Why you should stop using demographics in personas

In reality, demographic information often introduces bias to our decision making processes. Demographics may help us predict human behaviour that can be hugely helpful in our marketing and sales strategies when assessing market viability but rarely help us problem solve around specific needs and barriers (the role of the designer) in fact they often mask them.

To create personas, demographic detail is often gathered from existing market research, analytics tools or from real people you may have encountered (hopefully never on our assumptions alone). The reason we create personas is to:

1. Help create a lens through which we can view products or services from the perspective of real people
   
2. Build empathy
   
3. Personify needs and requirements, so we be can better understand the context in which they exist.
   

Let’s take age as an example. It’s often considered to be a barrier to adopting digital products, but age alone is never the root cause of the barrier and therefore doesn’t provide us with the insight we require to problem solve around these barriers.

Only recognising an audience’s age actually encourages us to ignore important factors and introduce our own bias towards our user-base.

When we talk about age as a relevant factor what might we mean in terms of behavioural variability? Accessibility requirements? The propensity to adopt digital products? Likelihood to engage with a brand? These are all variables that could be relevant to audiences at any age and are certainly worth investigating so let's not hide them behind unhelpful social stereotyping.

Of course, we shouldn’t throw the baby out with the bathwater. There are exceptions to this rule, in some cases age, gender, ethnicity, location or economic background, for example, may be the only lens through which we can accurately understand a user group’s goals or attitudes. But we should share this as a facet that explains their behaviour.

Using demographics alone can encourage us to shut out potential users or inaccurately prioritise expensive features.

By focusing our thinking on behaviour rather than demographics, we’re designing for what people do (or don’t do) not who they are. As designers, we’re creating a system of interactions. Understanding behaviour allows us to explore and optimise each of those interactions.

To support this, we’ve also scrapped profile photographs from our personas for the most part. Instead, we use illustrations with minimal detail to help the people using the personas with differentiation and recall. It feels refreshing to dump the stock photographs that always made me feel a bit uncomfortable anyway.

One concern I had when we started this new approach was that I was stripping out elements that I knew were traditionally included to provoke empathy for the end user.

The term empathy gets used a lot in the UX industry. You’ll find many UX practitioners keen to tell you how much empathy they have for the people they’re designing for, sometimes to the point of forming an emotional bond with each persona they create.

To some extent, I think this is a misconception of what is required of UX Practitioners and indeed what is meant by empathy. It may show you’re passionate and caring, but empathy is as much about truly understanding people (cognitive empathy) as it is about feeling their pain (affective empathy). It’s important we don’t neglect the former to satisfy the latter.

Reminding people that what they’re designing is actually going to be used by real people is no longer the responsibility of personas alone, we have popularised a whole toolkit of different methods and outputs, from Empathy Mapping to Bodystorming and everything in between.

Despite a fair amount of confusion around the value of personas, I believe they’re still a crucial tool in helping us communicate and develop a shared understanding of who we’re designing for but we need to ensure they remain helpful and reflect the best practise we’re developing together.

By dumping the demographics you might find you’re able to surface far more compelling information about user behaviour, or maybe you’ll be left with some important questions you need to answer to truly understand your users. Either way you’ve learnt something important that was previously hiding behind demographic detail.