Self-initiated exploration. Not commissioned or deployed.
Publishing as a system
The institutional question: how does an institution keep a thousand pages current when nobody's job is "the website"?
Figure ratio — 4:3
The full study
Why it is difficult
Ownership is distributed across hundreds of publishers; content mirrors the org chart; review happens only when something breaks publicly.
Constraints & stakeholders
Registrar, departments, HR systems and communications each hold authoritative fragments; accuracy carries legal and reputational weight.
Hypothesis
Structure outperforms heroics: owners and review cycles attached to content types — not pages — keep a large site current without a heroic web team.
Approach
A content model with governance fields on every type; pages become surfaces composed from one source, so a fact updates everywhere at once.
Decisions & tradeoffs
Governance states visible to editors, invisible to visitors; the study holds itself to three types to keep the model legible.
What it demonstrates
Content modeling, governance design and structured data — the machinery beneath accessible, findable institutional publishing.
What we would measure
Proposed, not achieved: authoring effort per update; share of pages past their review date; duplicate-content rate across units.
To test with a real institution
Whether the workflow fits a real CMS and real editors' capacity — governance that ignores workload fails quietly.