Information Architecture Explained
22nd Jul 2026 12 minutes read 16 sections
Overview
A website becomes easier to use when people can understand what information is available, where it belongs and how to reach the next useful step.
This guide explains what information architecture is, the decisions it involves and how planning it well can help people find information while making the website easier for an organisation to manage.
What is information architecture?
Information architecture is concerned with how information is grouped, labelled, ordered and connected across a website.
Good information architecture helps people answer three basic questions:
- Where am I?
- What can I find here?
- Where should I go next?
Someone should not need to understand the whole website to use one part of it. They need enough information to recognise whether they are in the right place and which route is most likely to help.
The decisions involved can affect:
- routes designed around tasks, subjects or audiences
- the sections within a website
- the relationship between parent and child pages
- navigation and category labels
- landing pages that introduce several choices
- links between related information
- search filters and groupings
A sitemap is only part of the picture
A sitemap is one possible representation of a website’s hierarchy. It may show the main sections, the pages beneath them and how many levels the structure contains. This makes a sitemap useful, but it’s not the information architecture itself.
A neat diagram does not prove that people will understand the choices it contains. It may not explain why information has been grouped in a particular way, whether people recognise the labels or how someone might move between related subjects in different sections.
A sitemap may also struggle to show information that belongs within more than one journey. A page could have one agreed home while still needing links from several other places. Different audiences may require different starting points to reach the same information.
A sitemap records part of the proposed structure. It does not replace the research and decisions behind it.
Why website structure can become confusing
Website structures usually become confusing over time rather than because of one poor decision.
A department may add a new section because none of the existing options feels suitable. Elsewhere, another team may publish similar content without realising that a version already exists. Each addition may seem reasonable, but over time the website accumulates overlapping sections, duplicated information and labels that only make sense to staff.
The structure may still make sense to staff because they know which team runs each service and understand the internal language. Visitors do not begin with that knowledge.
Someone trying to report pollution in a local river, for example, should not need to decide whether the issue belongs to a regional office, an incident team or a regulatory department. The website can still reflect the organisation where that information is useful, but it should not make internal knowledge a requirement for finding help.
Organising information around what people need to do
Website structures often work better when they reflect what people are trying to do rather than how the organisation is arranged. ‘Report an environmental concern’, for example, is likely to be more useful than the name of the team responsible for handling it.
Organising a website around tasks will not suit every situation. Some websites need to use established professional language, while others serve people who already understand the organisation and want quick access to a particular department or technical area.
In some cases, people may need more than one useful route to the same information. They might begin with a task, subject, service name or audience section, depending on what they already know.
What information architecture covers
A few key areas affect how well a website is organised.
Organisation
How information is grouped and where one subject, service or task ends and another begins.
Hierarchy
Which information people see first, what sits underneath it and how much detail appears at each level.
Labelling
The words used for sections, categories, links and choices. Labels should help people predict what they will find.
Relationships
How information is linked across the website, including subjects that belong in more than one section or task.
Findability
Whether people can find information through navigation, links, search or another route, even when they do not know what the organisation calls it.
These elements need to work together. Information may be grouped sensibly, but people still need a label that tells them what they will find. A good label cannot rescue a confused section, and a well-placed page can still be hard to find without useful links.
What makes a website structure work?
There is no single structure that suits every organisation, but people should be able to recognise the available options without having to learn internal terminology first. Similar subjects should also be grouped and labelled consistently, so understanding one part of the website helps them use another.
Sections need a recognisable purpose. When several areas appear to cover the same subject, people have to guess which one contains the right information. Related information should also remain connected when a task crosses between departments or content areas.
New information should fit into the existing structure without creating another top-level menu item every time. Testing helps to check whether the groupings make sense and where people would look for something new.
Use evidence, not assumptions
Stakeholder views matter, but they should not determine the structure on their own.
Internal expertise is important. Staff understand services, responsibilities, regulatory requirements and practical constraints that people outside the organisation may never see. The difficulty begins when that familiarity is treated as evidence that everybody else will think in the same way.
Useful evidence may also come from:
- website search terms
- analytics and common routes
- enquiries and support requests
- existing research
- a review of current content
- recurring questions and observations from customer-facing teams
Card sorting and tree testing are two useful ways to test information architecture.
Card sorting
People group topics in ways that make sense to them. This can reveal common patterns, differences between audiences and category labels or groupings that may need further investigation.
Tree testing
People are given tasks and asked where they would look within a text-only version of the website’s hierarchy. Because page layouts and visual design are removed, the test can reveal whether the categories and labels help people find the right area. Patterns of hesitation, wrong turns or failed tasks can then guide changes before the structure reaches design and development.
Neither method gives you a final answer. The findings still need to work for different audiences and take account of any legal, operational or technical needs.
When audience sections are useful
Creating a separate section for each audience can work when those groups need different information or use different language. For example, professional advisers and members of the public may need different ways into the same service.
Audience sections become less useful when they repeat the same information in several places. They can also be difficult for people who fit more than one group. Someone might be a resident, volunteer, parent and professional partner, so asking them to choose one audience before seeing the available information may create an unnecessary decision.
- Before separating information by audience, useful questions include:
- Can each section be maintained accurately?
- Does this group need different information or a clearer route to shared information?
- Will people recognise which audience description applies to them?
- Could someone belong to more than one group?
- Would the separation hide related information?
- Can each section be maintained accurately?
How information architecture and navigation work together
Information architecture explains how a website is organised and how its different parts connect.
Navigation gives people ways to move through that structure. The main menu is one part of it, but people may also use search, breadcrumbs, filters, landing pages, calls to action and links between related content.
A main menu has limited space, so it needs to focus on the choices that will help the greatest number of people. Trying to include every service, subject and audience at the top level usually makes it harder to scan and use.
That does not mean each page needs only one route. A page can have one main home while still being reached through search, a related link or a landing page organised around a particular task.
Information architecture connects those routes across the website. Navigation is how people find and use them.
Common signs that your website’s structure isn’t working
Problems often become noticeable as the website grows and people use it in different ways. Common signs include:
Everything ends up in the main menu
New options keep being added because people lack other reliable ways to find information.
People need to know which team is responsible
Someone has to understand the organisation before they can find the right service or information.
The same subject appears in several places
Similar content has been copied across sections because it lacks a single agreed-upon home.
Labels only make sense to internal team members
Section names and categories use language that people outside the organisation may not recognise.
Only journeys from the homepage have been planned
Visitors arriving through search, email or a shared link struggle to understand where they are or what to do next.
New content has nowhere obvious to go
Every addition leads to another exception, compromise or request for a new top-level section.
People search for basic information
Visitors repeatedly use website search because the navigation does not make the right section easy to recognise.
None of these signs automatically means the whole structure is failing. Search may be a useful way to find information, and some subjects may need links from several parts of the website even when the content has one main home.
Planning your website’s information architecture
The amount of work required will depend on the size and complexity of the website. A small service website may only need a focused review, while a university, membership organisation or public body may need to organise thousands of pages for several different audiences.
The process will depend on the website and the decisions being made, but these stages are a useful place to begin.
1. Understand the website’s purpose
Establish what the website needs to help people understand or do. This gives the structure a basis beyond the current menu or organisation chart.
2. Understand people and their main tasks
Use research and operational evidence to identify why people visit, what they expect to find and the language they use.
3. Review the existing information
Identify what exists, where subjects overlap, what appears to be missing and which labels or groupings cause confusion.
4. Identify useful groups and relationships
Explore which subjects belong together, where tasks cross and which information needs to be reached from more than one place.
5. Develop a proposed hierarchy
Use what’s been learned to decide which choices people see first and what sits beneath them. Introduce detailed options only after the main choices are clear.
6. Test the structure
Try the proposed structure with real tasks. Card sorting, tree testing or usability testing can reveal where people hesitate or head in a different direction.
7. Refine the labels and routes
Replace or explain terms people do not understand, and review choices that are too similar or difficult to tell apart.
8. Use the structure across the website
Apply the agreed structure to navigation, landing pages, page planning and internal links. Use the same principles when new content is added.
The stages may not happen in this order every time. Testing can send the team back to an earlier decision, and reviewing the content may show that separate sections need to be brought together.
Information architecture supports user journeys
Research and journey mapping can show what people need at each stage of a task. Information architecture helps to identify where that information should sit and which pages need to link together.
A fuller explanation of planning and reviewing complete tasks is available in User Journeys Explained.
Maintaining information architecture after launch
A website’s structure still needs attention after launch. Services change, teams are renamed and new requirements emerge. Without clear responsibility, individual additions can gradually recreate the same problems the original work was meant to solve.
Ongoing work may include:
- reviewing common search terms and tasks where people regularly get stuck
- removing sections that are no longer needed
- checking whether a new top-level menu item is really necessary
- keeping labels consistent across the website
- deciding where new content should sit
- adding or updating links between related subjects
- testing major structural changes before rolling them out more widely
Information architecture should make change easier to manage, not prevent it. Teams need an agreed way to decide where new information belongs and how it connects with what is already there.
A structure people can understand
A website should not expect people to understand the organisation before they can find its services. The organisation needs to do that work on their behalf.
Good information architecture does not remove organisational complexity. It stops passing so much of that complexity on to the people using the website.
