Category: Eywa Explains
Abstract / Executive Summary
Information architecture (IA) is the discipline concerned with how information is organised, labelled, connected, navigated and retrieved within digital environments. Although often associated with website menus and sitemaps, IA encompasses a broader system of structures and relationships that enables people to find information, understand where they are, and determine where to go next. Rosenfeld, Morville and Arango’s established framework identifies organisation, labelling, navigation and search as central components of information architecture, while also emphasising metadata, classification and the relationship between users, content and organisational context (Rosenfeld et al., 2015).
This article examines the conceptual foundations of IA, its relationship with usability and accessibility, common methods for designing and evaluating information structures, and its relevance to complex digital platforms. Research indicates that information organisation and navigation can materially affect users’ ability to retrieve information, although the strength and generalisability of individual studies vary. More recent research also demonstrates that methods such as card sorting and tree testing can reveal differences between intended structures and users’ expectations.
The central argument is that IA should be treated as an underlying system rather than a purely visual design exercise. Effective IA connects user needs, content, terminology, navigation and technology into a coherent structure that can evolve as information and user needs change.
Keywords
Information architecture; UX design; information organisation; navigation; findability; taxonomy; usability; digital platforms
Main article
1. Introduction
Digital systems increasingly function as information environments rather than collections of isolated pages. A website may contain hundreds or thousands of pages, documents, services, datasets and interactive functions. An enterprise platform may combine structured records, reports, workflows, search functions and external data sources. In both cases, users need to answer deceptively simple questions: Where am I? Where can I find what I need? What does this category mean? What should I select next?
These are fundamentally information architecture problems.
Information architecture is important because the existence of information does not guarantee that people can find or understand it. A study of an academic website, for example, examined how information was categorised, labelled and presented and how navigation and access were facilitated. Among 24 participants, just over half of the information-seeking questions were answered successfully, illustrating how the structure of a website can affect information retrieval independently of the information itself (Gullikson et al., 1999).
The objective of this article is therefore to explain what IA is, what it consists of, how it differs from visual design and software architecture, how it can be designed and evaluated, and why it matters for contemporary digital platforms. The discussion draws on foundational IA literature, peer-reviewed research and current accessibility guidance.
2. Conceptual / Theoretical Background
What does “information architecture” mean?
There is no single definition that captures every use of the term. In contemporary digital design, however, IA generally concerns the organisation and structuring of information environments so that people can find, understand and use information.
Rosenfeld, Morville and Arango describe IA through several interrelated systems, particularly organisation, labelling, navigation and search, alongside supporting elements such as metadata and classification (Rosenfeld et al., 2015). Their fourth edition, published in 2015, continues to frame IA as applicable across websites, mobile environments and other digital information spaces.
The distinction between IA and visual design is important. Visual design determines how an interface looks and communicates visually. IA determines, among other things, what belongs together, what something is called, how users move between information, and how information can be retrieved.
Similarly, IA should not be confused with software architecture. Software architecture concerns the technical organisation of software components, services and infrastructure. IA is primarily concerned with the structure of the information environment from the perspective of people and information retrieval.
The four core systems
A useful starting point is to consider four interconnected systems:
| IA component | Central question | Typical examples |
| Organisation | How should information be grouped? | Categories, hierarchies, topic groupings |
| Labelling | What should each group or item be called? | Menu labels, headings, terminology |
| Navigation | How can users move through the information space? | Global navigation, local navigation, breadcrumbs |
| Search | How can users retrieve information directly? | Search boxes, filters, facets, search results |
These systems should not be designed independently. Changing a category can require new labels; changing labels can affect navigation; changes to metadata can alter search results.
IA is therefore better understood as a system of relationships than as a single deliverable such as a sitemap.
Users, content and context
One of the important principles in the IA literature is that structure cannot be designed independently of users, content and context (Rosenfeld et al., 2015).
The same information may need to be organised differently for different audiences. A technical platform used by data specialists may legitimately expose terminology and metadata that would be inappropriate for a public-facing service. Conversely, a structure designed entirely around internal organisational terminology may make sense to staff while remaining difficult for external users to understand.
This is why IA involves both information analysis and user research.
3. Literature and Evidence Review
Research has repeatedly linked information organisation and navigation with the ability of users to retrieve information.
The 1999 study by Gullikson et al. examined an academic website and found substantial difficulty among participants in answering typical information-seeking questions. The study specifically identified categorisation, labelling, presentation, navigation and access as important aspects of the site’s information architecture. Its small sample means the numerical findings should not be generalised directly to all websites, but the study provides empirical evidence that information structure can affect task performance.
Subsequent research has examined particular IA techniques. Card sorting, for example, asks participants to group information items according to categories that make sense to them. A study of a medical information website found that card sorting could both validate existing categories and reveal emerging information needs that were not sufficiently represented in the site’s menu structure.
Research has also shown that information classification is not necessarily universal. A cross-country study involving Danish and Pakistani university students identified differences in participants’ classification of website information and related task performance. This suggests that assumptions about how information should be categorised may not transfer automatically between user groups or cultural contexts.
More recent work has examined tree testing, a method used to evaluate navigation structures independently of detailed visual interface design. A 2025 study involving 180 participants and 1,800 task completions compared several tree-testing variants with interactive prototypes. It found that the representation used during testing could affect results and that different tree-testing approaches did not always produce identical interaction patterns. This is an important methodological qualification: testing IA is useful, but the test method itself can influence what is observed.
Overall, the literature supports several broad conclusions:
- Information structure can affect users’ ability to retrieve information.
- User expectations do not necessarily correspond to organisational structures created by internal teams.
- Card sorting can help investigate users’ classification and terminology.
- Tree testing can help evaluate whether a navigation hierarchy supports task completion.
- IA is not static; information needs and structures can change over time.
At the same time, evidence is often based on specific websites, user groups or controlled tasks. IA research therefore supports the importance of user-centred structuring without establishing a universal formula for what every digital product’s architecture should look like.
4. Analysis and Discussion
IA is about making information findable
A useful way to understand IA is through the concept of findability: the degree to which information can be located when someone needs it.
Findability depends on more than search. A user may reach information through a navigation menu, a related-content link, a filter, a taxonomy, an internal search function or a combination of these mechanisms.
This is reflected in accessibility standards. WCAG 2.2’s navigation guidance states that users should have ways to navigate, find content and determine where they are. Its guidance recognises that navigation serves two fundamental functions: communicating the user’s current location and enabling movement to another location.
The implication is significant: a well-structured information environment should support orientation as well as retrieval.
IA is not just a sitemap
A sitemap is one representation of an IA, but it is not the architecture itself.
Consider a digital platform containing projects, opportunities, documents, requirements, tasks, agreements and reports. A sitemap might show where each section sits in the interface. It does not necessarily explain:
- how an opportunity relates to a project;
- whether a document belongs to an opportunity or an agreement;
- which terminology users understand;
- what metadata should be attached to a record;
- whether users should browse or search for an item;
- how related information should be surfaced.
These relationships are part of the underlying information architecture.
For complex platforms, IA consequently intersects with taxonomy, metadata, content modelling, search and data relationships.
IA and accessibility
Accessibility is another reason IA should be treated as structural rather than cosmetic.
W3C guidance emphasises meaningful headings, descriptive links, clear navigation mechanisms and multiple ways of locating content. WCAG 2.2 Success Criterion 2.4.5, for example, requires more than one way of locating a page within a set of pages, except where the page is part of a process.
Headings are particularly important because assistive technologies can use semantic structure to help users navigate content. Likewise, meaningful link text enables users to understand destinations without relying solely on surrounding visual context.
Consequently, accessibility and IA are closely connected. An architecture that is ambiguous, inconsistent or overly dependent on visual cues can create barriers even when the interface appears visually polished.
IA in complex digital ecosystems
The importance of IA becomes more pronounced as systems become more complex.
Eywa Systems’ materials describe digital platforms that consolidate biodiversity, geospatial and monitoring data into a single source for conservation management, alongside GIS, environmental analytics, carbon registries, climate finance and other systems.
In such environments, IA is not simply about deciding which items appear in a navigation bar. It may involve defining how different information domains relate to one another, how records are classified, which terminology is used across systems, and how users move between datasets, reports and workflows.
This also illustrates why IA should be considered early in digital transformation projects. If information structures are poorly defined, later interface development may reproduce those structural problems at a more expensive stage.
5. Challenges, Limitations, and Counterarguments
There is no universally “correct” architecture
Information architecture is partly dependent on context. Different users can have different mental models of the same information.
A structure that works for technical specialists may not work for policymakers, field practitioners or members of the public. The cross-country research on university websites provides evidence that classification preferences can vary between groups.
This means that IA should not simply reflect the organisation’s internal hierarchy.
Organisational structures can conflict with user structures
Organisations frequently structure information according to departments, programmes or internal responsibilities. Users, however, may think in terms of tasks or outcomes.
For example, an organisation might divide information between three internal departments, while a user simply wants to “apply for funding”. Reproducing the organisational chart as the navigation structure may therefore force users to understand the organisation before they can complete their task.
The appropriate solution cannot be assumed in advance. User research and task analysis are needed to establish how the relevant audience actually seeks information.
IA changes over time
Information environments are not static. Content grows, terminology changes, new services appear and user behaviour evolves.
Research on a medical information website illustrates this point: a structure that remained broadly valid after a year nevertheless revealed additional themes and information needs through renewed card sorting and analysis of actual search queries.
IA should therefore be treated as an ongoing governance activity rather than a one-off design exercise.
Testing methods also have limitations
Card sorting can reveal how users group information, but it does not necessarily tell designers how the resulting structure will perform in a complete interface. Tree testing can isolate aspects of navigation, but recent research shows that different testing variants can produce different results.
A robust IA evaluation therefore benefits from multiple forms of evidence rather than relying on a single research technique.
6. Implications
For organisations developing websites, enterprise platforms or data-intensive digital systems, several practical implications follow.
First, IA should begin with information needs rather than interface components. Before deciding what appears in a menu, teams should understand what users need to accomplish and what information they need to complete those tasks.
Second, terminology deserves deliberate attention. Labels are not merely copywriting. They determine whether users recognise where a particular piece of information belongs. W3C accessibility guidance similarly emphasises headings, labels and descriptive link text.
Third, structure should be tested with representative users. Card sorting can explore how users classify information, while tree testing can evaluate whether a proposed hierarchy supports realistic tasks. Evidence from recent research supports using such methods while recognising their methodological limitations.
Fourth, search should not compensate for poor organisation. Search is an important access mechanism, but IA also determines metadata, categories and relationships that influence how information can be discovered and understood.
Finally, IA should be integrated with data and technology strategy. In complex digital environments, information architecture connects user-facing navigation with underlying content models, metadata, search and system relationships. This is particularly relevant to platforms that bring together multiple information domains, such as geospatial, environmental and monitoring data.
7. Conclusion
Information architecture is the structural foundation that enables people to find, understand and navigate information within a digital environment. It encompasses much more than menus or sitemaps: organisation, labelling, navigation, search, classification, metadata and relationships between information all contribute to the architecture.
The evidence indicates that information structure can materially influence information retrieval and usability, while research into card sorting and tree testing demonstrates the value of testing structures against users’ expectations. However, there is no universal architecture that can simply be transferred from one digital environment to another. User groups, organisational contexts, information domains and technological constraints all matter.
For this reason, effective IA is best understood as an iterative, user-centred and evidence-informed discipline. Its purpose is not to make a digital product look organised, but to make the underlying information environment coherent enough that people can locate what they need, understand where they are, and move confidently through the system.
References
- Gullikson, S., Blades, R., Bragdon, M., McKibbon, S., Sparling, M., & Toms, E. G. (1999). The impact of information architecture on academic web site usability. The Electronic Library, 17(5), 293–304. doi:10.1108/02640479910330714.
- Nawaz, A., Clemmensen, T., & Hertzum, M. (2011). Information classification on university websites: A cross-country card sort study. IRIS: Selected Papers of the Information Systems Research Seminar in Scandinavia, 2, 109–122.
- Rosenfeld, L., Morville, P., & Arango, J. (2015). Information architecture: For the web and beyond (4th ed.). O’Reilly Media.
- Toms, E. G., et al. (1999). The impact of information architecture on academic web site usability. The Electronic Library, 17(5), 293–304. doi:10.1108/02640479910330714.
- World Wide Web Consortium. (2023). Web Content Accessibility Guidelines (WCAG) 2.2.
- World Wide Web Consortium Web Accessibility Initiative. (2026). Understanding Guideline 2.4: Navigable.
- World Wide Web Consortium Web Accessibility Initiative. (2026). Understanding Success Criterion 2.4.5: Multiple Ways.
- World Wide Web Consortium Web Accessibility Initiative. (2026). Understanding Success Criterion 2.4.9: Link Purpose (Link Only).
- Note: The 1999 study appears in the literature under the title “The impact of information architecture on academic web site usability”; it is cited once above in its verified bibliographic form rather than treated as a separate body of evidence.