Building better user personas for a better public service


September 28, 2016

At the end of last year, the UK civil service’s Performance Platform team completed a (re)discovery project. We wanted to look at how people use data to improve performance.

As part of this project we used Activity Theory to build personas. This helped us identify our primary user persona: Anna, the strategist.

Good and bad personas

Personas have always been a controversial topic in user research. People tend to love them or hate them.

Personas can be a waste of time if they are:

  • built based on assumptions that aren’t taken from user research data
  • produced without going through a rigorous process of analysis
  • clusters of superficial groupings based on demographic characteristics or job roles

However, personas can be very useful if they are:

  • developed based on sufficient research data and rigorous analysis
  • clusters of people who share similar goals, motivation, and behaviour/needs

Using Activity Theory to make better personas

I never used to be convinced about the value of personas. But this changed two years ago, when I was working with Charline Poirier, head of research at Canonical at the time. She introduced me to a framework based on Lev Vygotsky’s Activity Theory.

Let’s say there are two doctors. One wants to help people and save lives. The other one became a doctor to get rich.

They have the same job role. They appear to have the same goal and carry out similar tasks. But does that mean they would exhibit the same behaviour and needs? No.

The two doctors’ different motivations would result in a variety of behaviour and needs. Some of these might be the same, but others might be contradictory.


This is the value of Activity Theory. Rather than simply generating a list of needs in isolation, it adopts a holistic approach to understanding people’s activities. It explains actions and needs by exploring the context in which they occur.

This allows us to see the reasons behind these actions and needs and hear what people do not say. This makes Activity Theory useful in constructing research questions for user research and in guiding the analysis for identifying personas.

How we used Activity Theory to make personas for Performance Platform

For the Performance Platform rediscovery project, we looked at a number of elements based on the framework. These were:

  • goal and motivation: what is the person trying to achieve, and why?
  • needs: what needs does the person have in the context of motivation and goal?
  • activities/tasks: what kind of activities does the person carry out and why?
  • environment: what is the person’s physical and social context?
  • interaction/community: who are the people and communities the person is interacting with, and why?
  • tools/objects: what are the tools/objects the person uses, and why?
  • rules: are there any rules governing their decisions and actions?

I found the last element particularly interesting as it shows how people can prioritise one thing over another when they face tensions. The choices they make often reveal their motivation.

We also added additional elements, such as ‘performance’. These were either specific to our project, or were new themes that emerged during the analysis.

We then followed these steps:

  1. Analyse each individual according to this framework, using sticky notes. It’s important to arrange the framework elements vertically. This makes it easier to compare between individuals. And makes it easier to add to or move the research around.
  2. Transfer the sticky notes into a digital format. This process lets you sense-check the analysis results and spot any gaps that need more work.
  3. Identify personas considering the bigger context. We compared the similarities and differences between individuals in terms of the Activity Theory framework components, while referring back to the context of each individual. We then grouped people with the most similar patterns to form one persona.


The results

For our project, we interviewed 55 people, ranging from executive directors and analysts to open data activists. We identified 7 personas, including Anna the strategist, our primary persona.

We will iterate the Anna persona as we conduct more research with users. This is because personas should not be something static and created once. They should be living documents that reflect our increasing understanding of users.

This article comes from the User Research blog, where you can find similar guidance from the UK Government Digital Service’s research team.

About the author
Inline Feedbacks
View all comments
The Mandarin Premium

Insights & analysis that matter to you

Subscribe for only $5 a week


Get Premium Today