Question vs Statement Subject Line Tests
Does "Ready to cut your churn?" beat "Cut your churn this quarter"? A question vs statement subject test answers that for your specific list — and marginal lets an AI agent run the whole experiment from your editor.
Why question vs statement matters
Questions and statements pull different psychological levers. A question invites the reader to fill in a gap and can boost curiosity, while a declarative statement signals confidence and clarity. Which wins depends entirely on your audience, offer, and inbox context — guessing is how good subject lines die.
Instead of debating phrasing in a meeting, set up a subject line A/B test and let open and click data decide. The framing that works for a SaaS trial nurture rarely matches what works for a re-engagement blast.
- Question framing: drives curiosity, can feel softer or more conversational
- Statement framing: direct value, works well for time-bound or benefit-led sends
- The only reliable answer is a head-to-head test against your own list
Running the test with marginal's email MCP server
marginal is a hosted email MCP server that exposes a small set of tools to your AI agent. You describe the test in plain language and the agent generates both framings, launches the send, and reports results — no dashboard juggling required.
A typical question vs statement workflow uses four tools in sequence.
- generate_variants — produce a question-framed and statement-framed subject for the same email
- launch_test — split your audience and send both variants
- get_results — pull open and click rates with within-test lift metrics
- recommend_next — get a suggested follow-up test based on what won
Designing a clean comparison
To isolate the question vs statement variable, keep everything else constant: same body copy, same send time, same audience segment. The only difference should be the subject framing, so the lift you measure actually reflects phrasing.
marginal tracks opens and clicks per variant and reports within-test lift, so you can see whether the question genuinely outperformed the statement or whether the gap is noise.
- Hold body, CTA, and segment identical across both variants
- Match send timing so time-of-day doesn't skew opens
- Read within-test lift, not just raw rates, before declaring a winner
- Repeat the winning frame in a follow-up to confirm it generalizes
Where AI agent email marketing fits
Because marginal speaks MCP, any compatible client can drive these experiments — Cursor, Claude Desktop, Claude Code, Windsurf, Cline, Continue, Zed, and OpenAI Codex among them. You connect once to https://marginal.sh/mcp with a Bearer API key and your agent handles variant generation and analysis.
This makes recurring AI agent email marketing experiments cheap to run. The free tier covers 100 experiments per month, which is enough to test question vs statement framing across several campaigns before you commit copy. Docs live at https://marginal.sh/docs/.
- Hosted — no self-managed infrastructure to run
- Registry: sh.marginal/mcp · Endpoint: https://marginal.sh/mcp
- Free tier: 100 experiments/month to iterate on subject framing
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