoned cart sequence. The infrastructure for each event is fully built out inside Kit before the summit opens.
The net growth (new subscribers minus unsubscribes) was +3,366. That's real, compounding list growth — driven by a repeatable event system, not one-off tactics.
"The summit model isn't just growing the list — it's creating a deeply activated audience willing to click through every single day."
— From the HobbyScool Kit MCP analysisThis is the actual dashboard Claude generated from my HobbyScool Kit account using Prompt 1. Scroll inside to explore the full output.
Prompt 2 — The Unsubscriber Analysis
This is where things got interesting. The second prompt digs into who's leaving and why.
Analyze my unsubscribers. Bucketize them by the forms and tags they came in through, find the commonalities, and tell me where I'm losing people. Then suggest how I could re-engage them before they fall off the list.
Claude pulled the 100 most recent cancelled subscribers, including their tags, and looked for patterns. The finding was clear and a little surprising.
Almost Every Recent Unsubscriber Had One Tag in Common
Nearly every person who unsubscribed in the past 30 days carried the tag: [Cold] In a cold automation.
This means the unsubscribes weren't random. They weren't people rage-quitting after a bad email. They were people who had stopped engaging weeks or months earlier — and the cold re-engagement sequence was finally reaching the end of the line for them.
The churn problem isn't really a churn problem. It's a cold subscriber problem. And the cold sequence is doing exactly what it's supposed to do.
Five Buckets of Unsubscribers
Summit Registrant → Went Cold → Exited
They came in through a summit, never bought anything, went cold within a few weeks of the event ending, and eventually got cleaned out by the cold sequence. This is expected behavior for event-driven list growth.
Multi-Summit Hopper → Never Converted → Gone
These people attended 3, 4, even 5+ summits across multiple years. They love the free content. But they never bought a VIP pass or a product. Eventually they went cold and left.
Paying Customer → Still Unsubscribed ⚠️
This is the most important finding. Real buyers — people who purchased VIP passes — still went cold and eventually unsubscribed. One subscriber had purchased two separate VIP passes across two different summits before leaving.
Lead Magnet → Never Converted → Gone
They came in for a freebie, went through the delivery sequence, and never connected to the summit ecosystem or bought anything. The lead magnet never bridged them into becoming regular readers.
Long-Tenure Subscriber → Finally Left
Subscribers from 2023 who stayed for 2-3 years, attended summits regularly, and eventually moved on. These are the least concerning. A long relationship that ends naturally is healthy.
Prompt 3 — Broadcast Format Analysis
The third prompt looks at which types of emails actually move the needle.
Analyze all my broadcasts. Which email format drives the most sales and the best click-through rates? Show me the patterns in what's working versus what's flopping.
With 200+ broadcasts analyzed, a clear hierarchy emerged.
The Format Leaderboard
Summit attendee emails are in a category of their own. When someone has already registered for an event and Day 1 opens, they click. The "Art of Expression Summit: Day 1 Has Officially Begun!" email hit a 69.8% open rate and a 37.9% click rate. That's not email marketing — that's an audience that showed up ready.
But the finding that surprised me most? Free gifts dramatically outperform hard sales.
The Finding That Changes How I Think About Promotions
HobbyScool ran a 12 Days of Christmas giveaway series — one free digital product delivered per day for 12 days. And a Black Friday sale series — multiple emails pushing a time-limited offer.
The giveaway emails averaged a 4–6% click rate. The Black Friday sale emails averaged 0.63–1.56%.
The free gift emails generated roughly four times the clicks of the sale emails — on a list of similar size.
This audience doesn't respond to urgency language. They respond to generosity. The strategy that works is: lead with a gift, let the offer follow naturally.
Black Friday/hard-sell emails: open rates drop 5-8 points, click rates hit their lowest. All-caps subject lines, countdown timers, and "deal ends tonight" framing consistently underperform on engaged, trust-based lists. If your audience has been trained to open your emails because the content is genuinely good, promotional language reads as out of character.
Prompt 4 — Subject Line Analysis
This is the prompt I'd recommend running for every creator with 6+ months of send history. The patterns are worth knowing.
Analyze my subject lines across all broadcasts. Identify the formula that works best for my list and my open rates. Give me the patterns so I can replicate them.
After analyzing 200+ subject lines alongside open and click rates, six formulas emerged as consistent top performers.
6 Subject Line Formulas That Work
The Reframe
Removes shame, transfers blame to something external, promises a better path. Works especially well when the content solves a common frustration your reader carries.
The Unexpected Metaphor
Creates a pattern interrupt. The reader didn't expect to see that word in your subject line. Curiosity pulls the open. Works best for craft, creative, and lifestyle content.
The Specific Craft Hook
Specificity plus accessibility is the combination your audience can't resist. The more specific the skill, technique, or material, the better this formula works.
The Permission Statement
Your audience wants someone to tell them it's okay to do things differently. Give them that permission in the subject line itself. Especially strong for January/February and self-care themes.
The "You're Already There" Nudge
Reduces the perceived distance to action. "You're already halfway there" removes the intimidation of starting. High open rate and tends to lead to actual clicks and action.
The Persona Call-Out
Creates an immediate "that's me!" moment of recognition. Strong for opens, but needs a compelling hook inside the email to convert recognition into clicks.
What Kills Open Rates
Five patterns consistently push open rates 5-8 points below baseline:
Countdown urgency ("Lifetime Access closes at 9am tomorrow! ⏰") dropped one email to a 46.7% open rate — nearly 10 points below the newsletter average.
All-caps sale labels ("CYBER MONDAY: Every deal is back!") hit 49%. Your list has trained itself to tune out shouting.
Final hours framing ("Final hours. All deals end at 9am") consistently underperforms. It reads like the tenth sale email they saw that week.
Social proof without specificity ("Thousands have already registered — have you?") doesn't move this audience. They don't care if thousands registered. They care if it's right for them.
Indecision framing ("Still thinking about it?") subtly positions the reader as someone who can't make up their mind. That's not the frame you want to open with.
Prompt 5 — Bake It Into Your Writing Workflow
This last prompt is the one I'd add to every email session going forward.
Before I finalize this email, pull my top-performing subject lines from Kit and generate 3 subject line options based on the formulas that work best for my list. Include matching preview text for each.
When you run this prompt alongside the Kit MCP, Claude pulls your recent broadcast stats, identifies which subject line patterns are performing on your specific list, and generates three options across different formulas. You get a data-informed shortlist instead of starting from a blank page.
This isn't about letting AI write your subject lines. It's about using your own data to inform the decision. There's a big difference.
Key Takeaways for Any Kit User
Here's what I'd want every creator to take from this analysis — regardless of list size.
Your unsubscribes are probably not a content problem. They're almost certainly a cold subscriber problem. If you have a re-engagement sequence in Kit, the exits are likely happening at the end of that sequence — which means the automation is working correctly. The real question is what happens in the 4-8 weeks after each new subscriber joins, before they go cold in the first place.
Paying customers need their own email track. If a subscriber buys something from you — a VIP pass, a course, a membership — they shouldn't receive the same newsletter as someone who's never given you a dollar. They need content connected to their purchase. This is an often-missed retention opportunity.
Free gifts consistently outperform sale urgency. If you're running Black Friday or big sale campaigns, test leading with a free gift first. Let the generosity do the heavy lifting. The offer can follow naturally. Manufactured urgency tends to depress both open rates and click rates on engaged creator audiences.
Your best subject line formula is hiding in your own data. The six formulas I extracted from HobbyScool are specific to that audience and that content. Your audience may respond differently. Run the subject line analysis prompt on your own broadcasts. The pattern will surface.
The summit Day 1 email is your highest-leverage send. If you run any kind of live event — summit, challenge, workshop, webinar — the email you send when Day 1 opens is your single highest-performing email. Treat it that way. Make it effortless to click through to the content.
This is the full analysis Claude generated from Prompts 2 through 5 — covering unsubscriber buckets, broadcast format rankings, subject line formulas, and the writing workflow. Scroll inside to explore.
Your Turn: Run This on Your Own List
Here are all five prompts in sequence. Run them in order, with the Kit MCP connected to your Claude account.
1. Using the Kit MCP for [your brand], build me a dashboard with all the information you can find in my Kit account. Pull subscriber counts, list growth, tags, forms, sequences, and broadcast performance. Then give me your top insights and anything that surprises you. 2. Analyze my unsubscribers. Bucketize them by the forms and tags they came in through, find the commonalities, and tell me where I'm losing people. Then suggest how I could re-engage them before they fall off the list. 3. Analyze all my broadcasts. Which email format drives the most sales and the best click-through rates? Show me the patterns in what's working versus what's flopping. 4. Analyze my subject lines across all broadcasts. Identify the formula that works best for my list and my open rates. Give me the patterns so I can replicate them. 5. From now on, when I write an email, call the Kit MCP and generate 3 subject line options based on my top-performing subject lines.
The whole audit takes one session. The dashboard and analysis files are exportable and shareable. And once you've run it once, you have a baseline to compare against every quarter.
This is exactly the kind of system-level visibility that used to require a data analyst or an expensive tool. Now it's five prompts.
Is Your Creator Business Built to Scale?
Take the free Creator's MBA Scorecard to find out where your biggest growth opportunities are — and what to fix first.
Take the Free Scorecard →Frequently Asked Questions
The Kit MCP (Model Context Protocol) is a connector that gives Claude direct access to your Kit account. Once installed, Claude can pull your subscriber data, tags, forms, sequences, and broadcast stats in real time — without you exporting anything manually. You just ask Claude questions in plain English and it retrieves and analyzes the data for you.
The Kit MCP can access subscriber counts and growth stats, tags and their subscriber counts, forms and landing pages, sequences with subscriber and email counts, broadcast performance including open rates, click rates, unsubscribes, and total clicks, cancelled subscriber records including tags, and account details like sending addresses and plan information.
Based on analysis of 200+ broadcasts from a 41K subscriber list, six formulas consistently outperform others: The Reframe ("It's not you. It's the system."), The Unexpected Metaphor ("Your recycling bin is full of beautiful things"), The Specific Craft Hook ("Watercolor for beginners — everything you need to know"), The Permission Statement ("You're allowed to slow down"), The Nudge ("You're already halfway there"), and The Persona Call-Out ("This is for the 'I don't know what I want yet' crowd"). Urgency language, CAPS, and countdown framing consistently depress open rates.
For most email creators using Kit, the majority of unsubscribes are triggered by cold re-engagement sequences — not by random dissatisfaction. The pattern is: subscriber enters through a summit or lead magnet, goes cold within 4-8 weeks, and then 2-6 months later gets cleaned out by the cold automation. This is actually healthy list hygiene. The more concerning issue is when paying customers go cold, which signals a gap in post-purchase nurture.
Summit or event attendee Day 1-3 emails consistently deliver the highest click rates (25-40%), because subscribers have already opted in and are actively engaged. For the main newsletter list, skill-based tutorial emails outperform lifestyle, promotional, and philosophical content on clicks. Free gift or challenge emails also significantly outperform hard-sell promotional emails on both open and click rates.
