Put information where people expect it.

List the tasks and questions the website must support, then map them to existing or proposed pages. Note where several pages compete to answer the same question and where no page owns it. A page purpose should be specific enough to guide inclusion of material and distinguish neighbouring topics. Do not force unrelated information together solely because the same department supplies it. Equally, avoid splitting a short explanation across multiple pages when the reader needs the parts together to understand the answer.

Use language that helps visitors predict what they will find. A clever heading may be memorable while giving little clue about the service underneath it. Ask representative readers where they would look for a concrete piece of information and what they expect a label to contain. Listen for mismatches between internal terminology and the words people use in their own enquiries. Preserve those differences in the research notes rather than immediately teaching participants the organisation’s preferred vocabulary and calling the structure understood.

When new material arrives, decide whether it extends an established page, adds a distinct task or creates a new grouping. Update the relevant routes and remove obsolete distinctions instead of allowing the structure to reflect every historical publishing decision. Keep an inventory that identifies purpose and ownership, not just URLs. The notes explore page scope, naming and content relationships through practical website situations. A coherent architecture helps people locate an answer and recognise where they are without requiring a large navigation system or a prior explanation from staff.