Managing internal consensus was a challenge for the British Library’s web redevelopment programme. The organisation was somewhat ‘silo-ed’. Each silo worked to a different understanding of its audience, their requirements and how to meet them.

The Library had already commissioned some personas, articulating a discipline-based segmentation model. Unfortunately, the result was 14 highly subjective personas, one for each of the key business units within the organisation.

With such a compromised audience segmentation and no clear idea of user needs, making evidence-based strategic and design decisions was going to be impossible. In order to rectify this, we proposed re-framing their understanding of Library users. Our aim was to demonstrate a new, task-based segmentation and undertake research with real users to provide the Library with meaningful insights.

Working with the existing personas, we drafted a task-based segmentation model. Our primary segmentation was between what might be described as traditional ‘researchers’ – academics and professionals looking for primary source material – and ‘discoverers’ – users with less defined research goals and understanding of the topic of their research. This admittedly crude distinction was enough to demonstrate to stakeholders that ‘their’ users had much more in common with each other than they had realised.

Next, we created a mental model diagram based on interviews with users across the library’s existing discipline-based segmentation and our proposed task-based model. This gave us the opportunity to gain insights into how users research and to iterate our segmentation in order to draft new personas. After analysing transcripts of interviews with more than thirty users, we were able to build a model which validated our original segmentation, provided plenty of talking points for the Library and gave us a framework with which to understand the value to users of proposed content and features.