{
  "acceptance": "The comparison should reproduce the selected first and last interaction for a known journey, use one conversion definition and disclose any tracking gaps. A change in assigned channel credit should not be presented as a change in observed conversions.",
  "automated_checks": true,
  "batch_id": "analytics-editorial-20261104-v1",
  "built_at": "2026-11-04T11:50:28.566595+00:00",
  "claims": [
    {
      "id": "c1",
      "section": "Explanation and verification steps",
      "source_numbers": [
        1
      ],
      "statement": "Umami first-click attribution gives full credit to the first interaction."
    },
    {
      "id": "c2",
      "section": "Explanation and verification steps",
      "source_numbers": [
        1
      ],
      "statement": "Last-click attribution gives full credit to the final interaction before conversion."
    },
    {
      "id": "c3",
      "section": "Explanation and verification steps",
      "source_numbers": [
        1
      ],
      "statement": "A conversion step is required for the attribution insight."
    }
  ],
  "description": "Hold conversion definition and date range fixed before interpreting different channel credit.",
  "diagram": [
    "Fix the conversion definition",
    "Order observed interactions",
    "Compare first and last credit",
    "State the model in the report"
  ],
  "draft_hash": "7427c0aa403d5cd774a6af1f675a03d8c86439cf5858d8de49b9067cdc431e39",
  "evidence": {
    "observed_at": "2026-11-03T21:15:50.611897+00:00",
    "period": {
      "end": "2026-11-03T21:15:50.611897+00:00",
      "start": "2026-11-03T21:15:50.611897+00:00",
      "timezone": "UTC"
    },
    "rows": [],
    "scope": "Official documentation; no customer or own-site traffic measurement"
  },
  "evidence_kind": "documentation",
  "figure_caption": "Original explanatory diagram based on the dated sources below. Any example values are hypothetical; this is not a screenshot, customer measurement or AI platform result.",
  "figure_kind": "comparison",
  "figure_notes": [
    "Keep the conversion step and date range unchanged.",
    "Retain the actual sequence rather than only the winning channel.",
    "Different credit models can select different interactions.",
    "Assigned credit is not proof of causal contribution."
  ],
  "figure_title": "First-click vs last-click attribution: compare the same conversion",
  "group": "Measurement and interpretation",
  "human_review_claim": false,
  "intent": "first-last-click-attribution-comparison",
  "interpretation": "An attribution model assigns credit; it does not change the historical order of interactions. Umami’s documentation distinguishes first-interaction credit from final-interaction credit and requires a conversion step. Before calling two dashboards inconsistent, check whether they answer the same credit question and use the same conversion definition. This is useful when comparing an AI referral with earlier discovery or a later return visit.",
  "keyword_volume": "Not obtained",
  "limits": "Documentation defines the model; it does not verify your journey coverage. Cross-device gaps or missing referral evidence must remain unresolved unless separately established.",
  "paid_api_calls": 0,
  "path": "/ai-analytics/guides/first-last-click-attribution-comparison/",
  "policy": "session-evidence-editorial-v1",
  "preparation_mode": "session_editorial_batch",
  "problem": "Two attribution views assign the same conversion to different channels.",
  "product": "/ai-analytics",
  "published_at": "2026-11-04T12:15:11.453421+00:00",
  "reader": "Site owners and analysts investigating two attribution views assign the same conversion to different channels.",
  "related": [
    "missing-ai-referral-data"
  ],
  "related_links": [
    {
      "path": "/ai-analytics/guides/missing-ai-referral-data/",
      "title": "Handle missing AI referrers without inventing attribution"
    }
  ],
  "research_document_ids": [
    "5d7955810091b5e523407b647452935a"
  ],
  "slug": "first-last-click-attribution-comparison",
  "source_signature": "b705722b2ad3eea2c640b21e7416d06215bbde8891a375fad4a2b0de27aab787",
  "sources": [
    {
      "hash": "9cdb05c7a30aa1797d988fd8e750b91a0b3794872b4f41720f8cdd57b459ddb2",
      "locator": "Models # / b4; Models # / b5; Parameters # / b8",
      "observed_at": "2026-11-03T21:15:50.611897+00:00",
      "title": "Attribution",
      "url": "https://docs.umami.is/docs/attribution"
    }
  ],
  "sources_required": [],
  "steps": [
    "Write down the exact conversion step: a specified page view or event. Record the date range, filters and attribution model beside the result.",
    "Construct a permitted test journey with distinguishable source labels and a known sequence; keep it out of customer performance reporting.",
    "Run both model views with the same conversion step and scope. Identify which interaction each model selects before comparing totals.",
    "When reporting channel performance, name the credit model and note missing journey coverage. Keep causal claims separate from model-assigned credit."
  ],
  "title": "First-click vs last-click attribution: compare the same conversion",
  "worked_example": "Hypothetical journey: a visitor first arrives through a search campaign, later follows an AI referral, then returns through an email link and submits a request. First-click and last-click models give credit to different endpoints of that sequence. Neither model, on its own, proves that the middle interaction was useless or that any channel caused the conversion."
}