Pet Products Review Request Email Test
Post-purchase review requests are where pet brands earn social proof — and where small subject line changes move open rates a lot. This guide shows how to run a pet products review request email test with marginal, the hosted email MCP server, so your AI agent can launch and measure the experiment for you.
Why review requests matter for pet products
Pet owners are emotionally invested in what they buy — food, toys, beds, supplements — which makes them unusually willing to leave reviews when asked well. The challenge is timing and tone: ask too clinically and the email gets ignored; lean into the pet and you get replies.
An email A/B test removes the guesswork. Instead of debating whether to reference the customer's order or their pet by name, you send both versions to comparable segments and let the open and click data decide.
- Test product-specific vs. brand-level review asks ("Rate your dog's new harness" vs. "How was your order?")
- Compare incentive framing — loyalty points vs. a simple thank-you
- Try timing language in subject lines: "Already settling in?" after a bed delivery
Setting up the review request subject line A/B test
marginal exposes four MCP tools your agent calls in sequence. For a review request campaign, you start with generate_variants to produce candidate subject lines tuned to the pet products vertical, then launch_test to send them as a managed split.
Because marginal handles the send, tracking, and within-test lift metrics, you don't wire up a separate sending stack. Each variant gets its own open and click measurement so the subject line A/B test produces clean comparisons.
- generate_variants — draft review-request subject lines (playful, direct, pet-name personalized)
- launch_test — split your post-purchase list and send managed variants
- get_results — pull open and click rates with lift between variants
- recommend_next — get the suggested winner and a follow-up direction
Driving the test from your AI agent
marginal is built for AI agent email marketing. Connect any MCP client — Cursor, Claude Desktop, Claude Code, Windsurf, Cline, Continue, Zed, or OpenAI Codex — to the endpoint at https://marginal.sh/mcp and authenticate with your Bearer API key.
From there you can prompt the agent in plain language: "Run a review request email test for our new cat tree, generate three subject lines, and report back the winner." The agent calls the email MCP server tools and returns results you can act on, all within the same workflow you already code in.
- Hosted — no self-hosted MCP server to maintain
- Free tier covers 100 experiments per month
- Endpoint: https://marginal.sh/mcp · Registry: sh.marginal/mcp
Reading results and iterating
After the test completes, get_results shows which review request subject line earned more opens and clicks, with within-test lift so you know the margin between variants — not just raw rates. For pet products, you'll often find personalization ("How's Bella enjoying the new bowl?") outperforms generic asks.
Use recommend_next to plan the follow-up: roll the winner out to the rest of your post-purchase flow, then test the next lever — preview text, send timing, or the call-to-action. See the docs at https://marginal.sh/docs/ for tool parameters and examples.
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