Testimonial Subject Line A/B Tests
Social proof works in the inbox — but only if the subject line earns the open. This guide shows how to run a testimonial subject line A/B test through marginal, the hosted email MCP server, so your AI agent can write, launch, and measure variants without leaving your editor.
Why testimonials belong in the subject line
Customer quotes, star ratings, and named results are some of the strongest open-rate levers in email marketing. The hard part is knowing which framing wins: a direct quote, a number, or a named brand. A subject line A/B test answers that with data instead of opinion.
Testimonial subjects fail in predictable ways — too long to render on mobile, quotes that feel generic, or social proof buried after the fold. Testing a few angles side by side surfaces the version your audience actually responds to.
- Quote-led: "'This saved us 10 hours a week' — Priya, Ops Lead"
- Number-led: "4,200 teams switched last quarter — here's why"
- Brand-led: "How Acme cut churn 18% with one change"
- Star-led: "⭐ 4.9/5 from 900+ reviews — see what they said"
Running the test through marginal
marginal exposes four MCP tools your agent calls directly. For a testimonial subject line A/B you'll typically use all of them in sequence, from drafting variants to reading the winner.
Because marginal is hosted at https://marginal.sh/mcp, there's no server to deploy — you authenticate with a Bearer API key (marg_live_...) and the tools handle managed sends plus open and click tracking.
- generate_variants — draft 3–5 testimonial-driven subject angles
- launch_test — send the split and start tracking opens and clicks
- get_results — pull within-test lift metrics per variant
- recommend_next — get the suggested winner and a follow-up angle
A practical workflow with an AI agent
AI agent email marketing shines here: you describe the testimonial you want to feature, the agent proposes framings, and marginal handles the experiment mechanics. A typical loop looks like the steps below.
Keep each variant testing a single idea — quote vs. number vs. brand — so the lift you read maps cleanly to a decision.
- Tell your agent the testimonial source and the offer or content behind the email
- Call generate_variants to produce distinct testimonial subject angles
- launch_test against your segment to split traffic evenly
- Read get_results once opens stabilize, then recommend_next for the follow-up
Getting started
marginal connects to Cursor, Claude Desktop, Claude Code, OpenAI Codex, Windsurf, Cline, Continue, and Zed, so you can run subject line A/B tests inside whatever environment your team already uses.
The free tier includes 100 experiments per month — enough to validate testimonial framings across several campaigns. See the docs at https://marginal.sh/docs/ for tool schemas and example calls.
- Endpoint: https://marginal.sh/mcp
- Auth: Bearer API key (marg_live_...)
- Registry: sh.marginal/mcp
- Free tier: 100 experiments/month
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