{
  "schema_version": 1,
  "slug": "ai-marketing-workflows",
  "title": "AI marketing workflows",
  "purpose": "Prepare useful marketing artifacts with bounded prompts and explicit human review.",
  "intro": "Each entry includes a full input worksheet, three concrete prompts, a manually authored fictional example and task-specific review criteria. The workbooks are downloadable and can be used independently of the website.",
  "updated_on": "2026-10-08",
  "license": "CC-BY-4.0",
  "attribution": "Jon Bishop",
  "selection_note": "A selected library for SaaS, AI and developer-tool founders. Source-backed observations are separate from our advice; inclusion is not a ranking or a tested performance claim.",
  "categories": {
    "research": {
      "label": "Research and synthesis",
      "intro": "Use a bounded prompt sequence with explicit inputs and review criteria.",
      "compare": "The model helps prepare an artifact; a responsible person verifies evidence and decides what happens next."
    },
    "content": {
      "label": "Content and messaging",
      "intro": "Use a bounded prompt sequence with explicit inputs and review criteria.",
      "compare": "The model helps prepare an artifact; a responsible person verifies evidence and decides what happens next."
    },
    "measurement": {
      "label": "Measurement and experiments",
      "intro": "Use a bounded prompt sequence with explicit inputs and review criteria.",
      "compare": "The model helps prepare an artifact; a responsible person verifies evidence and decides what happens next."
    },
    "operations": {
      "label": "Planning and operations",
      "intro": "Use a bounded prompt sequence with explicit inputs and review criteria.",
      "compare": "The model helps prepare an artifact; a responsible person verifies evidence and decides what happens next."
    }
  },
  "filter_fields": {
    "execution": "Execution",
    "format": "Format"
  },
  "attribute_labels": {
    "open_source": "Open source",
    "self_hosted": "Self-hosted",
    "pricing": "Pricing model",
    "integration": "Verified connection",
    "subcategory": "Subcategory",
    "license": "License scope",
    "format": "Format",
    "provider": "Provider",
    "task": "Task",
    "artifact": "Page type",
    "pattern": "Pattern",
    "metric": "Metric",
    "period": "Data period",
    "cohort": "Cohort",
    "execution": "Execution",
    "output": "Output",
    "stage": "Use at this point"
  },
  "records": [
    {
      "slug": "homepage-draft",
      "title": "Source-grounded homepage draft",
      "category": "content",
      "summary": "A complete three-prompt workbook for a claim-evidence ledger followed by a homepage draft.",
      "use_when": "You have the stated inputs and want a reviewable draft with evidence boundaries.",
      "limitations": "This is a manually authored prompt workbook, not a model-tested automation or an executed integration. Output quality and cost depend on the chosen model and inputs.",
      "sources": [
        {
          "url": "/directories/ai-marketing-workflows/downloads/homepage-draft.md",
          "title": "Jon Bishop original workbook",
          "scope": "Original resource inspected locally; examples are fictional",
          "checked_on": "2026-10-08",
          "reference_id": "original:ai-marketing-workflows/homepage-draft.md"
        }
      ],
      "sections": [
        {
          "heading": "Inputs and output",
          "paragraphs": [
            "Product facts with source IDs, target visitor, desired action, supported proof and brand constraints",
            "A claim-evidence ledger followed by a homepage draft"
          ]
        },
        {
          "heading": "Prompt sequence",
          "paragraphs": [
            "Inspect evidence and missing inputs; prepare the artifact using supported facts; review and correct the result against task-specific checks."
          ]
        },
        {
          "heading": "Worked example \u2014 fictional",
          "paragraphs": [
            "A fictional API monitor detects failed requests; input evidence supports alerts but provides no quantified savings.",
            "Draft: Detect failed API requests and investigate the trace. Omit any percentage savings claim."
          ]
        },
        {
          "heading": "Review criterion",
          "paragraphs": [
            "Every capability and outcome maps to supplied evidence; the CTA describes a real available route."
          ]
        }
      ],
      "steps": [
        "Fill the input worksheet and remove sensitive material that your model cannot receive.",
        "Run the prompts in sequence in an approved model, carrying forward the evidence ledger.",
        "Check the resulting artifact independently and keep unresolved questions visible."
      ],
      "attributes": {
        "execution": [
          "Manual prompt sequence; model untested"
        ],
        "format": [
          "Markdown prompt workbook"
        ],
        "output": [
          "A claim-evidence ledger followed by a homepage draft"
        ]
      },
      "status": "Complete original workbook; model execution untested",
      "reviewed_on": "2026-10-08",
      "resource_filename": "homepage-draft.md"
    },
    {
      "slug": "interview-synthesis",
      "title": "Customer interview synthesis",
      "category": "research",
      "summary": "A complete three-prompt workbook for a theme table preserving quotes, contrary cases and unanswered questions.",
      "use_when": "You have the stated inputs and want a reviewable draft with evidence boundaries.",
      "limitations": "This is a manually authored prompt workbook, not a model-tested automation or an executed integration. Output quality and cost depend on the chosen model and inputs.",
      "sources": [
        {
          "url": "/directories/ai-marketing-workflows/downloads/interview-synthesis.md",
          "title": "Jon Bishop original workbook",
          "scope": "Original resource inspected locally; examples are fictional",
          "checked_on": "2026-10-08",
          "reference_id": "original:ai-marketing-workflows/interview-synthesis.md"
        }
      ],
      "sections": [
        {
          "heading": "Inputs and output",
          "paragraphs": [
            "Sanitized transcripts with interview IDs, dates, recruitment context and the research question",
            "A theme table preserving quotes, contrary cases and unanswered questions"
          ]
        },
        {
          "heading": "Prompt sequence",
          "paragraphs": [
            "Inspect evidence and missing inputs; prepare the artifact using supported facts; review and correct the result against task-specific checks."
          ]
        },
        {
          "heading": "Worked example \u2014 fictional",
          "paragraphs": [
            "Fictional: two of three interviews mention procurement review; one self-serve buyer does not.",
            "Output: procurement appeared in two supplied interviews; population prevalence is unknown. Preserve the self-serve counterexample."
          ]
        },
        {
          "heading": "Review criterion",
          "paragraphs": [
            "Quotes are exact, respondent counts are accurate and sample observations are not generalized to the whole market."
          ]
        }
      ],
      "steps": [
        "Fill the input worksheet and remove sensitive material that your model cannot receive.",
        "Run the prompts in sequence in an approved model, carrying forward the evidence ledger.",
        "Check the resulting artifact independently and keep unresolved questions visible."
      ],
      "attributes": {
        "execution": [
          "Manual prompt sequence; model untested"
        ],
        "format": [
          "Markdown prompt workbook"
        ],
        "output": [
          "A theme table preserving quotes, contrary cases and unanswered questions"
        ]
      },
      "status": "Complete original workbook; model execution untested",
      "reviewed_on": "2026-10-08",
      "resource_filename": "interview-synthesis.md"
    },
    {
      "slug": "onboarding-emails",
      "title": "Lifecycle onboarding sequence",
      "category": "content",
      "summary": "A complete three-prompt workbook for a trigger-and-message plan with draft emails.",
      "use_when": "You have the stated inputs and want a reviewable draft with evidence boundaries.",
      "limitations": "This is a manually authored prompt workbook, not a model-tested automation or an executed integration. Output quality and cost depend on the chosen model and inputs.",
      "sources": [
        {
          "url": "/directories/ai-marketing-workflows/downloads/onboarding-emails.md",
          "title": "Jon Bishop original workbook",
          "scope": "Original resource inspected locally; examples are fictional",
          "checked_on": "2026-10-08",
          "reference_id": "original:ai-marketing-workflows/onboarding-emails.md"
        }
      ],
      "sections": [
        {
          "heading": "Inputs and output",
          "paragraphs": [
            "Verified activation event, product paths, consent and suppression rules, eligibility, tone and approved product claims",
            "A trigger-and-message plan with draft emails"
          ]
        },
        {
          "heading": "Prompt sequence",
          "paragraphs": [
            "Inspect evidence and missing inputs; prepare the artifact using supported facts; review and correct the result against task-specific checks."
          ]
        },
        {
          "heading": "Worked example \u2014 fictional",
          "paragraphs": [
            "Fictional: an account created a project but has not sent its first test event.",
            "Draft one reminder linking the verified test-event guide; suppress it after the first valid event."
          ]
        },
        {
          "heading": "Review criterion",
          "paragraphs": [
            "Triggers are observable, destinations work, opt-out ownership is explicit and no email is sent by this workflow."
          ]
        }
      ],
      "steps": [
        "Fill the input worksheet and remove sensitive material that your model cannot receive.",
        "Run the prompts in sequence in an approved model, carrying forward the evidence ledger.",
        "Check the resulting artifact independently and keep unresolved questions visible."
      ],
      "attributes": {
        "execution": [
          "Manual prompt sequence; model untested"
        ],
        "format": [
          "Markdown prompt workbook"
        ],
        "output": [
          "A trigger-and-message plan with draft emails"
        ]
      },
      "status": "Complete original workbook; model execution untested",
      "reviewed_on": "2026-10-08",
      "resource_filename": "onboarding-emails.md"
    },
    {
      "slug": "technical-tutorial",
      "title": "Technical tutorial brief",
      "category": "content",
      "summary": "A complete three-prompt workbook for an evidence-linked tutorial outline and technical verification checklist.",
      "use_when": "You have the stated inputs and want a reviewable draft with evidence boundaries.",
      "limitations": "This is a manually authored prompt workbook, not a model-tested automation or an executed integration. Output quality and cost depend on the chosen model and inputs.",
      "sources": [
        {
          "url": "/directories/ai-marketing-workflows/downloads/technical-tutorial.md",
          "title": "Jon Bishop original workbook",
          "scope": "Original resource inspected locally; examples are fictional",
          "checked_on": "2026-10-08",
          "reference_id": "original:ai-marketing-workflows/technical-tutorial.md"
        }
      ],
      "sections": [
        {
          "heading": "Inputs and output",
          "paragraphs": [
            "Current official documentation excerpts, versions, reproduced commands, audience prerequisites and the desired working result",
            "An evidence-linked tutorial outline and technical verification checklist"
          ]
        },
        {
          "heading": "Prompt sequence",
          "paragraphs": [
            "Inspect evidence and missing inputs; prepare the artifact using supported facts; review and correct the result against task-specific checks."
          ]
        },
        {
          "heading": "Worked example \u2014 fictional",
          "paragraphs": [
            "Fictional: a local sample endpoint accepts a test payload and returns a request ID.",
            "Outline setup, the exact tested request and its observed response; mark deployment and scale behavior as unverified."
          ]
        },
        {
          "heading": "Review criterion",
          "paragraphs": [
            "Commands match the tested version; no invented API or claimed reproduction replaces an actual local run."
          ]
        }
      ],
      "steps": [
        "Fill the input worksheet and remove sensitive material that your model cannot receive.",
        "Run the prompts in sequence in an approved model, carrying forward the evidence ledger.",
        "Check the resulting artifact independently and keep unresolved questions visible."
      ],
      "attributes": {
        "execution": [
          "Manual prompt sequence; model untested"
        ],
        "format": [
          "Markdown prompt workbook"
        ],
        "output": [
          "An evidence-linked tutorial outline and technical verification checklist"
        ]
      },
      "status": "Complete original workbook; model execution untested",
      "reviewed_on": "2026-10-08",
      "resource_filename": "technical-tutorial.md"
    },
    {
      "slug": "experiment-plan",
      "title": "Marketing experiment design",
      "category": "measurement",
      "summary": "A complete three-prompt workbook for a proposed design with assumptions and missing statistical inputs.",
      "use_when": "You have the stated inputs and want a reviewable draft with evidence boundaries.",
      "limitations": "This is a manually authored prompt workbook, not a model-tested automation or an executed integration. Output quality and cost depend on the chosen model and inputs.",
      "sources": [
        {
          "url": "/directories/ai-marketing-workflows/downloads/experiment-plan.md",
          "title": "Jon Bishop original workbook",
          "scope": "Original resource inspected locally; examples are fictional",
          "checked_on": "2026-10-08",
          "reference_id": "original:ai-marketing-workflows/experiment-plan.md"
        }
      ],
      "sections": [
        {
          "heading": "Inputs and output",
          "paragraphs": [
            "Decision, assignment unit, primary metric definition, baseline data, eligibility, constraints and practical effect threshold",
            "A proposed design with assumptions and missing statistical inputs"
          ]
        },
        {
          "heading": "Prompt sequence",
          "paragraphs": [
            "Inspect evidence and missing inputs; prepare the artifact using supported facts; review and correct the result against task-specific checks."
          ]
        },
        {
          "heading": "Worked example \u2014 fictional",
          "paragraphs": [
            "Fictional: ten weekly eligible trials make a quick multi-variant test unlikely to be informative.",
            "Recommend checking instrumentation and traffic feasibility before choosing sample size; do not invent a statistically adequate duration."
          ]
        },
        {
          "heading": "Review criterion",
          "paragraphs": [
            "Assignment, denominator, contamination and stopping rules are explicit; qualified review is needed for inferential claims."
          ]
        }
      ],
      "steps": [
        "Fill the input worksheet and remove sensitive material that your model cannot receive.",
        "Run the prompts in sequence in an approved model, carrying forward the evidence ledger.",
        "Check the resulting artifact independently and keep unresolved questions visible."
      ],
      "attributes": {
        "execution": [
          "Manual prompt sequence; model untested"
        ],
        "format": [
          "Markdown prompt workbook"
        ],
        "output": [
          "A proposed design with assumptions and missing statistical inputs"
        ]
      },
      "status": "Complete original workbook; model execution untested",
      "reviewed_on": "2026-10-08",
      "resource_filename": "experiment-plan.md"
    },
    {
      "slug": "competitor-evidence",
      "title": "Competitor evidence comparison",
      "category": "research",
      "summary": "A complete three-prompt workbook for a scoped comparison matrix with unknown cells.",
      "use_when": "You have the stated inputs and want a reviewable draft with evidence boundaries.",
      "limitations": "This is a manually authored prompt workbook, not a model-tested automation or an executed integration. Output quality and cost depend on the chosen model and inputs.",
      "sources": [
        {
          "url": "/directories/ai-marketing-workflows/downloads/competitor-evidence.md",
          "title": "Jon Bishop original workbook",
          "scope": "Original resource inspected locally; examples are fictional",
          "checked_on": "2026-10-08",
          "reference_id": "original:ai-marketing-workflows/competitor-evidence.md"
        }
      ],
      "sections": [
        {
          "heading": "Inputs and output",
          "paragraphs": [
            "Dated primary-page excerpts with source IDs, the buyer question and equivalent product editions",
            "A scoped comparison matrix with unknown cells"
          ]
        },
        {
          "heading": "Prompt sequence",
          "paragraphs": [
            "Inspect evidence and missing inputs; prepare the artifact using supported facts; review and correct the result against task-specific checks."
          ]
        },
        {
          "heading": "Worked example \u2014 fictional",
          "paragraphs": [
            "Fictional: competitor A documents SSO on Enterprise; competitor B does not mention SSO on the supplied page.",
            "A: Enterprise SSO documented. B: unknown from supplied evidence, rather than absent."
          ]
        },
        {
          "heading": "Review criterion",
          "paragraphs": [
            "Absence of a mention stays unknown; dates and edition scopes travel with each claim."
          ]
        }
      ],
      "steps": [
        "Fill the input worksheet and remove sensitive material that your model cannot receive.",
        "Run the prompts in sequence in an approved model, carrying forward the evidence ledger.",
        "Check the resulting artifact independently and keep unresolved questions visible."
      ],
      "attributes": {
        "execution": [
          "Manual prompt sequence; model untested"
        ],
        "format": [
          "Markdown prompt workbook"
        ],
        "output": [
          "A scoped comparison matrix with unknown cells"
        ]
      },
      "status": "Complete original workbook; model execution untested",
      "reviewed_on": "2026-10-08",
      "resource_filename": "competitor-evidence.md"
    },
    {
      "slug": "performance-review",
      "title": "Monthly marketing performance review",
      "category": "measurement",
      "summary": "A complete three-prompt workbook for a reconciled metric table and investigation memo.",
      "use_when": "You have the stated inputs and want a reviewable draft with evidence boundaries.",
      "limitations": "This is a manually authored prompt workbook, not a model-tested automation or an executed integration. Output quality and cost depend on the chosen model and inputs.",
      "sources": [
        {
          "url": "/directories/ai-marketing-workflows/downloads/performance-review.md",
          "title": "Jon Bishop original workbook",
          "scope": "Original resource inspected locally; examples are fictional",
          "checked_on": "2026-10-08",
          "reference_id": "original:ai-marketing-workflows/performance-review.md"
        }
      ],
      "sections": [
        {
          "heading": "Inputs and output",
          "paragraphs": [
            "Metric definitions, sanitized aggregate data, periods, spend, cohort rules, tracking changes and business questions",
            "A reconciled metric table and investigation memo"
          ]
        },
        {
          "heading": "Prompt sequence",
          "paragraphs": [
            "Inspect evidence and missing inputs; prepare the artifact using supported facts; review and correct the result against task-specific checks."
          ]
        },
        {
          "heading": "Worked example \u2014 fictional",
          "paragraphs": [
            "Fictional: 20 of 100 eligible trials activated this month; last month used a different activation definition.",
            "Report 20% for the current definition and flag the historical comparison as inconsistent."
          ]
        },
        {
          "heading": "Review criterion",
          "paragraphs": [
            "Arithmetic and denominators are checked independently; attribution is separated from causal explanations."
          ]
        }
      ],
      "steps": [
        "Fill the input worksheet and remove sensitive material that your model cannot receive.",
        "Run the prompts in sequence in an approved model, carrying forward the evidence ledger.",
        "Check the resulting artifact independently and keep unresolved questions visible."
      ],
      "attributes": {
        "execution": [
          "Manual prompt sequence; model untested"
        ],
        "format": [
          "Markdown prompt workbook"
        ],
        "output": [
          "A reconciled metric table and investigation memo"
        ]
      },
      "status": "Complete original workbook; model execution untested",
      "reviewed_on": "2026-10-08",
      "resource_filename": "performance-review.md"
    },
    {
      "slug": "launch-plan",
      "title": "Evidence-based launch plan",
      "category": "operations",
      "summary": "A complete three-prompt workbook for a launch checklist with dependencies and decision gates.",
      "use_when": "You have the stated inputs and want a reviewable draft with evidence boundaries.",
      "limitations": "This is a manually authored prompt workbook, not a model-tested automation or an executed integration. Output quality and cost depend on the chosen model and inputs.",
      "sources": [
        {
          "url": "/directories/ai-marketing-workflows/downloads/launch-plan.md",
          "title": "Jon Bishop original workbook",
          "scope": "Original resource inspected locally; examples are fictional",
          "checked_on": "2026-10-08",
          "reference_id": "original:ai-marketing-workflows/launch-plan.md"
        }
      ],
      "sections": [
        {
          "heading": "Inputs and output",
          "paragraphs": [
            "Target audience, release scope, verified readiness checks, approved channels, dependencies, budget and owners",
            "A launch checklist with dependencies and decision gates"
          ]
        },
        {
          "heading": "Prompt sequence",
          "paragraphs": [
            "Inspect evidence and missing inputs; prepare the artifact using supported facts; review and correct the result against task-specific checks."
          ]
        },
        {
          "heading": "Worked example \u2014 fictional",
          "paragraphs": [
            "Fictional: the signup route works, but the tutorial still contains an untested installation command.",
            "Keep the tutorial verification as an unresolved gate; draft the channel plan without claiming launch readiness."
          ]
        },
        {
          "heading": "Review criterion",
          "paragraphs": [
            "Every task has an owner, dependencies are visible and external posting or spending requires its own authorization."
          ]
        }
      ],
      "steps": [
        "Fill the input worksheet and remove sensitive material that your model cannot receive.",
        "Run the prompts in sequence in an approved model, carrying forward the evidence ledger.",
        "Check the resulting artifact independently and keep unresolved questions visible."
      ],
      "attributes": {
        "execution": [
          "Manual prompt sequence; model untested"
        ],
        "format": [
          "Markdown prompt workbook"
        ],
        "output": [
          "A launch checklist with dependencies and decision gates"
        ]
      },
      "status": "Complete original workbook; model execution untested",
      "reviewed_on": "2026-10-08",
      "resource_filename": "launch-plan.md"
    }
  ],
  "collections": [],
  "methodology": [
    {
      "heading": "Original resources",
      "paragraphs": [
        "These workbooks were authored for this library. Fictional examples illustrate expected evidence boundaries; they are not customer outcomes or actual model runs."
      ]
    },
    {
      "heading": "Execution status",
      "paragraphs": [
        "The downloadable text and links were checked locally. Models, integrations and external actions were not executed. A workflow is published as a prompt workbook, not as a proven automation."
      ]
    },
    {
      "heading": "Review",
      "paragraphs": [
        "The person using the workflow verifies claims, calculations, commands, permissions and unresolved evidence before acting."
      ]
    }
  ]
}