What is information architecture?

A short definition

Every website has an information architecture, whether anyone planned it or not. It is the answer to four questions:

1. Grouping. Which pieces of content belong together?
2. Labelling. What is each group, and each page, called?
3. Addressing. Where does each page live, and what is its URL?
4. Connecting. Which pages link to which, and how do people move between them?

When those answers match how your visitors think, the site feels obvious. When they match how your organisation is arranged internally, or how content happened to be added over the years, visitors have to learn your logic before they can find anything.

The parts of an information architecture

The hierarchy

The hierarchy is the tree of sections and pages: home at the top, main sections below, then subsections and pages. It is the backbone of the structure and the part people picture when they hear the word "sitemap". The Australian Government Style Manual describes hierarchies as sorting content into categories and levels, which helps users see how topics relate, and notes that website information architecture can be built on them. Its advice to keep hierarchies shallow applies to sites as well as documents: each level is another choice a visitor has to get right.

Labels

Labels are the words on menus, section headings, page titles and buttons. They are the main way people decide where to click. Good labels use the words your visitors use, are specific enough to rule things out, and stay consistent across the site. A label like "Resources" tells a visitor almost nothing about what they will find; "Forms and templates" tells them a lot.

Navigation

Navigation is how the hierarchy shows up on the page: the main menu, local menus within sections, breadcrumbs, footers and utility links. It is covered in detail in the guides on accessible navigation patterns and breadcrumbs and wayfinding. Two WCAG 2.2 success criteria apply directly. Consistent Navigation (3.2.3, level AA) expects repeated menus to keep their order across pages. Multiple Ways (2.4.5, level AA) expects at least two routes to each page.

Taxonomy

A taxonomy is the controlled set of labels used to classify content: categories, tags, topics, audiences, locations. It powers listings, filters and related-content blocks. Where the hierarchy gives each page one home, a taxonomy lets content be found from several angles. The taxonomy guide explains how to design one.

Page addresses

Every page has a URL, and the pattern of those URLs is part of the structure. Readable addresses help people judge a link before clicking it and help search engines understand the page. The URL structure guide covers the choices.

Internal links

Links between pages are the paths people and search engines use to move around. Google's documentation advises linking each important page from "at least one other page on your site". A structure without deliberate links leaves good pages stranded. See internal linking for findability.

Search

Site search is a second way in for people who know what they want or who cannot find it in the menus. It depends on the same labels and taxonomy as everything else: search results filtered by a messy taxonomy are messy too.

Where information architecture fits in a project

Information architecture sits between research and design. Research tells you who your visitors are and what they come to do. Information architecture turns that into a structure. Design then gives the structure a visual form, and development builds it.

That order matters. If design starts first, the structure gets decided by default, which can mean copying the old site's menu into the new look. Settling the structure first gives designers a clear brief: these sections, these labels, these page types, these navigation elements.

How information architecture is done

The methods are practical and mostly low-tech.

• Content inventory. A list of every page, built by crawling the site or exporting from the content system.
• Task analysis. A list of what visitors come to do, drawn from search logs, support enquiries, analytics and conversations with staff.
• Card sorting. Participants group content into categories that make sense to them. Digital NSW describes it as people "organising topics into categories that make sense to them", and suggests at least five to six participants.
• Tree testing. Participants try to find answers to tasks using only a text version of a proposed structure. The European Commission's web guide describes it as evaluating the information structure with tasks and scenarios from user research, to check that the "structure corresponds to users' logic".
• Sitemaps and diagrams. The documented structure, shared with everyone who builds and writes for the site.

Our guide to card sorting and tree testing explains the two research methods in detail.

What information architecture can and cannot do

A sound structure makes content easier to find and understand. It reduces duplicated pages, keeps menus manageable as the site grows and gives writers a clear place to put new content. Google's SEO Starter Guide notes that a logical structure "can help search engines and users understand how your pages relate to the rest of your site".

It cannot make weak content strong, make a slow site fast or guarantee search rankings. It also cannot be finished once and forgotten. New services, new audiences and new teams all put pressure on a structure, which is why a good plan includes rules for how it changes.

Who does information architecture

On small teams, information architecture is part of someone else's job: the web manager, a content designer or the developer building the site. On larger teams it may be a dedicated role, or a stage of work given to a specialist. What matters is that someone owns the structure as a whole, rather than each team owning its own corner and nobody looking at how the corners fit together.

Signs your structure needs attention

• Visitors phone or email to ask questions the website already answers.
• Several pages cover the same topic and disagree with each other.
• Staff cannot agree where a new page should go.
• Menu labels are department names or internal jargon.
• Site search is the busiest page on the site.

If several of these sound familiar, an information architecture audit will show which causes are structural and which are not. If you would rather start by looking for yourself, paste an outline into the hierarchy visualiser.

Questions about this topic

Is information architecture the same as UX design?

It is one part of user experience design. UX covers the whole experience, including interaction design, visual design, research and content. Information architecture focuses on how content is organised, labelled and connected.

Is information architecture the same as a sitemap?

A sitemap is one output of information architecture work. The work also covers labels, taxonomy, page addresses, navigation and the links between pages.

Do small websites need information architecture?

Every website has one, planned or not. A small site may only need an hour of thought about its menu labels. A large site with several audiences benefits from a written plan.

Sources checked 4 November 2026 for what is information architecture?

1. Hierarchical structure, Style Manual
2. SEO Starter Guide, Google
3. Structure testing, European Commissionsecondary
4. Multiple Ways (2.4.5), W3C
5. Consistent Navigation (3.2.3), W3C
6. Card sorting, Digital NSW
7. Links, Google