Shopify Cross-Sell Email A/B Tests with marginal
Cross-sell emails are where Shopify stores earn their next order — but the subject line decides whether anyone opens them. This guide shows how to run a Shopify cross-sell email test using marginal, the hosted email MCP server that lets your AI agent generate variants, launch tests, and read lift metrics without leaving your editor.
Why cross-sell emails need their own test
Cross-sell campaigns target buyers who already trust you, recommending complementary products after a purchase. Because the audience is warm but the offer is incremental, small framing changes move revenue. A subject line that leads with the original product ('Goes great with your new boots') often beats a generic discount line.
Treating each cross-sell flow as a measurable Shopify cross-sell email test — rather than a one-off send — gives you data to compound. Every campaign teaches you what your customers respond to.
- Reference the prior purchase to anchor relevance
- Test product-led versus benefit-led subject lines
- Segment by category so recommendations stay credible
- Measure click-through, not just opens, to gauge real intent
Running the test through marginal
marginal exposes four MCP tools your AI agent calls directly: generate_variants drafts subject line options, launch_test ships them as a managed send, get_results returns open and click data, and recommend_next suggests the follow-up. The whole loop runs from an agent in Cursor, Claude Code, Windsurf, or any MCP client.
Because marginal is hosted at https://marginal.sh/mcp, there's no infrastructure to stand up. You connect with a Bearer API key and start an email A/B test in minutes.
- generate_variants — produce multiple cross-sell subject lines
- launch_test — run a managed subject line A/B test
- get_results — pull within-test lift, opens, and clicks
- recommend_next — let the AI agent propose the next iteration
A workflow for Shopify stores
Start with a clear hypothesis tied to a Shopify segment — for example, customers who bought a coffee grinder in the last 30 days. Ask your agent to generate four cross-sell subject lines positioning beans, filters, and a scale as add-ons.
Launch the email A/B test, let it run until you have meaningful open and click volume, then read the results. marginal's within-test lift metrics tell you which variant won so your AI agent email marketing loop can scale the winner and retire the rest.
- Define the post-purchase segment and the complementary products
- Generate and launch a subject line A/B test for the cross-sell offer
- Track opens and clicks with managed sends
- Use recommend_next to evolve the campaign with each cycle
Getting started
The free tier includes 100 experiments per month, which is enough to iterate through several Shopify cross-sell flows. Point your MCP client at https://marginal.sh/mcp, add your Bearer key, and your agent can launch its first test.
See the docs at https://marginal.sh/docs/ for tool schemas and examples covering subject line A/B tests and result reporting.
- Hosted MCP server — no self-hosting required
- Works with Cursor, Claude Desktop, Cline, Continue, Zed, and more
- Free tier: 100 experiments per month
- 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