{
  "version": "1.0",
  "authored_date": "2026-10-29",
  "status": "editorial source contract before fixture generation and initial AI run",
  "scenario": "GreenLeaf Wellness UK launch, Week 2",
  "disclosure": "GreenLeaf and the business context are fictional. This fixture is a new synthetic reconstruction; no underlying reporting CSV was recovered from the course. Course statements below are scenario facts or teaching principles, not real-world evidence. Relative source paths resolve from this JSON file. Original paid slide images and asset exports are not part of the public evidence pack.",
  "governing_documents": [
    "../../../../doc/2026-10-29-reporting-course-alignment.md",
    "../../../../doc/2026-10-29-connected-case-study-direction.md",
    "../../../../docs/superpowers/specs/2026-10-29-reporting-worked-example-design.md"
  ],
  "course_sources": [
    {
      "id": "business-context",
      "classification": "course_scenario_fact",
      "source": "../../../course/ai-native-media-operations/slides/module-2-research-strategy.md#m2-meet-greenleaf-wellness-our-running-example",
      "used": "Fictional subscription supplement company, US-to-UK expansion, daily greens/protein/adaptogen products, ingredient traceability and transparent dosing proposition.",
      "boundary": "Course growth, price positioning and named category competitors are fictional teaching context, not verified current market data. No health efficacy claim is established."
    },
    {
      "id": "persona-hypothesis",
      "classification": "course_scenario_fact",
      "source": "../../../course/ai-native-media-operations/slides/module-2-research-strategy.md#m2-oliver-the-audit-davies-profile",
      "supporting_sources": [
        "../../../course/ai-native-media-operations/slides/module-2-research-strategy.md#m2-oliver-s-decision-criteria-and-objections",
        "../../../course/ai-native-media-operations/slides/module-2-research-strategy.md#m2-validate-with-real-research"
      ],
      "used": "Oliver supplies a skeptical proof-seeking buyer hypothesis and an interview question to validate, not a real interviewed customer.",
      "boundary": "The fixture's two audience buckets and illustrative message wording are newly authored. Course persona conversations do not validate them."
    },
    {
      "id": "kpi-ladder",
      "classification": "course_scenario_target",
      "source": "../../../course/ai-native-media-operations/slides/module-3-planning-measurement.md#m3-the-kpi-ladder-every-metric-connects-to-a-business-objective",
      "supporting_sources": [
        "../../../course/ai-native-media-operations/slides/module-5-reporting-insights.md#m5-kpis-vs-optimization-metrics-build-it-into-your-prompts"
      ],
      "used": {
        "quiz_completion_rate": 0.45,
        "new_uk_customer_revenue_monthly_usd": 60000,
        "unaided_awareness_rate": 0.15,
        "ltv_at_12_months_usd": 480
      },
      "boundary": "Targets are not measured outcomes. CTR/CPC/CPM are diagnostic levers; the paid-cohort export does not measure awareness, realized LTV, incrementality or full-month all-channel revenue."
    },
    {
      "id": "activation-handoff",
      "classification": "course_teaching_principle",
      "source": "../../../course/ai-native-media-operations/slides/module-4-activation.md#m4-greenleaf-pre-launch-checklist-paid-owned-and-earned",
      "used": "Carry measurement, naming, paid/owned/earned planning and quiz journey checks into reporting. The course states a $312 breakeven CAC guardrail.",
      "boundary": "No completed live pre-launch checklist is claimed. Media CPA is not full CAC; the guardrail needs reconciled cost/customer scope and economic assumptions."
    },
    {
      "id": "insight-and-audiences",
      "classification": "course_teaching_principle",
      "source": "../../../course/ai-native-media-operations/slides/module-5-reporting-insights.md#m5-every-insight-must-answer-three-questions",
      "supporting_sources": [
        "../../../course/ai-native-media-operations/slides/module-5-reporting-insights.md#m5-the-multi-audience-unlock",
        "../../../course/ai-native-media-operations/slides/module-5-reporting-insights.md#m5-step-2-ai-drafts-multi-audience-reports"
      ],
      "used": "WHAT / WHY / WHAT'S NEXT; one brief and dataset, one prompt, founder/CEO summary, media team analysis, content & creative brief. Each supports a different decision while preserving facts."
    },
    {
      "id": "brief-and-memory",
      "classification": "course_teaching_principle",
      "source": "../../../course/ai-native-media-operations/slides/module-5-reporting-insights.md#m5-step-1-continued-what-goes-in-the-reporting-brief",
      "supporting_sources": [
        "../../../course/ai-native-media-operations/slides/module-5-reporting-insights.md#m5-step-1-continued-the-brief-connects-to-everything-you-ve-built",
        "../../../course/ai-native-media-operations/slides/module-5-reporting-insights.md#m5-the-client-chat-as-a-compounding-product"
      ],
      "used": "Five brief components and upstream research/measurement/activation/institutional-memory handoffs. Record reviewed learnings in brief v2 without promoting hypotheses into facts."
    },
    {
      "id": "source-checks-and-review",
      "classification": "course_teaching_principle",
      "source": "../../../course/ai-native-media-operations/slides/module-5-reporting-insights.md#m5-the-data-trust-problem",
      "supporting_sources": [
        "../../../course/ai-native-media-operations/slides/module-5-reporting-insights.md#m5-dimensions-context-and-the-common-sense-check",
        "../../../course/ai-native-media-operations/slides/module-5-reporting-insights.md#m5-step-3-the-25-that-makes-reports-actually-useful",
        "../../../course/ai-native-media-operations/slides/module-5-reporting-insights.md#m5-step-4-from-reactive-to-proactive-reporting"
      ],
      "used": "Check source definitions and dimensions; human review evaluates decisions using context. Detect, analyze, route, review with human authority for released recommendations.",
      "boundary": "The human 25% label is a teaching metaphor, not measured time saved or superiority. The example is a static retrospective simulation, with no live alerts or external messages."
    },
    {
      "id": "mobile-teaching-signal",
      "classification": "course_illustrative_signal",
      "source": "../../../course/ai-native-media-operations/slides/module-5-reporting-insights.md#m5-greenleaf-week-2-ai-assisted-data-analysis",
      "governing_interpretation": "../../../../doc/2026-10-29-reporting-course-alignment.md",
      "used": "Reconstruct paid-cohort mobile quiz completion at 45% in Week 1 and 28% in Week 2, with starts as denominator.",
      "boundary": "No underlying CSV was recovered. The separate asset label 4.2% quiz CVR has no shared denominator definition here. Do not import an asset's $36K risk estimate, 3x segment result, Reddit-driven branded-search explanation or budget move as a fixture finding."
    },
    {
      "id": "template-structure",
      "classification": "course_reusable_framework",
      "source": "../../../course/ai-native-media-operations/templates/04-reporting-setup-guide.md",
      "used": "Brief fields, naming discipline and stakeholder-aware prompt structure.",
      "boundary": "Its NovaBrew case is a separate example and does not replace GreenLeaf. Paid template/slide assets are not exported in full."
    }
  ],
  "reconstruction_choices": {
    "classification": "newly_authored_synthetic_assumptions",
    "authority": "../../../../docs/superpowers/specs/2026-10-29-reporting-worked-example-design.md",
    "semantics": "../en/dictionary.md",
    "brief": "../en/brief-v1.md",
    "dates": { "week_1": ["2026-09-01", "2026-09-07"], "week_2": ["2026-09-08", "2026-09-14"], "snapshot": "2026-09-22T23:59:59Z", "timezone": "UTC", "date_meaning": "paid-click cohort date", "observation_and_attribution_days": 7 },
    "currency": "USD integer cents; first-purchase revenue after discounts, excluding tax/shipping; no fixture refunds",
    "grain": "date x channel x device x audience x creative; 336 canonical rows, 337 raw rows including one exact duplicate",
    "channels": ["google-search", "linkedin", "reddit-paid"],
    "devices": ["mobile", "desktop"],
    "audiences": ["skeptical", "exploring"],
    "creatives": ["traceability", "value"],
    "campaigns": { "google-search": "gl-uk-google-search", "linkedin": "gl-uk-linkedin", "reddit-paid": "gl-uk-reddit-paid" },
    "allowed_raw_campaign_alias": { "gl_uk_search": "gl-uk-google-search" },
    "attribution": "Each synthetic new customer and first-order revenue amount assigned once to last eligible paid click within seven days. Exclusive aggregate construction is not independent identity/attribution evidence.",
    "first_order_revenue_per_customer_cents": 8000,
    "first_order_revenue_boundary": "$80 is an authored constant for arithmetic clarity, not course pricing, realized LTV or a contribution-margin estimate.",
    "synthetic_patterns": "Daily costs/counts, creative/channel offsets, audience composition and within-stratum rates are constructed; publish exact generator choices with the fixture. One desktop zero-activity row tests undefined ratios. No causal or empirical finding follows from the construction.",
    "fixed_mix_rule": "Mobile channel x audience x creative strata, Week 1 quiz-start weights applied to both weeks; disclose undefined rates, excluded weights and counts. Check composition and within-group movement before diagnosis.",
    "roles_and_alert": "Fictional decision owners, weekly/monthly/quarterly cadence and mature paid-mobile below-45% routing rule are newly authored operational details, not live automation."
  },
  "supplemental_context": {
    "classification": "prewritten_fictional_update",
    "source": "../runs/greenleaf-2026-10-29/context-note.md",
    "release_rule": "Write and hash before initial AI run; introduce after initial response as newly learned context. Do not withhold known v1 history or retrofit to the model response. Update may leave the recommendation unchanged and cannot prove causality."
  },
  "derived_results": {
    "status": "No generated dataset, checked metric pack, actual AI response or human-review outcome is asserted by this manifest.",
    "permitted_pre_run_arithmetic": "45% to 28% is -17 percentage points, approximately -37.8% relative. All other fixture findings require generation, validation and independent calculation.",
    "record_later": "Actual data hashes, source cleanup locations, metric values, run metadata, unedited outputs and review decisions belong in their respective generated/run artifacts, not invented course facts."
  }
}