Skip navigation
Insights / 25 Sep 2020

Personas - are they needed?

Personas are used in many web projects as part of the preliminary study work. The purpose is to illustrate the target groups by making models in the form of a number of fictitious, but "real" and complete personalities. That is, to describe the U in UX (User Experience). But are personas really necessary? The answer depends on a number of factors, and I will try to clarify these.

Demographics or empathy?

To describe typical customers, you can look at data from an existing website, interview people in town or brainstorm characteristics. The individuals are then appropriately grouped into clusters with similar characteristics and needs. Now we have a first sorting to start from. When it comes to internet or e-commerce, it can be useful to look at purchase patterns (how much, how often, how quickly, product categories) or internet habits (digital native, novice, mobile-only). Of course, you can then start designing for these clusters of needs, but it turns out that the solutions will be better if you can get into the customers' situation "for real", i.e. that you empathize with the target groups even if you do not belong to the target group yourself. It is a reason to create more complete customer types, or personas.

What should a persona contain?

There are countless templates for personas online, but some typical and necessary ingredients are: - A descriptive name (Molle Mobil, Staffan Stamkund, Nördiga Nils) - Image (preferably not "stock photo") - A "story" that gives a little background and perhaps describes the persona's everyday life - Needs (immediate and longer term) - Influencers (people, situations) - Obstacles (which we need to take into account) - Economic importance (% of turnover)/purchasing patterns - Internet habits (frequency, devices, etc.) - Other important knowledge/skills (professional skills, disability)


I usually always want to come up with scenarios that describe common or expected paths through a website - regardless of whether it is an information website or an e-commerce. Where is the visitor from? What do we want him to do? What actions do we want to drive? With the persona's different glasses, you then test how it would probably go. Scenario briefings are good to do before regular user tests to avoid "dead ends" or "dark alleys" where the visitor can get stuck. Personas can also be useful when designing campaigns or buying keywords.

Forget bias

Something to watch out for when developing your personas is that you don't just use yourself and your circle of acquaintances as models. This has also proven important in other IT contexts, such as when teaching an AI for example. Personas with radically different backgrounds should be included (unless there is a good reason not to). Ethnicity and faith are two factors that can influence how a web service is perceived or functions - it can be about the visual language, captions, etc. Some "edge personas" can also be good (Angry Arne, Doubting Tilda).

So, are they needed?

It all depends on the project's level of ambition and budget. But personally, I think that personas are a good way to illustrate target groups and needs and better than statistics to have as a guide in design work. It is also useful to see solutions from many different perspectives during the development work. Personas also work independently of areas of expertise. A developer can understand the benefit of a certain function more easily if there are "living" personas to relate to, and build fantastic solutions for.

/David Aler, Strategist

More interesting news