Demo Request Follow Up Subject Test
The follow-up email after a demo request is one of the highest-leverage messages in your funnel — yet most teams ship a single subject line and hope. This guide shows how to run a demo request follow up subject test with marginal, the hosted email MCP server, so an AI agent can generate variants, launch the test, and report which line gets more opens.
Why the demo follow-up subject deserves a real test
Prospects who request a demo are warm, but their attention decays fast. The subject line on your follow-up determines whether the email gets opened in the first hour or buried by tomorrow's inbox. Small wording shifts — urgency, personalization, value framing — can move open rates meaningfully on this audience.
Because demo requests arrive steadily, you usually have enough volume to learn from a subject line A/B test instead of guessing. The goal is to find the framing that consistently earns the open, then let it inform future sequences.
- Compare a curiosity angle vs. a direct "book your time" ask
- Test whether including the rep's first name lifts opens
- Measure urgency phrasing against a calm, helpful tone
- Check if referencing the product use case beats a generic recap
Running the test with marginal's email MCP server
marginal exposes email experimentation as MCP tools at https://marginal.sh/mcp, so an AI agent inside your editor can run the whole loop without you switching to a separate dashboard. You describe the follow-up, and the agent handles variant creation, send, and analysis.
The four tools map cleanly to a subject test workflow:
- generate_variants — draft multiple demo follow-up subject lines from your brief
- launch_test — send the variants as a managed A/B test with open/click tracking
- get_results — pull open rates and within-test lift between subjects
- recommend_next — get a suggested winner and the next subject to try
A practical AI agent workflow
Connect marginal to your MCP client — Cursor, Claude Desktop, Claude Code, Windsurf, Cline, Continue, Zed, or OpenAI Codex — using a Bearer API key (marg_live_...). From there, AI agent email marketing becomes a conversation rather than a manual setup chore.
A typical run looks like: ask your agent to generate five subject variants for the demo follow-up, launch the test across your inbound list, wait for opens to accumulate, then ask for results and a recommendation. The agent reads the lift metrics and tells you which line to keep sending.
- Prompt: "Generate 5 subject lines for a demo request follow up."
- Prompt: "Launch a subject line A/B test with these variants."
- Prompt: "Show me open-rate lift and recommend the winner."
What you can do on the free tier
marginal is fully hosted — there's nothing to self-host or operate. The free tier includes 100 experiments per month, which is plenty for iterating on a demo request follow up subject test and a few adjacent sequences before committing to a winner.
Open and click tracking are built in, and within-test lift metrics make it easy to see whether one subject genuinely outperformed the rest. Full setup details and tool references live in the docs at https://marginal.sh/docs/.
- Hosted MCP server — no infrastructure to run
- 100 experiments/month on the free tier
- Built-in open/click tracking and lift metrics
- Endpoint: https://marginal.sh/mcp
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