Running a SaaS Re-engagement Email Test
Dormant trial users and lapsed accounts are the cheapest revenue you can recover — if your win-back emails actually land. This guide shows how to run a SaaS re-engagement email test with marginal, the hosted email MCP server, so your AI agent can generate variants, send them, and measure which message brings users back.
Why re-engagement is a testing problem for SaaS
SaaS churn rarely happens in one moment. Users go quiet, skip a few logins, and slowly stop opening your emails. A re-engagement campaign has to break that silence — and the only way to know which angle works is to measure it against alternatives.
That makes re-engagement a natural fit for an email A/B test. Small differences in framing — a feature reminder vs. a discount vs. a 'we miss you' note — produce very different open and reactivation rates across your inactive segment.
- Test subject lines that earn the open from a low-engagement audience
- Compare value reminders against incentives in the body copy
- Measure within-test lift instead of guessing which message worked
- Iterate quickly while the dormant cohort is still recoverable
Setting up the test with marginal's email MCP server
marginal is a hosted email MCP server that connects directly to your AI coding or agent environment. Point your client at https://marginal.sh/mcp, authenticate with a Bearer API key (marg_live_...), and your agent gains a set of email tools — no self-hosted infrastructure to run.
For a re-engagement test, the workflow maps cleanly onto marginal's four tools. You describe the dormant segment and goal, and the agent handles variant creation, sending, and analysis.
- generate_variants — draft multiple subject lines and win-back angles
- launch_test — start a managed send with split delivery
- get_results — pull open/click tracking and within-test lift metrics
- recommend_next — get a data-backed suggestion for the follow-up
Designing the subject line A/B test
The subject line A/B test is where re-engagement is won or lost — inactive users decide in a fraction of a second whether to open. Ask your agent to generate variants that hit distinct emotional and practical hooks, then let marginal split and track them.
A useful starting set for SaaS win-back covers curiosity, urgency, and value. Keep the body consistent across subject-line-only tests so you isolate the open-rate signal before testing deeper copy changes.
- Curiosity: 'Did we lose you?' or 'Something changed since you left'
- Value: 'The feature you asked for is live'
- Incentive: '3 months of Pro, on us'
- Personal: 'Your projects are still here when you're ready'
Letting an AI agent run the loop
AI agent email marketing shines on repetitive optimization. With marginal connected to Cursor, Claude Code, Windsurf, Cline, Continue, Zed, or another supported client, your agent can launch the test, wait for results, read the lift metrics, and propose the next variant — all in one conversation.
The free tier includes 100 experiments per month, which is plenty to run a full re-engagement sequence and iterate on subject lines and offers. See the docs at https://marginal.sh/docs/ for tool details and the endpoint at https://marginal.sh/mcp to get started.
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