Information Architecture.

Information architecture consists of four building blocks. The organisation system determines how content is grouped: by service, by target group, by industry or by process. The labelling system determines what these groups are called; this is where most problems arise, because internal terms are rarely the words searchers use. The navigation system determines how people move around: main navigation, footer, cross-references, breadcrumb trail. The search system determines what happens when someone bypasses the structure and searches directly. The discipline was shaped largely by Louis Rosenfeld and Peter Morville and is older than the term UX design.

Information architecture is tested with people, not with opinions in a meeting room. In card sorting, participants sort content cards into groups: open when the groups still have to be found, closed when an existing structure is being checked. In tree testing, participants see only the menu structure without any design and have to complete a task in it; this shows cleanly whether the structure or the design is the problem. The site's internal search and the queries in Google Search Console add the words people actually use.

A concrete case: over the years a service provider has accumulated around forty pages. Its services sit partly under 'Offer' and partly under 'About us', and prices are in a PDF in the download area. A tree test with ten people and the task 'Find out what an initial consultation costs' ends up in the wrong branch in six cases. Restructuring into five clearly named main sections, a dedicated pricing page and cross-links to it from every service page not only reduces drop-offs but makes the pricing information indexable in the first place; as a PDF in the download area it was effectively invisible.

For search engines, information architecture is not a side effect but the foundation. It determines the URL structure, internal linking and the click depth of important pages. A flat, thematically clean structure also prevents keyword cannibalisation, because every topic has exactly one responsible page. The user flow shows how the structure turns into a concrete sequence. To be honest about the limits: information architecture does not repair weak content. If the service page does not answer the questions about approach and price, the visitor finds it faster but still leaves. How we approach structure and interface together is described on the page UX Agency Zurich.