Fossilizing people into models. What do we gain and what do we lose?
Back in the 1980’s Apple democratized computers by integrating Xerox Parc’s intuitive GUI, which was more usable and easy to understand by the common population.
Several decades later Google Nest humanized thermostats by making it more intuitive and energy efficient.
Both of these ideas have one thing in common — “people at the center ‘’ and a careful eye to understand their pain points, needs, wants, frustrations and disappointments with the current experiences. These products were not perfect in any sense but they genuinely solved some real problems of the people. Policies, products and services without real understanding of people are doomed to fail like the millions that fail every year.
But how does one capture the requirements of hundreds of thousands of people who are inherently different from one another?
The answer is — by reducing people into models based on identifying the distinct similarities within inherent differences. There are different ways to model people based on whether we want to model the width to understand the marketplace (consumer base) or depth to understand the innate experience ( their needs, wants, goals, frustrations, aspirations) within those widths.
This idea of profiling people based on distinct similarities within differences also inspired the research of the famous FBI’s Behavioral Sciences Unit, that led to the coining of the term “serial killers” which brought a paradigm shift in crime investigation. (Please watch Mind Hunter if you haven’t already.) But such practice of profiling people had already emerged as a tool of contemporary marketing in the Mid 20th Century before the FBI adopted it. The marketers called it “Customer Segmentation” developed by Madison Avenue executives in the 1950s. These segmentations have traditionally defined the width of a marketplaces or customer base by analyzing demographic data or behavioral data. These help an organization differentiate between the types of customers that exist and understand their interests.
Why have such models evolved over time?
Gradually when the markets became complex with several options for the users to choose from, stopping at the solution to a problem was not enough anymore. Businesses felt the need to shift from providing solutions in the form of products and services to providing unique experiences, in order to stand out from the competition and for that they needed to truly understand people and their personalities across these consumer bases and segments. This shift gave rise to the term “Experience Economy” first coined in 1988 by B. Joseph Pine II and James H. Gilmore in their article in Harvard Business Review.
A little later around the mid and late 90’s researchers had begun to reflect on methods that could best communicate an understanding of the users. Since then various concepts have emerged, such as user archetypes, user models, lifestyle snapshots, need states and Jobs-To-Be-Done. Around the same time in late 90’s designers adopted the fictional but fact-based archetype to model user profiles called “Customer/User Persona” pioneered by American software designer and programmer Alan Cooper who first mentioned it in his acclaimed book “The Inmates are Running the Asylum”.
What is the purpose and what do we gain from building these models?
Such models unarguably help focus design efforts but the biggest gain from reducing people to these models is that stakeholders within organizations who have never seen a user or don’t know who their users are can understand and develop empathy for users without being directly involved in qualitative user research. The purpose is to always remind , who they are designing for and bring these specific groups of customers to life. In the book, ‘The Persona Lifecycle’, Pruitt & Adlin describe personas as “detailed descriptions of imaginary people constructed out of well-understood, highly specified data about real people.
Today User persona is a popular method that guides designers in the product development process because they help understand the expectations, concerns, and motivations of target users. It prevents teams and different stakeholders from vaguely defining the user that they are designing for and keeps them mindful that users are inherently different from themselves. Similarly there are more evolved methods which practitioners arguably consider better than personas, such as “Jobs-To-Be-Done” which focuses on outcomes based on the idea that users “hire” (adopt) a product or service, to do a specific “job” or “need state- continuum” which is based on the idea that user needs and expectations from a product changes over time continuum.
What we may lose and what are the guardrails to create authentic models and avoid personally projected models?
However, such methods which reduce people to 2-dimensional models come with their own imperfections and naturally people and teams have been suspicious about these models. To many individuals, personas seem suspiciously definitive on the pretext that the users are too large to be summed up by a handful of personas. In extreme cases of design malpractices, these models are reduced to merely a check box in the design process filled with generic information that fails to nudge a thought or conversation. Even worse, the modelers themselves can impact these models by inducing their own biases and imaginary constructs to create an ideal persona that is far from reality.
Therefore the modelers should be careful and aware of their own impacts on these models.
Another mistake that designers may make is giving a concrete state to the personas early in the process. Personas shouldn’t be definitive and finalized in the early stages of design. As design is an iterative process designers should be open to improve personas along the way as they discover new information about the users. Design processes shouldn’t entirely rely only on these models and use additional methods and tools at their disposal to inform their design processes. Experts have been pondering over questions such as — What is an effective way of creating personas and how many users are enough to model the archetypes? Some have argued that 5 personas are enough to capture 80% of users’ needs. But there are no definitive answers. A lot of it depends on the team’s research planning and own judgement based on the pre-existing data. However it can be extremely counter-productive to build such models based on limited data that may encourage teams to add their own figments of imagination to these models without being aware of these biases.
One effective way to make a useful and honest persona is to adopt evidence- based practices where direct quotes from users heard during user interviews are added to the persona. Another way to add evidence is to add video snippets from user interviews to support the archetypes (of-course these videos should be used internally.
Discussing personas with other people may help uncover biases and assumptions. Showing the personas to the people they represent may be another way to ensure if the personas are close or far from reality.
It has been evident that creating these models are extremely valuable to the product development process if done correctly. But what is that correctness? I would argue nobody knows exactly. The struggle to build these models is real because it is never easy to model the experience of someone whose race, origin, or lived experience might be totally different from that of ours. Practitioners are extremely likely to commit errors while building such models and it may never reach the stage of perfection. This may be one pragmatic reason as to why these practices have been evolving over a period of time. However we as design practitioners can follow the above mentioned protocols to remain bias aware, diligently gather qualitative data and discuss processes and assumptions with teams and other stakeholders to reduce discrepancies and recognize what is real and what isn’t, so that these models become at the least useful if not perfect.
To close this argument, I would say that such models can be holy grails of design practices if we strive to understand the truth of people at a deeper level rather than merely checking the box.
