{"id":51464,"plugin_id":"plugin_asdk_app_6ac5f3338d4081918eeefc2733ba5f7a","kind":"skill","collection_source":"plugin_package","comparison_source":null,"observed_at":"2026-10-24T13:46:28.835Z","digest":"089eaac765c0e81071169bc247ded374cf75cc8c3143fedd756fb392fe3bf553","against":null,"payload":{"description":"Design and run an A/B test on a MisarMail campaign — subject line, content, sender name, or send time. Use for \"A/B test\", \"split test\", \"which subject works better\", or testing email variants.","included_files":[],"name":"ab-test-campaign","skill_md_contents":"---\nname: ab-test-campaign\ndescription: Design and run an A/B test on a MisarMail campaign — subject line, content, sender name, or send time. Use for \"A/B test\", \"split test\", \"which subject works better\", or testing email variants.\n---\n\n# A/B test a campaign\n\n## Check the audience is big enough — first\n\nCall `get_campaign` (or `list_campaigns`) for the recipient count.\n\nWith the default 20% sample split across two variants, an audience of 10,000\ngives ~1,000 per variant. Below that, normal variance swamps the effect and the\n\"winner\" is noise. **Say so plainly and recommend against testing** rather than\nrunning a test that produces a confident-looking but meaningless result.\n\n## Design\n\n`create_ab_test` takes `campaign_id`, a `type`, and 2–5 `variants`:\n\n| Type | Varies | Notes |\n|---|---|---|\n| `subject` | Subject line | Highest signal, easiest to interpret |\n| `content` | Body HTML | Test one change, not a redesign |\n| `from_name` | Sender name | Often larger effect than expected |\n| `send_time` | Delivery time | Needs a longer measurement window |\n\nFor subject tests, `generate_subject_lines` produces candidates. Test variants\nthat differ in *approach* (question vs. statement, specific vs. curiosity), not\nin wording trivia — two near-identical subjects cannot produce a real winner.\n\nSet `winner_metric` to match the goal: `open_rate` for subject tests,\n`click_rate` or `conversion_rate` for content.\n\n## Selecting the winner\n\n`select_ab_test_winner` **sends the winning variant to the entire remaining\naudience**. It is irreversible. Do not call it until results are in, and never\nwithout explicit confirmation.\n\nLet the sample run at least 4 hours — opens arrive over hours, and an early\nreading systematically favours whichever variant reached the more active\nsegment first.\n"},"changes":[],"summary":"First saved snapshot. No earlier version is available for comparison.","summary_kind":"deterministic","summary_metadata":{}}