Information Architecture

Organising meaning, not just content. In a digital environment, the difference between a system that works and one that frustrates users almost never comes down to technology — it lies in the invisible structure beneath it.

The way information is organised, labelled, interconnected and made accessible to the right people, at the right moment, through the right channel: this invisible structure has a name: Information Architecture.

Information Architecture is the discipline of designing and organising information environments — both physical and digital — in a semantically coherent and logically comprehensible way, making access straightforward and meaningful for users. It is not simply about deciding where to place content: it is about building the conceptual foundations on which the entire user experience rests.

Its scope extends across three distinct but interdependent levels. The first concerns internal coherence: every information system must be intuitively navigable, with logical structures that reflect the way users think, search and orient themselves. The second concerns external coherence: in an omnichannel world, the same architecture must preserve its meaning across platforms, devices and contexts of use — from website to mobile app, from portal to internal search system. The third, deeper and often overlooked, concerns relational coherence: the capacity to place each piece of content within a broader ecosystem of connections, where every item of information generates context, meaning and knowledge for those who encounter it.

The foundational pillars: the iceberg metaphor

To understand how Information Architecture truly works, it helps to think of an iceberg. What users see — navigation, labels, menus, search results — is the tip above the water. It is the interface, the visible point of contact between people and the system. But beneath the surface lies the larger, more important part: the part that determines whether the system performs well or poorly, whether it scales over time, whether it holds up as content grows and complexity increases.

Esario designs information architecture from the bottom up, because that is where the solidity of any system is built.

The seabed: Users, Content, Context

Every Information Architecture project begins with three fundamental questions: who are the users, and what are their information needs and behaviours? What content exists, with what structure and what meaning? Within what cultural, organisational and technological context does the system operate?

The deep structure: Metadata and Controlled Vocabularies

Above the foundations, but still invisible to the user, lies the most technical and most powerful layer of the architecture: metadata and controlled vocabularies.

A metadata element is information that describes other information: the title of a document, the category of a product, the type of a page. Without structured metadata, an information system is blind — it does not know what it contains, cannot retrieve content intelligently, and cannot connect it to related material.

Controlled vocabularies — taxonomies, thesauri, classification schemes — are the semantic infrastructure that gives coherence to the entire system. They define which terms are authorised to describe the concepts of a given domain, how those terms relate to one another (which are broader, which are narrower, which are associated without being in a hierarchical relationship), and how the natural language of users is translated into the structured language of the system.

A well-built thesaurus, for instance, knows that when a user searches with different words, they are looking for the same thing — and knows how to guide them towards the right content, even when the words they use do not match exactly those used by whoever produced that content. This semantic infrastructure is not a technical detail: it is what makes an information system navigable, retrievable, scalable and interoperable over time.

The intermediate structure: Strategies and Project Plans

With solid foundations in place, Esario builds the architectural strategy: how to organise content into categories and hierarchies, how to design navigation and search systems, how to define user journeys through the system, and how to ensure the architecture holds up as content and user needs evolve.

This phase produces the project documents that guide development: content maps, semantic relationship schemas, structural plans for navigation systems.

The surface: Wireframes, HTML Prototypes, Testing

Only at this point — with foundations, semantic structure and strategy consolidated — does Esario move to designing the visible interface: wireframes, mockups, blueprints and HTML prototypes. These artefacts are not the starting point of the project, but its first tangible outcome: the materialisation of all the decisions made at the layers below.

HTML prototypes are not merely presentation tools: they are environments in which to conduct usability testing with real users, verifying that the designed architecture genuinely responds to their needs and behaviours. Test results feed back into the structure, refining architectural choices before they are locked into final development.

Information Architecture and Artificial Intelligence

The growing complexity of digital information systems — portals with thousands of items of content, omnichannel platforms, distributed data ecosystems — has made it essential to complement traditional architectural design with the support of Artificial Intelligence. Esario integrates AI tools across all phases of the process, delivering concrete results in terms of quality, speed and analytical depth.

In the content and user analysis phase, large language models make it possible to process large volumes of texts, documents and data rapidly, identifying the conceptual patterns of the domain, the prevailing thematic areas, the most frequent terms, and the implicit semantic relationships between items of content. Work that would take weeks of manual analysis can now be produced in significantly less time, with a level of coverage that would be unachievable manually.

In the construction of controlled vocabularies, AI tools assist in defining taxonomies and thesauri: suggesting candidate terms, proposing hierarchical and associative relationships, identifying synonyms and linguistic variants, and flagging potential inconsistencies. The designer retains control over semantic choices — which always require expert judgement of the domain and the users — but AI dramatically reduces the exploratory and initial structuring work.

In architecture validation, computational semantic analysis techniques make it possible to verify the internal consistency of the vocabulary, detect conceptual overlaps between categories, and simulate search behaviours to test the quality of information retrieval before the prototype has even been built.

Finally, information systems equipped with well-structured semantic architectures are today the ideal substrate for powering more advanced AI applications — semantic search engines, recommendation systems, conversational assistants — because they provide the machine with a conceptual map of the domain on which to reason accurately and without ambiguity.