I’ve been wanting to write this for a while: a critique of the widespread use of “personas” in design work, especially where it’s informed by qualitative research or what we generously call “ethnography” in business contexts.
The persona, as a design artifact, is supposed to condense research findings into a set of fictional-but-representative character sketches. The idea is that these characters help product teams, designers and business stakeholders remember and empathize with real users. In practice, though, personas are less about empathy and more about expediency. They serve a particular purpose: they make complexity legible. They are simplification tools, narrative prosthetics for those who weren’t there when the research was conducted.
Real fieldwork data, however, doesn’t easily yield tidy typologies. It produces insights that are varied, sometimes contradictory and often destabilizing to whatever mental models product owners walked in with. Good qualitative research expands the frame; it challenges the problem definition. It makes us confront uncomfortable questions. The business reaction is often some version of: “This is confusing.” Not because the findings are incoherent, but because they resist reduction to a single dominant narrative — they often undermine or complicate existing assumptions rather than affirm them.
The persona steps in at this point as a compromise artifact. It funnels complexity into a format that feels actionable. However, this actionability comes at the expense of intellectual rigour. Personas flatten difference. They displace contradiction. They foreclose possibility in the name of process.
It’s worth reflecting on the etymology of the term “persona.” From the Latin, it referred to the masks worn by actors in the theatre — an outward-facing performance designed to convey a character. But these theatrical masks themselves trace their origins back to religious rituals in ancient Greece and Rome. Initially, they were not mere devices for character portrayal but ritual instruments used by priests to channel divine presence. The mask enabled a kind of sanctioned transformation — an entry into another identity for the benefit of the collective. The mask, then, was both technology and ritual: a threshold object.
In a sense, the persona in design work functions analogously. It is part of the ritual performance of modern product development. Personas are created unthinkingly, routinely, without deep interrogation of their epistemic role or their adequacy as artifacts. They are performed more than constructed — less the product of reflective synthesis, more a necessary rite in the sequence of design stages. Like the ancient ritual mask, they are enabling fictions. But unlike the sacred, design personas rarely aspire to deeper truths. They are instead used to usher a project forward, to validate pre-existing notions under the guise of empathic insight.
In Goffman’s dramaturgical sociology, the persona is the self as presented in everyday social interaction, a construct carefully managed for particular audiences. In both classical theatre and social theory, the persona is not the person; it is a mediated representation. It conceals as much as it reveals.
When design practitioners create personas, they are doing something structurally similar. These artifacts function as masks: they present a clean, coherent front while hiding the complexity, messiness and contradiction that lie behind. They are designed for an audience — the business stakeholders who crave a kind of utilitarian clarity, a bridge to urgent “next steps” — and they deliver that clarity, but only by staging a performance that, by necessity, misrepresents the richness of the underlying data.
In practice, the situation is often even more compromised. Designers and researchers frequently import their own “tried-and-true” starting points, drawing on previous persona templates or implicit assumptions about what’s worked in the past. “We must have at least one older person.” “We need an equal number of women and men.” “Let’s make sure we have a nonbinary person, it’s 2025!” These patterns are often less about representation and more about box-ticking. They feed into the overall flattening effect. Too often, the templated personas are then adorned with stock photography — smiling, generically diverse images of beautiful people, the everymen and -women of our day. It starts to feel like buying a Halloween mask at the dollar store: something everyone uses one night a year before it ends up in landfill with the rest of the seasonal detritus.
That said, personas can occasionally serve a purpose. In the narrow domain of software requirements development, they can function as useful heuristic devices. When evaluating or refining a specific feature, it can be helpful to imagine how different positional stand-ins might interact with it. Can the older person with limited tech experience navigate the drop-down? Can someone who only accesses the internet via smartphone complete this form? Does the process take too long for a busy mom balancing multiple responsibilities? In this limited context, personas can scaffold the imagination — providing lightweight proxies that help teams consider edge cases or usability breakdowns.
Unfortunately, this constrained use is rarely the norm. More often, the interplay between “problem statement” or “market opportunity” and persona development is weakly theorized but energetically pursued. Personas have become something that must be completed, regardless of their relevance to the problem space. Design research practitioners have so thoroughly routinized the production of personas that they function like what Science and Technology Studies scholars of the 1990s called a “black box” — a once-contested instrument whose internal logic is no longer interrogated. It is assumed to work. It is accepted as canon. And that, perhaps more than anything, is the peril.
Personas have become part of the standard design lexicon. They’re expected, they travel well and they give the impression that something solid has been produced. Yet what is actually being delivered is a premature narrowing of insight. Research’s destabilizing potential is neutralized in favour of forward momentum or rote pattern-following.
This dynamic bears a strong resemblance to what Bruno Latour identifies in We Have Never Been Modern as the “purification” impulse in scientific modernity — the shearing-off of process, uncertainty and messy social entanglement to present something clean, stable and fact-like. Just as science masks its contingent, constructed nature in order to appear authoritative, so too do personas mask the contingent, plural and unresolved nature of qualitative research. The artifact becomes a rhetorical device that says: “Here is what we know,” when the more honest claim would be: “Here is what we’re still trying to understand.”
Part of what makes this so difficult to undo is the quiet pressure to productize design research itself. For internal design teams and external consulting firms alike, personas have become a pivotal deliverable — a recognizable milestone in a process that increasingly needs to be legible, repeatable and salable. They help make the process itself into a product. And this, too, discourages critical interrogation. The persona, in this context, is a sacred cow. It is not examined but celebrated. Invoked with ritual joy rather than subjected to methodological scrutiny.
What we need instead are practices and artifacts that can hold complexity without becoming paralyzing. We need representations that reflect the richness and ambiguity of real fieldwork. We need ways to resist the temptation to tidy up before the mess has taught us what it needs to teach.
That might involve composite narratives. It might mean more speculative synthesis. Or perhaps it requires live, participatory workshops in which contradictions are explored rather than resolved. I don’t have the definitive answer. But I do know this: personas, as currently deployed in most product design contexts, aren’t it.
While I may not have a formula to replace personas outright, I’ve observed in my management consulting work that something else often emerges organically. The kind of fieldwork I do — long-term discovery through numerous interviews, and simply by “being around” my client’s teams, working alongside them — ends up producing a richer, more dynamic understanding. It’s classic participant observation, and it slowly develops into something more potent than any static artifact.
Given enough time and proximity, I begin to develop a feel for the organization — its voices, its internal logics, its tensions. In design sessions, this allows me to bring those perspectives to bear in ways that are responsive and contextually informed. Not as fixed roles or scripted personas, but as impressions grounded in real interactions. I wouldn’t claim to know exactly how this works, but it seems to operate with a certain reliability.
Perhaps this too is a form of foreclosure. But if it is, it’s of a different kind — less an endpoint than a state of readiness. A cognitive, dialogical mode in which I can query an invisible knowledge base, one built from immersion, proximity and practice. It’s dynamic. It adapts.
In an odd sense, this knowledge is stored in my mind in a way that I imagine as not too dissimilar from the way LLMs use vectors to store information about the vast corpus of information they learned from. I hope that I’m less prone to hallucination and getting it wrong, though.