Testing Welcome Series Subject Lines in Zed
Your welcome series sets the tone for every subscriber relationship — and the subject line decides whether anyone reads it. With marginal connected to Zed as an MCP server, you can generate, launch, and measure subject line A/B tests for each welcome email straight from your editor.
Why the welcome series deserves a subject line test
Welcome emails consistently earn the highest open rates of any lifecycle message, which makes them the best place to learn what resonates with new subscribers. A subject line test on email one tells you whether a warm greeting beats a value-first hook before that pattern compounds across the rest of the series.
Because welcome series messages fire automatically and at volume, even a small lift per step adds up. Running a structured subject line A/B test on each stage turns vague intuition into data you can act on.
- Email 1: greeting vs. immediate offer or expectation-setting
- Email 2: education framing vs. social proof
- Email 3: nudge to first action vs. curiosity-driven prompt
- Test one variable at a time so each result is attributable
Running the workflow inside Zed
Zed's MCP support lets an AI agent call marginal's tools directly in your workflow. Instead of switching to a separate dashboard, you describe the welcome email and the agent handles variant generation, launch, and measurement through the email MCP server.
The four marginal tools map cleanly to a welcome series test loop:
- generate_variants — draft multiple subject lines for a welcome email from a short brief
- launch_test — start the subject line A/B test as a managed send
- get_results — pull open and click tracking plus within-test lift metrics
- recommend_next — get a data-backed suggestion for the winning subject line and the next step
A practical example in Zed
Suppose you're onboarding new trial users. Ask the agent in Zed to generate five welcome subject lines that balance friendliness with a clear next action, then launch a test across your new-subscriber segment. marginal handles the send and tracks opens and clicks so you can compare variants on equal footing.
Once enough data lands, call get_results to read the within-test lift and recommend_next to decide which line graduates into your evergreen welcome flow. This keeps AI agent email marketing grounded in measured outcomes rather than guesswork.
- Brief the agent with audience, tone, and goal for the welcome email
- Review generated variants before launching
- Let the test run until results stabilize, then promote the winner
- Repeat per step to optimize the full welcome series
Getting connected
marginal is a hosted MCP server, so there's nothing to deploy or self-host. Point Zed at the endpoint at https://marginal.sh/mcp and authenticate with your Bearer API key (marg_live_...). The free tier covers 100 experiments per month, enough to test every stage of a welcome series.
See the docs at https://marginal.sh/docs/ for client setup details and tool reference once your key is in place.
- Endpoint: https://marginal.sh/mcp
- Auth: Bearer API key (marg_live_...)
- Free tier: 100 experiments/month
- Works with Zed and other MCP clients like Cursor, Claude Code, and Windsurf
Get started with marginal
marginal is a hosted email marketing MCP server at marginal.sh. Sign up free, create an API key, and connect https://marginal.sh/mcp from Cursor, Codex, or Claude Desktop.
- 100 experiments/month on the free tier
- Four MCP tools: generate_variants, launch_test, get_results, recommend_next
- Listed in the MCP Registry as sh.marginal/mcp