Re-engagement Email Tests for Online Courses
Dormant learners are the hidden churn of any online courses business. This guide shows how to run an online courses re-engagement email test with marginal, the hosted email MCP server, so your AI agent can find the subject line that pulls students back into a lesson.
Why re-engagement needs testing for online courses
Learners go quiet for predictable reasons: they finished a busy week, lost momentum after module two, or never opened the first nudge. A single generic "We miss you" email rarely cuts through. The phrasing that re-engages a coding bootcamp student differs from what works for a language learner or a professional certification.
An email A/B test removes the guesswork. By splitting your dormant segment and measuring opens and clicks per variant, you learn which framing actually drives returns — not which one sounded best in a meeting.
- Progress nudges: "You're 2 lessons from your certificate"
- Curiosity hooks tied to the next module
- Time-sensitive cohort or live session reminders
- Value recap: what they'll lose by quitting now
Running the test with marginal
marginal exposes four MCP tools your AI agent calls directly: generate_variants drafts competing subject lines, launch_test sends them across your dormant segment, get_results reports open and click metrics, and recommend_next suggests the follow-up. The endpoint at https://marginal.sh/mcp is hosted, so there's no server to run yourself.
A typical loop: ask your agent to generate five re-engagement subject lines for students inactive 14+ days, launch a subject line A/B test, then read within-test lift to pick the winner before your full re-send.
- generate_variants — re-engagement subject line ideas at scale
- launch_test — managed send with built-in split assignment
- get_results — opens, clicks, and within-test lift metrics
- recommend_next — data-backed next step for the segment
AI agent email marketing in your editor
Because marginal speaks MCP, AI agent email marketing happens inside the tools you already use. Connect Cursor, Claude Desktop, Claude Code, Windsurf, Cline, Continue, Zed, or OpenAI Codex and prompt your agent to design and launch the re-engagement test conversationally.
Authentication uses a Bearer API key (marg_live_...). Once configured, the agent handles variant generation and the email A/B test end to end while you review results.
- Iterate on copy without leaving your editor
- Let the agent run the subject line A/B test, not just draft it
- Free tier covers 100 experiments/month for steady testing
- Full setup at https://marginal.sh/docs/
A re-engagement testing playbook
Start narrow and let the data widen your reach. Define a dormant cohort, test two to four distinct angles, and only send the winner to the broader inactive list. Re-engagement is iterative — what wins this month may fade as your messaging fatigues, so retest quarterly.
- Segment by inactivity window (7, 14, 30 days)
- Test one variable — the subject line — per round
- Promote the winning variant to the full dormant segment
- Use recommend_next to plan the follow-up cadence
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