{"id":"search-mgr-hybrid-policy-sources-pick","audience":"managers","category":"ai-search","track":"research","technique":"verification","title":"Pick credible sources for a hybrid work policy proposal","free":false,"task":"You're proposing a hybrid schedule for your 300-person company and only trust sources that are from 2025 or later, surveyed companies (not just workers) with 100 to 1,000 employees, and explain their method. From the 7 results, write exactly 4 numbered lines: lines 1-3 each name one source that passes all three tests with a reason under 12 words, and line 4 starts with Skip: and names the most misleading result and why. Plain text, under 100 words.","materialLabel":"Your search results (attached to your prompt automatically)","material":"AI search results for \"hybrid work policy benchmarks mid-size companies\"\n\nA. Midsize Employers Council, \"2026 Workplace Policy Survey\" (Apr 2026): 640 companies with 100-999 employees; method and questions published.\n\nB. \"Remote work is dead, says CEO\" - Business Daily news article (2025), quotes one CEO of a 40,000-person bank.\n\nC. OfficeNest (sells hot-desk booking software), \"Why 3 days in office wins\" (2026): survey of its own customers, no method shared.\n\nD. Carver School of Management, \"Hybrid Schedules in Mid-Size Firms\" (Jan 2026): 212 firms with 150-800 staff; methods appendix included.\n\nE. Worker sentiment poll, PulseCheck (2026): 5,000 employees asked what they prefer; no company data.\n\nF. State Chamber of Commerce, \"Employer Practices Report\" (Sept 2025): 380 member businesses, 100-1,000 employees section; method on page 3.\n\nG. Blog post \"Hybrid stats you need\" (2022), no sources listed.\n\nMy notes: proposal to the exec team Nov 21. CFO asked \"what are similar companies actually doing?\"","rules":[{"type":"numberedItems","count":4},{"type":"maxWords","max":100},{"type":"plainText"}],"mistakes":[{"title":"Not giving your tests","why":"The AI picks the big worker poll because 5,000 sounds convincing, though it says nothing about what companies do.","fix":"Add: \"2025 or later, company surveys at 100-1,000 employees, method explained.\"","checks":["r0","r1"]},{"title":"Skipping the warning","why":"The AI lists good sources but doesn't flag the vendor survey dressed up as research.","fix":"Add: \"Line 4: Skip: the most misleading result and why.\"","checks":["r2"]},{"title":"Long or formatted output","why":"AI reviews all 7 results in bold.","fix":"Add: \"Exactly 4 numbered lines, plain text, reasons under 12 words.\"","checks":["numberedItems","maxWords","plainText"]}],"promptHints":[{"label":"Gives the three tests","patterns":["2025","100.{0,5}1,?000","method","compan(y|ies)"]},{"label":"Asks for exactly 3 picks","patterns":["\\b(3|three)\\b.{0,20}(source|pick)","exactly"]},{"label":"Asks what to skip","patterns":["skip","misleading","worst"]},{"label":"Caps the reasons","patterns":["under \\d+ words","short reason"]},{"label":"Asks for plain text","patterns":["plain","no bold","no markdown"]}],"hint":"The CFO asked what similar companies do. Give the AI tests that match that question so it skips loud but off-target sources.","rubricLabels":["Lines 1-3 pick exactly the Midsize Employers Council survey (A), the Carver School study (D) and the State Chamber report (F), in any order","Each of lines 1-3 gives a reason tied to the tests (2025 or later, company survey at 100-1,000 employees, method explained), and none of the picks is the worker poll (E) or the OfficeNest survey (C)","Line 4 (after its number) begins with Skip: and names the OfficeNest vendor survey (C), the single-CEO news article (B) or the 2022 blog (G), with the test it fails"]}