Building theme-based information architectures

By Tim Caynes @timcaynes

There is often a temptation to dive into information architecture (IA) design based solely on acquired knowledge and a well-articulated business objective. It’s quite possible to produce meaningful taxonomies and content structures in this way, especially for discrete, closely scoped projects. However, some of the most effective structures evolve when iterative analysis of research findings and discussion outputs start to surface emerging themes.

I recently worked on the development of a new sales and investments channel for one of our financial services clients. This kind of content is not an unusual proposition and there is a depth of experience within the team that might predict some of the outcomes. We might even consider sketching out the concepts in our heads just to start the conversation, since the business objective is reasonably clear. In doing that, we might even get it done quicker than anybody planned, or budgeted for. I might even get home in time for tea.

How we really maximise the potential of the design phase is if we take the insights from a well-executed discovery phase, invest meaningful time to digest and analyse the messages we’re getting from customers, and start to build up a picture of an ideal customer journey. I’m not talking here about spending a day in a dark room with a stationery cupboard’s worth of post-it notes and walking away with a full site map. All we should really be aiming for at the end of an initial design session is identifying those emerging themes and understanding how accurately they reflect the voice of the customer.

Themes are really just a logical clustering of questions, statements and pain-points related to a customer’s interaction. Since they’re deliberately vague at the early stages of design, they’re not distinct categorisations with well specified hard attributes, but more of an expression of the customer needs and desires. If that doesn’t sound pompous enough, I’d also suggest that they often describe the soft attributes of the customer’s emotional engagement with the site, for example, ‘I’m worried about the future’, or ‘I’m not sure I should be putting all my eggs in one basket’. Identifying these themes early provides a clear customer-centric reference when iterating into task-based journeys and navigation models.

Themes are also a great way to solicit empathetic feedback. It’s more meaningful to describe a potential customer journey to a client, for example, if they can clearly empathise with the customer. Using needs-based descriptions with more open language, through themes, it can be much easier to evaluate a customer journey than by simply referencing a label on a navigation entry point. Consider the difference between ‘Protecting my future’ as a theme and ‘Savings’ as a label. While one might not map directly to the other, there is still a much clearer expression of the customer need through a theme-based approach, which sets the context for discussion and iteration.

There are a number of approaches to how we take discovery outputs and begin to build out information architectures that place the customer at the heart of the discussion. Using a theme-based approach is one of those that I rather like and it has proved successful in articulating the voice of the customer, particularly when talking with clients. What are your experiences of or opinions on theme-based IA?

Tim Caynes

I’m a Principal Designer at Foolproof which means I’m responsible for the integrity of the design thinking that shapes the work that we do.

View Tim's profile

What do you think?