1. What is the minimum viable operating model to get started?

    Start with a small central enablement team that owns standards, tooling, and support, plus 1–2 pilot domains with clear business pressure and committed stakeholders.

  2. Who should own data products: central team or domains?

    Use a hybrid model. Central owns what must be consistent (standards, platform capabilities, enablement, portfolio transparency). Domains own the data products that support their business outcomes.

  3. How do we pick the first domains?

    Pick 1–3 domains where you already have strong relationships, urgent use cases, and leaders willing to invest. Start small, prove the model, then expand.

  4. What roles do we need in a domain team?

    A minimal viable setup is a data product manager, a data analyst, and a data engineer. In very small setups, roles can be combined temporarily, but accountability must stay clear.

  5. What decision bodies do we actually need?

    Use domain-specific councils for focused prioritization and trade-offs. Keep one federated council (or similar forum) to align cross-domain standards and platform needs.

  6. How do we avoid the “part-time steward” trap?

    Do not rely on involuntary volunteers to run the operating model. Use professionals for core responsibilities. Use stewards as tactical contributors with low overhead and clear scope.

  7. How do we scale without bureaucracy?

    Standardize the basics, keep decision forums small and focused, and measure outcomes (adoption, time-to-value, product health). If a process does not improve outcomes, remove it.