Many of the organisations that we work with today understand the importance of design research and, for a handful, it’s now integral to their business. Much of that research centres around customers and there is no doubt that organisations who truly put a focus on the customer at the top of their priority list have a distinct edge over those who do not.
The complexity and diversity of challenges we’re asked to lead companies through have never been greater, from wholescale digital reinvention to new concepts that can help them grow. Design research plays a critical role in our work across that whole spectrum of initiatives.
As Mark Wilson covered in his recent article ‘Customer-led design won’t lead you somewhere new’ we are also very mindful of how customers are engaged, and why, at times, we have to leap ahead and ‘invent on their behalf’. There’s no ‘correct’ way to embed customers in the design process.
I have long specialised in deeply embedding customer insight into our work, and, I hope, into the mindsets of our clients. Keeping the customer visible in the design process is a key part of this work, and that generally requires the creation of some assets that represent them. In this article, I’m going to attempt to explain the key differences between three of the ways that customers are represented, and why we favour some new techniques over the oldest of them all: personas.
Personas have become a common way for organisations to signal ‘we’re listening to our customers’. It’s increasingly common for a new client to ask us to work with a set of personas that they use already, or to create them as part of our work. That’s a great sign that the value of customer informed-thinking is more broadly appreciated, but it’s also frequently a sign that there’s still a long way to go in applying it.
According to their widely-regarded creator, Alan Cooper, personas are ‘not real people, but they represent them throughout the design process.’
He goes on, ‘They are hypothetical archetypes of actual users. Although they are imaginary, they are defined with significant rigour and precision. Actually, we don’t so much ‘make up’ our personas as discover them as a byproduct of the investigation process.’
A rigid model of the customer
I didn’t want to come out swinging but personas have some significant downsides. Set aside that there is rarely enough research behind them or that even good ones can be badly misused within an organisation. The biggest issues for me are that they’re static and can lead to some unhealthy biases.
Let’s touch on biases first: many of the characteristics that typically make up personas become inadvertent signifiers for class and status. The name given to a persona can allude to ethnicity and the stock photograph chosen can easily lead to mental stereotyping. I won’t dwell on this, but it’s an issue I’ve seen many times, and it relates directly to the issue of them usually being too static.
“Here are our personas. They were done a few years ago”. Sound familiar?
Poor Sandra will always be that 22 year-old intern, who routinely binges on fast-fashion, likes to socialise with her friends and struggles to manage her finances. Forget that four years later Sandra is now the CEO of a startup, and today’s persona-equivalent Sandra has substantially different characteristics.
99 times out of 100 personas are ‘created’ — often at the start of a design process — and become frozen in time. They form a locked-in perception of a customer, and that’s risky.
Personas are frequently pitched as the ‘voice’ of the customer in the design process but it’s important to recognise that they almost always represent existing customers and their characteristics. So the more out of date they become the less they represent the people you serve today — let alone the next wave of customers you could serve.
I should also say that we often see a kind of hybrid between marketing segments and personas (‘As a Flighty Urbanite Sam represents 22% of our target demographic’) which muddies the water even more.
There are better tools than personas
I am skimming a broad subject and focusing more on the negatives here of course. Before you all reach for your shredders, personas can still serve a purpose, especially if they are invested in, refreshed and curated (the memory of “we have 50 personas” still wakes me up in the night — more on that shortly) on a regular basis.
They’re quite a blunt tool to work with, and I wouldn’t use them when grappling with complex or high-risk challenges, but they at least get a version of the customer front and centre, and that’s always a good thing.
However, for most applications there are better tools than personas, so let me cover two of the approaches we’ve used instead of personas over the last few years.
Behavioural archetypes are in many ways an evolution of personas. They add functional depth while removing the specific ‘person’ from the representation. Behavioural archetypes focus on ‘who needs or does what, when and why?’ more than ‘who is this person?’
As a result, they’re better suited to informing real-world design decisions because they more directly represent customer behaviour than personal characteristics do, and fewer of them are able to span a broader range of target users.
Instead of a 29 year old person called Terry, we might have a behavioural archetype called ‘long-term strategist’ whose age, sex, race or favourite car is irrelevant. Their behavioural characteristics are what matter: from their professional needs to the key outcomes of their work. We can therefore shape a service or a feature that meets their needs, and by doing that we will address the needs of anyone who behaves that way.
Behavioural archetypes can radically simplify customer intelligence
Let’s take a real world example that I alluded to earlier. We conducted a major innovation programme with a global organisation whose business is built on the delivery of timely and insightful data, information and analysis.
Our remit was to invent a new generation of their services that could ultimately replace a host of legacy products that generated hundreds of millions of dollars each year.
Over the years, as the organisation’s offering had grown and their activities had diversified globally, so too had their library of personas. An original set of decade-old personas had proliferated to better ‘reflect’ perceived differences in global customer behaviour. This makes sense because if you’re looking at the characteristics of a typical buyer in Hong Kong, they are — in theory — quite different to their UK equivalent.
Yes, you’ve guessed it: this was when I was introduced to their library of 50 personas. There was clearly no way that anything can be designed intelligently against fifty target user personas.
Over the course of the next few weeks, we were able to synthesise and refresh this unusable library into six robust behavioural archetypes.
We excluded gender, age, roles, language and organisation size or value, refocusing everything on behavioural needs and motivations that were actually relevant. Each archetype has far more actionable detail than the original personas with none of the mental model rigidity that attaching a person to them would bring.
The results were transformative. Geographic infighting stopped, with teams swiftly adopting the new, radically simplified customer framework. The six archetypes allowed us to drive the pace of innovation and ultimately product delivery much faster, breaking through a host of previously ‘immutable truths’ of the business along the way and helping unlock much more progressive thinking for the new generation services.
The final materials proved to be a valuable asset for teams far beyond their product and innovation teams, bringing a deeper understanding of tangible customer behaviours into areas like marketing, sales and customer service.
A note on names.
Naming behavioural archetypes well is more important, and harder, than you think. As I’ve already covered, archetypes should steer clear of humanising names (sorry Drew, Fonda and Mo, you’re out) and focus on functional naming. This sounds illogical for something that is intended to bring people closer to the people they serve, but it’s an important factor in how people engage with them.
From experience, purpose-orientated names are really effective (our long-term strategist from earlier, for example) because you’re trying to create an identifier that represents behavioural characteristics easily. Think about what is the core purpose of each archetype you’ve identified is and try to capture that in the name you use. It sounds like a small point, but it really is worth spending time to get the names right.
We’ve found that behavioural archetypes are a much more sophisticated way to represent target customers, and yet are much easier to understand. They’re a genuinely powerful tool for making strategic decisions about the direction a product or services should take.
Ultimately, behavioural archetypes focus on characterising customer traits in a more useful way than personas can. But people — customers — also exhibit a lower level of more fluid behavioural states that we call behavioural modes.
In a way, behavioural modes are a lower-level build on behavioural archetypes — again, they avoid any hint of a face, a name, or a personal backstory — but they are designed to help design teams empathise with and reflect differing emotional states in their services.
Modes dig deeper into how any given user’s current mindset emotional state might affect how they use and respond to a service.
One person can, and will, exhibit many modes
Modes acknowledge that any given individual’s behaviour — through changing circumstance, confidence, mood, etc. — can change how they will perform the same activity.
Modes are fluid, but research can help us uncover the range of modes a service needs to support so that it can support each customer in whatever mode they are in at that time.
It’s not uncommon to see an individual start a process in one mode (e.g. confused), shift into a different one (anxious) before flipping into yet another (relieved) because of how the experience unfolds, while another goes through the exact opposite.
People experience the same service in different ways based on the mode they’re in
Identifying, articulating and designing for modes is incredibly useful for organisations who provide a service to customers with a wide variety of age, experience or needs, or where the nature of the service itself is likely to cause or support different emotional states.
Let’s use a quite serious example that we can all equate to: experiencing a worrying medical symptom and needing to get a diagnosis. Let’s assume two people are using a virtual diagnosis system provided by the NHS.
Person A finds a lump in their leg. They have a family history of benign cysts, so, even though their condition could be really serious, their mode is quite calm and confident because they expect it not to be. They fly through the symptom checker and get an appointment for a biopsy in 10 minutes, switch off and go read a book.
Here, the service needs to support their confidence, get out of their way, and make the process fast and simple.
Person B has a recurring pain in their chest. They don’t know their family history, but they know this could be heart-related, so their mode is anxious, perhaps plain terrified. They want to understand every step they’re taking, understand the implications of everything and be comfortable with the end outcome: an appointment for a scan. Two hours later, they’ve been back and forth, in and out of supporting content and videos and have had a dozen questions answered.
The same service has had to support their lack of confidence and their extreme anxiety: trying to rush them through like a supermarket checkout would have been disastrous.
Person A and Person B, are, of course, the same person so this is really scenario A and B.
In scenario B above, “How will I tell my family? I could be off work for months, how am I going to pay my mortgage? Who’s going to look after my dog?” might all be rattling around their mind while they’re doing what they need to do, whereas “I need to bring the washing in” might be top of mind in scenario A.
When designing a service, we can predict the functional steps needed, like a step-by-step symptom checker and a booking system but articulating behavioural modes help us make the same core system effective across widely differing states: in this example, reassurance, trust and in-depth support when it’s needed and simplicity and efficiency when it’s not.
A deeper empathy with customers
Taking the time to explore and understand modes can really pay off: it deepens customer empathy and leads to more well-rounded design solutions. We often work with clients to identify ‘magic moments’ in their service experience, and they, more often than not, manifest as small acts of kindness or consideration that relate to understanding the mode of the customer at that time.
Behavioural modes are not a standalone tool, but as a complement to more functional archetypes they bring a richness to the design process that is hard to match.
Ultimately, it’s all about them
Of the three tools I’ve covered here, it’s pretty clear that we generally favour the latter two, because we find them more appropriate in our work. But that doesn’t mean we’ll never create another persona, and it doesn’t mean you shouldn’t either: if they are the best tool for a particular challenge use them. Making the right choice boils down to understanding the respective strengths and weaknesses of each, and how you might adapt and apply them to best fit your purpose.
All design tools are just tools, helping you figure out a solution to a defined problem. No one tool will answer or solve all of those problems, and sometimes you have to set them all aside to make the leaps you need.
In my opinion, that’s fine, because ultimately whether they’re standing next to us shaping the solution, or we’re creating something that they can’t imagine using until it’s in their hand, we’re always trying to create something that works brilliantly for customers. Any tool that helps us achieve that is a good one in my book.