Personas are primarily a tool for user modeling, which means that they are an abstract representation (i.e. model) of various user groups. In that sense, they serve as a communication tool that helps design decisions. Personas have specific needs, goals and they perform certain activities. These elements derive from real research data gathered from qualitative methods, such as interviews, observations and diary studies. Therefore, they can attributed with human characteristics: real photo, name, age, profession and behaviour variables. Many people confuse personas with market segments which are mostly produced by marketing teams. Although the two tools are complementary and each one informs the other, there are some significant differences between them. In one sentence, market segments show who your audience is but not what your audience does, needs or wants. Market segments are mostly based on quantitative data that come from demographics and not from real user research. Thus, they don’t offer a lot of insights to designers. However, market segments facilitate finding the right participants to recruit for user research.
There are a lot of inspiring books (“The inmates..”, “About Face”, “Persona lifecycle” to mention some) written by persona 'evangelists' go into some deatil in how to use them and why they are so important. Like any other user modeling tool, personas allow designers to deal with friendly artefacts rather than raw research data on an affinity map. Personas are expected to incorporate all the answers needed to justify your choices. For example, if ‘Jane’ was our primary persona, we could ask “would Jane want to use a mobile app?”. The answer should lie in Jane’s description, behaviour variables and goals. Choosing different types of personas (primary, secondary and anti-persona) means that the design addresses the needs of those groups differently. The primary persona must be the most satisfied with the design.
In theory, everything sounds sensible and straightforward. Conversely, creating a persona in practice can be really tricky! Combining elements from research data into a persona is likely to raise some critical questions, such as the following:
This question sounds similar to “how many participants must be recruited for a research project” in order to have valid data and sufficient evidence for our design decisions - especially if you wanted to follow a user-centred design process. The biggest challenge when creating a persona is to avoid stereotypes and use only real data. For example, if the persona is moving around in the city by car or by public transportation, it affects the way s/he interacts with a mobile device. We must ensure that we gather all the data we might need to shape a persona’s character.
Selecting a primary persona means that the design satisfies primarily the needs of that particular user group, which is usually consisted of the typical users. But what dictates the typical users? Did findings from the market/user research reveal what most people need and want?
specifying requirements and conceptual design typically follows user modeling in the design process. But what is the role of personas at these stages? As mentioned before, personas are ideally used as a resource to refer to when design decisions must be made. Is this usually the case or does a consolidated affinity diagram still hold a primary position as a team’s point of reference? In my experience, if not everyone in the design team agrees on the value and the validity of personas, they won’t be used as they are supposed to. This also leads to the point that by no means personas should be considered an alternative to user testing.
Acknowledging the needs, the goals and the challenges of your target audience is an incomparably valuable aspect in the design process. However, choosing to accommodate one part of your target audience, e.g. the mainstream rather than the expert users, has a huge impact on design decisions. Personas are supposed to represent all the information needed to justify these decisions. Nevertheless, the fear of not creating valid and useful personas remains. Apart from that, there are also different views on how exactly personas can beused in the design process. Are they simply a deliverable to the client that classifies the different types of audiences, or are they a useful tool that UX designers actually use to compare their designs against users’ needs and goals? The truth is potentially somewhere in the middle. If you have any experience with using them (effectively or not), your comments are more than welcome. In a next post, I will try to show how personas and other subjective measurements can contribute to designing for user engagement, so watch this space!