Information Architecture
Information Architecture
Information Architecture (IA) is the invisible skeleton of digital products. When it is done well, users navigate to what they need without thinking about how they got there. When it is wrong, users wander, backtrack, and eventually leave.
IA is not about visual design. It is about structure — how content is organised, how it is labelled, and how users can move through it. The same visual design on bad IA is still unusable. Great IA makes even average visual design navigable.
The Core Problem IA Solves
Users do not browse products like they browse libraries. They arrive with a goal ("I need to change my billing address," "I want to find last month's invoices," "I need to cancel my subscription"). The IA's job is to support that goal with the fewest number of steps and the lowest cognitive load.
The fundamental tension: product teams organise information around their internal logic (teams, product lines, technical categories). Users organise information around their own tasks and mental models. These are rarely the same. Card sorting (covered in User Research) is the primary tool for bridging this gap.
Taxonomy and Hierarchy
Most digital products organise content hierarchically — a tree structure with broad categories at the top that branch into increasingly specific subcategories. The key decisions:
Breadth vs. depth: A shallow, broad taxonomy (many items at top level) requires fewer clicks but more scanning. A deep, narrow taxonomy (few top-level items that branch deeply) requires more clicks but each choice is simpler. Research suggests users prefer shallower hierarchies — 3–4 items per level rather than 2–3 levels deep.
Category naming: Labels must match users' language, not the organisation's. "Account Management" may be internal terminology; users think "settings." Testing label comprehension is one of the highest-leverage IA research activities.
Faceted classification: Unlike strict hierarchies, faceted systems let users filter across multiple dimensions simultaneously (e-commerce: filter by colour, size, brand, price range all at once). More powerful for browsable content; unnecessary overhead for task-focused navigation.
Navigation Patterns
Global navigation provides access to all major sections from anywhere in the product — typically a top navigation bar or persistent sidebar. It must answer: "Where am I? Where can I go?" at every moment.
Local navigation shows the structure within the current section — a sidebar menu in a settings panel, tabs within a content category. It handles the next level of hierarchy below global.
Breadcrumbs show the user's path through the hierarchy: Home > Account > Billing > Change Address. They support wayfinding and provide one-click returns to parent levels. Essential when users can enter the system from search (landing deep in the hierarchy without traversing the top).
Contextual navigation links to related content within the current page — "See also," related articles, next steps. It supports discovery and reduces dead ends.
Search is an alternative to navigation, not a substitute for it. Users who cannot find what they need through navigation will resort to search. A high search rate for basic content is a warning sign that IA is failing.
Hamburger Menus and Mobile Navigation
The hamburger menu (≡) hides navigation behind a tap. Research consistently shows lower discoverability and engagement compared to visible tab bars. Yet mobile screens cannot always accommodate a full tab bar. The tradeoff:
- Tab bars (visible): higher discoverability, limited to 4–5 items
- Hamburger menus: more items, lower discovery, better for secondary navigation
- Hybrid: primary 4 items in tab bar, overflow in hamburger
Bottom navigation (tab bar at the screen bottom) is now the iOS and Material Design standard — it respects thumb reach zones.
Progressive Disclosure
Progressive disclosure hides complexity until users need it. Advanced settings are collapsed behind "More options." Complex forms show only the relevant next step. Long content is truncated with a "Read more." This reduces cognitive load and avoids overwhelming novice users while keeping advanced functionality available.
The risk: burying frequently-needed options. Progressive disclosure requires research to establish which features are truly advanced versus which are routinely used.
Content Audits and IA Review
Before redesigning IA, audit the existing content: what exists, how much of it there is, whether it is outdated, duplicated, or missing. A content audit creates the inventory IA is built around. Many IA problems are content problems — too much content competing for attention, no clear hierarchy in the content itself, missing bridging content that helps users move between sections.