Standards Library

    Product Operating Standards

    Reference operating documents that define expectations for documentation, governance, and evidence in mature product organizations. These are not templates—they are standards that establish what good looks like.

    Documentation

    Product Requirements Standard (Enterprise PRD)

    Defines the structure, content expectations, and review criteria for product requirements documentation in enterprise environments.

    Problem This Standard Solves

    Inconsistent requirements documentation leads to scope ambiguity, implementation errors, and failed procurement evaluations.

    What Good Looks Like

    Clear problem statements, measurable success criteria, explicit constraints, stakeholder sign-off evidence, and traceability to business objectives.

    When Required

    Before engineering commitment for features exceeding two sprints of effort, or for any customer-facing capability in regulated products.

    Owner

    Product Manager

    Reviewed By

    Product Leadership, Engineering Lead, relevant stakeholders

    Governance

    Decision Records (Product ADRs)

    Framework for capturing significant product decisions with context, alternatives considered, and rationale for the chosen path.

    Problem This Standard Solves

    Undocumented decisions create institutional knowledge loss, repeated debates, and inability to explain past choices to auditors or new team members.

    What Good Looks Like

    Timestamped records with clear decision statement, context, alternatives evaluated, trade-offs acknowledged, and decision-makers identified.

    When Required

    For any decision that affects architecture, introduces new dependencies, changes pricing or packaging, or establishes precedent.

    Owner

    Decision Sponsor (varies)

    Reviewed By

    Relevant stakeholders, documented in record

    Measurement

    KPI & Metric Definitions

    Standard format for defining, calculating, and reporting product metrics to ensure consistency and auditability.

    Problem This Standard Solves

    Inconsistent metric definitions lead to misaligned goals, unreliable reporting, and inability to compare performance across products or time periods.

    What Good Looks Like

    Precise calculation methodology, data sources identified, refresh frequency specified, ownership clear, and edge cases documented.

    When Required

    For any metric reported to leadership, included in customer contracts, or used for compensation decisions.

    Owner

    Product Analytics / Product Manager

    Reviewed By

    Data Engineering, Finance, relevant stakeholders

    Technical Governance

    Architecture Evidence Expectations

    Requirements for C1/C2 context diagrams, data flow documentation, and disaster recovery evidence in enterprise products.

    Problem This Standard Solves

    Missing architecture documentation delays security reviews, fails procurement evaluations, and creates operational risk during incidents.

    What Good Looks Like

    Current C1/C2 diagrams, documented data flows with classification, DR/BC procedures with tested RTOs, and regular review cadence.

    When Required

    For all production systems, updated quarterly or after significant changes, and before enterprise sales evaluations.

    Owner

    Engineering Lead / Architect

    Reviewed By

    Security, Platform Engineering, Product

    Compliance

    Risk & Compliance Evidence Checklist

    Comprehensive checklist of evidence requirements for regulated product environments and enterprise procurement responses.

    Problem This Standard Solves

    Incomplete compliance evidence delays sales cycles, triggers audit findings, and creates regulatory risk.

    What Good Looks Like

    Maintained evidence repository with clear ownership, regular attestation schedule, gap analysis, and remediation tracking.

    When Required

    Continuously maintained for regulated products, reviewed before major releases and procurement responses.

    Owner

    Compliance / GRC

    Reviewed By

    Legal, Security, Product Leadership

    AI Governance

    Model & Agent Evaluation Standard

    Defines expectations for evaluation sets, ground truth, guardrails, and post-deployment monitoring for any product feature powered by LLMs, models, or autonomous agents.

    Problem This Standard Solves

    Shipping AI capabilities without formal evaluation creates unmeasured risk, unpredictable regressions, and an inability to answer enterprise buyer or regulator questions about model behavior.

    What Good Looks Like

    Documented eval sets with ground truth, task-level success criteria, adversarial and safety cases, versioned prompt artifacts, drift monitoring, and defined thresholds for rollback or human escalation.

    When Required

    Before production release of any model-driven capability, and continuously maintained across model, prompt, or context changes.

    Owner

    Product Manager (with ML/AI Lead)

    Reviewed By

    AI Governance, Security, Legal, Product Leadership

    AI Governance

    Agent Autonomy & Boundaries Standard

    Establishes classification, permission scoping, and human-in-the-loop requirements for autonomous and semi-autonomous agents operating on behalf of users or the enterprise.

    Problem This Standard Solves

    Undocumented agent capabilities create audit exposure, unclear liability, and unpredictable behavior in customer-facing or regulated workflows.

    What Good Looks Like

    Autonomy tier classification, explicit tool and data scopes, reversibility analysis, escalation criteria, kill-switch procedures, and observable action logs suitable for audit review.

    When Required

    For any agent capable of taking action beyond retrieval — including tool use, transactions, communications, or data modification.

    Owner

    Product Manager (with AI Lead)

    Reviewed By

    Security, Legal, Compliance, Executive Sponsor

    Commercial Governance

    Pricing & Packaging Change Standard

    Defines the evidence, approval chain, and communication requirements for any change to list pricing, packaging tiers, or entitlement structures in enterprise offerings.

    Problem This Standard Solves

    Ad hoc pricing and packaging changes create revenue leakage, contract renegotiation exposure, and customer trust erosion in enterprise segments where price stability is a procurement expectation.

    What Good Looks Like

    Written change proposal with segment-level revenue modeling, customer impact analysis, grandfathering policy, sales enablement plan, and executive sign-off recorded as an ADR.

    When Required

    For any change to list pricing, tier boundaries, entitlement rules, or usage metering that affects customer-visible contracts or renewal economics.

    Owner

    Product Manager (with Finance)

    Reviewed By

    CFO, CRO, Legal, Executive Sponsor

    Lifecycle Governance

    Feature & API Deprecation Standard

    Governs the retirement of customer-visible features, endpoints, and integrations with adequate notice, migration support, and evidence preservation.

    Problem This Standard Solves

    Silent deprecations damage enterprise trust, trigger contract disputes, and expose the organization to procurement claims about undocumented product changes.

    What Good Looks Like

    Deprecation ADR, minimum notice period appropriate to customer segment, published sunset date, migration guidance with named support owner, and retained artifact of the pre-deprecation behavior.

    When Required

    For any planned removal, replacement, or breaking change to a customer-visible feature, API surface, or integration contract.

    Owner

    Product Manager (with Engineering Lead)

    Reviewed By

    Customer Success, Legal, Enterprise Account Executives

    AI Governance

    AI Vendor Evaluation Standard

    Establishes the evaluation, due diligence, and ongoing monitoring requirements for third-party AI vendors and model providers integrated into the product.

    Problem This Standard Solves

    Undocumented AI vendor selection creates supply-chain risk, undisclosed sub-processor exposure, and evaluation gaps that surface only under enterprise procurement scrutiny.

    What Good Looks Like

    Written vendor evaluation record covering model provenance, training data disclosures, sub-processor chain, data residency, evaluation on our benchmark set, and a documented exit plan with a named alternative.

    When Required

    Before contracting with any AI vendor or model provider whose output is exposed to customers, and reviewed annually or on material capability change.

    Owner

    Product Manager (with Security and Procurement)

    Reviewed By

    Security, Legal, Compliance, Data Governance

    Enterprise Product Standards Pack

    Download all five standards as a complete operating reference for your product organization.

    Use the advisor