{"id":"res-mkt-benchmark-stats-right-segment","audience":"marketing","category":"research","technique":"grounding","title":"Pull email benchmarks for the right segment, with page numbers","free":false,"task":"You're building a pitch slide for a retail client and need email marketing benchmarks from the industry report. You want exactly 3 numbered stats for B2C retail only, each under 15 words and including the page, like \"(p. 8)\". Use the corrected figure where the report fixes itself, use only the report, and output nothing else in plain text.","materialLabel":"Excerpts from the email benchmark report (attached to your prompt automatically)","material":"2026 EMAIL BENCHMARK REPORT - Postmark Analytics Group (excerpts)\n\np. 6 - All industries: average open rate 34.1%, click rate 2.6%.\np. 8 - B2C retail: average open rate 31.8%, click rate 2.1%, unsubscribe rate 0.24%.\np. 9 - B2B software: average open rate 39.5%, click rate 3.4%.\np. 11 - B2C retail: emails sent on Tuesdays had the highest click rate (2.5%).\np. 12 - B2C retail: subject lines under 40 characters had open rates 4.2 points higher than longer ones.\np. 15 - Forecast: \"We expect B2C retail open rates to reach 35% by 2028.\"\np. 30 - Errata: The B2C retail click rate on p. 8 should read 1.9%, not 2.1%.\n\nClient: Fenwick Home Goods (B2C retail). Their current open rate: 27%.","rules":[{"type":"numberedItems","count":3},{"type":"plainText"}],"mistakes":[{"title":"Mixing segments","why":"The report leads with all-industry numbers, and the AI may use them or the higher B2B figures.","fix":"Say: \"B2C retail stats only.\"","checks":["r0"]},{"title":"Missing the errata","why":"Corrections sit at the back of reports, and the AI copies the first number it sees.","fix":"Add: \"Use the errata's corrected figure where there is one.\"","checks":["r1"]},{"title":"Turning it into advice","why":"The AI compares the client's 27% to every benchmark and suggests fixes.","fix":"Say: \"Output only the 3 stats.\"","checks":["r2"]},{"title":"Not saying how many","why":"The AI decides the length itself and usually returns more stats than you can use.","fix":"Be exact: \"exactly 3 numbered stats.\"","checks":["numberedItems"]},{"title":"Getting markdown back","why":"AI defaults to bold labels and headings, which show up as stray symbols in a slide.","fix":"Ask for \"plain text, no bold, no headings.\"","checks":["plainText"]}],"promptHints":[{"label":"Names the segment","patterns":["b2c","retail"]},{"label":"Handles the errata","patterns":["errat","correct","1\\.9"]},{"label":"Asks for pages","patterns":["page","p\\.\\s*\\d","\\(p\\."]},{"label":"Says exactly how many","patterns":["\\b(3|three)\\b.{0,25}(stat|line|number)","exactly \\d+"]},{"label":"Asks for plain text","patterns":["plain","no bold","no markdown","no formatting","no headings"]}],"hint":"Benchmark reports cover many segments and fix errors in the back. Tell the AI which segment you need, to use corrections, and to cite pages.","rubricLabels":["All 3 stats are B2C retail figures (from pp. 8, 11 or 12), not the all-industry or B2B software numbers, and none is the 2028 forecast","If the B2C retail click rate is used, it is 1.9% (the p. 30 errata), not 2.1%, and every other number matches the report","Each line includes a correct page reference, and nothing else is output: no comparison to Fenwick's 27%, no recommendations"]}