What is information architecture (IA)?

What information architecture covers

Jen Cardello of the Nielsen Norman Group (NN/g) describes information architecture in two parts. First, deciding which content and features the site has. Second, the organisation, structure and naming that define the relationships between them.

This work happens in spreadsheets and diagrams. Building a content inventory and choosing category names are part of it too. Cardello puts it briefly: the information architecture informs the interface.

Often confused with: navigation

Navigation is the face of the information architecture on screen. The main menu, side menu, breadcrumbs, filters and footer are navigation components. The information architecture is the structure that decides what those components show.

The order matters: structure first, then navigation. In Cardello’s example, a team picks an inverted-L menu because it likes how it looks. But that menu only suits sites that are no more than four levels deep.

According to Cardello, the information architecture doesn’t need to be final before you start on the wireframe. Still, I think the main branches should be clear. Before I design a screen, I ask which branch it sits in.

Example: an online store menu

Imagine a store with six products. The same products can be grouped by two different logics.

Tree A arranges the products like the shelves in a warehouse. Tree B starts from the question someone shops with. The thermos is in both trees, but next to different products.

Which one is right depends on the users’ mental model. You only find out by testing with users.

How to test it

In card sorting, participants sort labelled cards into groups that make sense to them. In tree testing, they are asked to find where something specific lives in a category tree with no visual design or content.

According to NN/g, card sorting usually doesn’t produce the final structure. You also need to confirm the resulting tree with a tree test.

Why it matters

If users can’t find what they’re looking for, the content doesn’t exist for them. That’s why information architecture sits at the foundation of usability.

However careful the visual design is, it won’t rescue a product in the wrong branch. The search box doesn’t always close this gap either.

Where to start

If you have a website, the first step is a content inventory. Put every page and every feature in a spreadsheet.

Then go through the inventory and mark outdated or duplicate content. Group the rest and give each group a name.

Naming is easy to underestimate. If the menu has “Solutions”, “Services” and “Products” side by side, users have to guess which one holds what they want. Choose names from the words customers use to describe these things.

Finally, test the tree, then draw the menu. If you reverse the order, you’ll have to squeeze content that doesn’t fit the menu in somewhere.

Related terms

• User-centred design: the approach of building the structure from the user’s side.
• Prototype: the first attempt at the structure on screen.
• UX: the wider discipline information architecture belongs to.