Win-Back Email Tests for DTC Brands
Lapsed customers are the cheapest revenue a DTC brand can recover — if you can find the message that earns the reopen. This guide shows how to run a DTC brands win-back email test with marginal, the hosted email MCP server that lets your AI agent generate variants, launch sends, and measure lift without leaving your editor.
Why win-back is a strong fit for A/B testing
Win-back flows target customers who already bought once but have gone quiet. Because the audience is finite and high-intent, every percentage point of open and click lift translates directly into recovered orders. That makes win-back one of the highest-leverage places for a DTC brand to run an email A/B test.
The variable that moves the needle most is the subject line. A discount-led subject reads differently than a curiosity-led or 'we miss you' angle, and the only honest way to know which works for your list is to test it head-to-head.
- Finite, well-defined segment (e.g. no purchase in 90 days)
- Clear conversion event to attribute lift against
- Strong emotional and offer-based angles to contrast
- Repeatable: rerun the winning structure each quarter
Setting up the test with marginal
marginal exposes four MCP tools that map cleanly onto a win-back workflow: generate_variants drafts competing subject lines, launch_test ships the managed send, get_results reports open and click metrics, and recommend_next suggests the follow-up angle. Your AI agent calls these directly once marginal is connected at https://marginal.sh/mcp.
A typical run: ask your agent to generate three win-back subject lines, launch a subject line A/B test across your lapsed segment, then check within-test lift after the send window closes.
- generate_variants — produce contrasting win-back subject lines
- launch_test — managed send with open/click tracking built in
- get_results — within-test lift metrics per variant
- recommend_next — what to try in the next round
Win-back subject line angles worth testing
DTC brands tend to default to a flat discount in win-back, but the test is more interesting when you contrast offer-driven copy against relationship-driven copy. Let the data decide rather than your gut.
- Offer-led: 'Here's 20% to come back'
- Curiosity-led: 'You left something behind'
- Restock/relevance: 'Your favorite is back in stock'
- Loyalty/recognition: 'We saved your spot'
AI agent email marketing in your existing tools
marginal is a hosted email MCP server, so there's nothing to deploy. Add the endpoint to any MCP-capable client — Cursor, Claude Desktop, Claude Code, Windsurf, Cline, Continue, Codex, or Zed — authenticate with your Bearer API key, and your agent can run the full win-back loop conversationally.
The free tier includes 100 experiments per month, which is plenty to iterate on subject line angles for a single DTC win-back flow. See the docs at https://marginal.sh/docs/ to wire up your first test.
- Hosted at https://marginal.sh/mcp — no self-hosting
- Bearer auth with a marg_live_ API key
- Works across major AI agent clients
- 100 experiments/month on the free tier
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