Personas: the good, the bad and the ugly

I’ve had a mixed feelings about using personas as user centred design tool. There are many situations where they can work and many where they fall short of the mark.

Following on from the discussions on Twitter earlier today around personas I thought it was time I shared my experiences. Alex, Emma, Andy & Mark shared opinions and no doubt blog posts will follow. This has been a point of discussion before of course.

Richard Caddick, my boss at cxpartners following the conversation mentioned how as as young UXer are you supposed to think about personas? Are personas a useful tool or not?

I thought it was time to share my thoughts.

A few years ago we all agreed any form of user centred design (UCD) in a project meant better results.

Fast forward 5 years. UX is more widely understood, more widespread and as a discipline has more collective experience. The UCD toolbox is bigger meaning we can select the right tool for the problem at hand.

Personas: the good

Back in the dark days when I first started working in UX. Any kind of UCD approach yielded results. By results I mean, a measurably better final product, a smoother project role out and meeting timescales and budgets.

User research is and always has been, expensive and time consuming. Personas are a great tool for representing the user if timescales and budgets don’t allow for research.

In the design process when there is a choice the difficulties arise. Should this call to action say this or that? Should we focus the IA on services, or products or something else?

Without UCD we see personal subjective opinion rule. Decisions are justified based on personal preference and opinion. “I think we should do this…” rules; personal, subjective opinion often drives design.

Personas give the user a voice. “Anne, Persona A, would think this…” helps give design direction. The final project is more user centred. Equally if the project team have a shared view of what the user wants, constant changes are avoided thus reducing costs and time.

Personas: the bad

In my experience the above rarely happens, certainly on the projects I’ve worked on.

I’ve worked on projects with great personas. Well researched, robust, representative. It’s not the individual personas that are the problem. It’s people.

A representation of something is by its very nature is not the original. Personas are a level of abstraction removed from the real user. Personas are designed to foster a sense of empathy with the people they represent – but using them requires a sense of empathy. The ability to put oneself in the shoes of the persona. To be good at UX we need to have empathy, not everybody has the same level of empathy as us.

Personas: the ugly

I’ve seen some really badly designed personas over the years. Often they have solid demographic underpinnings. Audience segmentation, extensive attitudinal research, market sizing etc are used to form a persona.

Our audience is 20 – 30 years old women, married / co-habiting with no children, ABC1s, living in the South East, earning £25 – £50K a year. They shop online frequently and use facebook everyday. Reads the Sunday Times.

And a nice little touch from James Chudley, “has 2 kittens called salt and pepper”. Often this wrapped up with a great stock photo of someone with perfect teeth.

The problem here is how do we address design decisions with the above?

We need to add behaviour to the mix. Let’s take travel as an example. We can include research that shows the above audience books a yearly holiday 3 moths before they go away. They research the hotel for 3 hours, share it with the partner by email, they book it at lunch time. Typically they have a short list of 4 hotels and compare across 3 other websites. Here a persona can become more useful.

Arguably the above data might be enough to make decisions. Knowing that the target audience has a cat let alone it’s name doesn’t help make a decision – it helps with empathy but isn’t useful. User research will give you this information. Packaging it as a persona may reduce it’s impact. Techniques like mental modelling are far more effective as a tool to describe behaviour but that’s another blog post.

You’ve gotta have faith

I made the facetious comment on twitter that personas share a lot in common with astrology.

Astrology like personas lacks a scientific basis, the month of your birth has an impact on who you are. Astrology looks to get a deeper understanding of people, what makes them tick, what makes them behave in a certain way, a way of predicting what will happen in the future. Sound familiar? Astrology like personas relies on a sense of empathy to succeed.

The very problem that personas are trying to solve – a lack of empathy with a user group – relies itself on empathy to succeed. Personas it seems, by design have a fundamental flaw.

But then hey, I’m an Aries, I would say that.

It’d be great to hear other people’s experiences either below or on Twitter, @mrjoe