FROM
TARGETING
TO BELONGING. — A Target Isn’t an Audience

In the bit, a Professor of Logic walks a neighbour through a chain of deductions. You have a doghouse, therefore you have a dog. You have a dog, therefore you have a family. You have a family, therefore you are a heterosexual male. It is a closed loop of archetypes that feels like a science until it meets a man who doesn’t own a doghouse. The logical conclusion? He must be gay.

The joke lands because Macdonald understood exactly what he was doing with it. The Professor is not a genius with a method. He is a bigot with assumptions. The punchline is not at the expense of the man without a doghouse. It is at the expense of a system of thinking so rigid, so committed to its own chain of inference, that it produces a slur where it intended to produce a conclusion. The logic was always the problem. The prejudice was what happened when the logic ran out of road.

We call it a persona when it’s trying to be commercially useful. We call it a stereotype when it isn’t. The cognitive operation is identical: a series of if/then leaps used to manufacture a fictional human. Meet “Marketing Mike” — 34, urban, values authenticity, drinks craft beer, makes considered purchase decisions and reads long-form on Sunday mornings.

Meet Marketing Mike. Statistically useful. Commercially inert. He lives in every brand brief and nowhere in real life.

He is statistically useful yet commercially inert. A high-resolution ghost.

This is the foundation most brand strategy is built on. And the foundation has a crack in it that the industry has been carefully stepping around for twenty years.

What Personas Actually Do

Before going further: personas are not useless. They help marketing teams align around a shared mental model of who they’re trying to talk to. They improve internal coherence. They give strategists, creatives and media buyers a common frame of reference. At their best, they get you to the right neighbourhood — the right general territory of who you’re trying to reach and what they care about.

All of that is valuable. But too many teams treat the neighbourhood as the destination. The persona becomes the brief and that gap between where the persona leaves you and where cultural participation actually begins is where most campaigns quietly die.

Personas were developed — and are still primarily deployed — to help brands direct messages more accurately. The concept traces back to Alan Cooper’s work in software design, later adopted wholesale by marketing departments as a segmentation and targeting tool. Better segmentation meant tighter targeting.

The problem is that targeting and activation are not the same thing. You can target someone with surgical precision and still leave no trace in their life. Understanding who someone is does not tell you what they will do, what they will share or what they will show up for.

That distinction is what most personas can’t close. And in an environment where the most valuable outcome is participation, that gap matters enormously.

That being said, modern personas aren’t Marketing Mike on a sticky note. The best current practices build dynamic segments that update as people move. That’s genuinely more sophisticated than a workshop archetype and the difference between a 2005 persona deck and a 2026 data-fed cohort model is real. But sophistication at the individual level doesn’t solve the structural problem. Even the highest-resolution profile of a million individuals still can’t tell you what those people will do together — because community behaviour is emergent. It isn’t an aggregation of individual attributes. It’s what happens in the space between people when shared identity gives them permission to act. No individual model, however dynamic, was designed to map that space.

The Structural Problem

The deeper failure of personas is that they flatten reality into digestible fiction. The more a person is simplified, the less predictive they become. Chapman and Milham’s landmark critique demonstrated that personas cannot be adequately verified or falsified and therefore have no demonstrable validity as a scientific method. Subsequent empirical work confirmed this across datasets of up to 10,000 respondents: persona-like descriptions with typical attribute counts matched vanishingly few real individuals. Salminen et al. later identified a compounding problem: traditionally constructed personas suffer from a “file drawer effect” — the underlying customer behaviour changes over time but the manually collected data does not change to reflect it.

A persona useful enough to brief from is, by definition, too simplified to explain how people actually behave — where shared values, shared humour and shared grievances drive decisions in ways no demographic bracket could ever capture.

David Beckham. Born 1975. Outer London. Global fame by 25. The most brand-safe man in commercial history. Photo: Wikimedia Commons, CC BY 3.0

Russell Brand. Born 1975. Outer London. Global fame by 25. A lightning rod for institutional distrust. Photo: Cambridge Union Society / Wikimedia Commons, CC BY-SA

David Beckham and Russell Brand are both British men born in 1975, raised in outer London, globally famous by their mid-twenties. Same demographic bracket, same era of cultural dominance. One became the most brand-safe man in commercial history. The other became a lightning rod for institutional distrust. A persona can’t tell you which is which — or more importantly, which community would carry your campaign forward and which would burn it down.

Brand himself makes the deeper case. In roughly fifteen years, he moved from Hollywood leading man to left-wing political activist to anti-establishment YouTube polemicist to baptised Christian aligned with an increasingly conspiratorial fringe-right audience. Same person. Same demographic data point. At least four entirely different communities, each with its own cultural logic, its own values, its own sense of who belongs and who doesn’t. A persona built in 2008 would have described a charismatic entertainer. By 2024, it would have been describing someone his 2008 audience barely recognised and in many cases actively rejected.

Beckham’s identity has shifted too. Footballer to fashion icon to UNICEF ambassador to Netflix documentary subject to Inter Miami co-owner. But every move was within brand logic. Each transition was commercially legible, strategically managed and persona-friendly. The model could track him because he moved inside the system. Brand moved outside it. That’s the distinction a persona can never make: it only works when people cooperate with it.

No research cycle is fast enough to track a public figure, let alone an ordinary person, moving between identities at that speed. And Brand is just the visible version of something happening everywhere. In a post-institutional, post-religious, internet-era landscape, subculture is more accessible than at any point in history. The barriers to entry that once constrained community participation — geography, cost, social gatekeeping — have collapsed. You don’t need to live near a scene to be part of it. You don’t need to be invited. You just need a phone and an opinion.

The result is that people now belong to multiple communities simultaneously, often ones that would have been mutually exclusive a generation ago. Someone can be part of a fitness community, a cryptocurrency community, a parenting forum and a political movement — all before lunch. They carry different identities into each, present different versions of themselves and respond to different cultural signals depending on which room they’re in. Society is more fractured, more layered and more fluid than any demographic model was designed to describe.

And then there are the identities personas can’t see at all. The ones people carry privately. The communities they lurk in but never post to, the political position they hold but don’t voice at work, the fandom they’d never admit to in a client meeting. The senior strategist who runs an anonymous meme account with 40,000 followers. The finance director who spends every evening in a gaming community her colleagues don’t know exists. The middle manager whose most passionate identity, the one that drives his purchasing decisions, his content consumption and his sense of belonging, is one he’s never once mentioned in any survey. These hidden identities are often the ones that drive the strongest behavioural signals. The anonymous purchase, the private share, the guilty-pleasure subscription. A persona can only model what people are willing to declare. It has no mechanism for capturing the identities people protect and those protected identities are frequently the ones that matter most.

The core problem is that personas are structurally incapable of reflecting the complexity of modern identity. They were built for a world where people had one job, one community, one postcode and one set of values that a researcher could observe and record. That world is gone. Demographics describe people. They do not explain what people do together.

The Real Shift: From Profiles to Participation

The internet didn’t just change how people consume content. It changed how people organise themselves.

Before algorithmic networks, community formation was constrained by geography and access. You were part of the communities available to you — your neighbourhood, your school, your workplace. Marketers could approximate these with demographics because geography and demographics were genuinely correlated.

Now, people self-organise globally around shared values, shared references, shared aesthetics and shared enemies. Muniz and O’Guinn’s foundational research identified this phenomenon as early as 2001 — brand communities that are “specialised, non-geographically bound,” organised around shared consciousness, rituals and a sense of moral responsibility. Henry Jenkins later described the broader cultural shift as “participatory culture”. Environments with low barriers to expression, strong support for sharing and a felt sense that contributions matter.

A teenager in Sheffield and a teenager in São Paulo might inhabit the same cultural world — the same fandoms, the same in-jokes, the same sense of who is in and who is out — while having almost nothing in common with their next-door neighbours.

This is identity-based community, which is not captured by a persona. It’s not even close to what a persona was designed to describe.

And this is where the structural crack becomes a strategic failure. Because these identity communities don’t just define how people see themselves. They define what those people participate in, share, amplify and defend.

These aren’t the same thing and collapsing them costs precision at every stage:

The persona was built to serve the first. It was never designed for the third — and it mistakes the second for both.

Two Campaigns. Zero Personas.

The evidence for this isn’t theoretical. Two campaigns, examined in depth in this journal, demonstrate what activating identity actually looks like and what distinguishes it from anything a customer profile could have produced.

Subway × Community — Brand as load-bearing structure

The Community × Subway integration worked because it didn’t target a specific persona. Dan Harmon, who knew his viewers with the granularity of someone who’d built a community around a specific kind of intelligence and humour, embedded Subway into the logic of the show so completely that removing the brand would have broken the plot. No persona predicted that. No demographic profile could have described it with enough precision to brief the creative work that followed.

What worked was understanding the cultural logic of the people inside the room. What they found funny, where the line was, what a betrayal of the aesthetic would feel like, then building from inside that logic rather than alongside it.

Subway as a character worked because the joke was built from Subway’s actual identity: franchised humanity, manufactured warmth, the sincere effort to seem like a person. That specificity comes from someone who genuinely understands the community they’re working within. Read the full analysis →

KitKat — Permission is structural, not reactive

In early 2026, 413,000 KitKat bars were stolen in transit from Italy. Within 48 hours, it was the most joyful brand story on the internet.

KitKat’s social team posted. The tone was well-judged. But the response was almost entirely beside the point.

Every piece written about this story credited the marketing response. The tone, the timing, the social team’s instinct to “lean in”. That sequence is accurate, but it was never the point. The conditions for that moment existed before the brand acted. The cultural permission, the implicit social licence that allows millions of people to find something funny and share it without calculating the social cost, was structural. No victim or moral complexity. No secondary harm potential. A corporation that could absorb the loss, a product enjoyed by millions and a heist so audacious it invited reluctant admiration.

The social team met the moment. They didn’t create it. But no customer profile in the world tells you how to engineer those conditions in advance. Read the full analysis →

Neither of these moments were built from customer profiles. They worked because they aligned with identity and they gave people something to participate in.

These moments also couldn’t have happened without audience understanding. Harmon knew his viewers with forensic intimacy. KitKat’s social team read the existing cultural permission before they posted. The difference isn’t data versus no data. It’s where the understanding came from — knowledge built inside the community’s own logic. That distinction separates alignment from activation.

The Hypocrisy Worth Naming

Here is the thing the industry doesn’t say out loud.

We live in a cultural moment where stereotyping is — correctly — understood as reductive, harmful and intellectually lazy. The idea that a person’s demographic profile predicts their values, behaviour or character is widely rejected in public discourse. Academic criticism of personas has flagged this tension explicitly: gendered and racial depictions in persona profiles risk reinforcing biased viewpoints rather than reflecting genuine consumer behaviour.

And then we go to work. Where we build customer profiles based on demographic aggregation, name them, dress them in stock photo avatars and use them to make multi-million pound decisions about how to speak to human beings.

The cognitive operation is the same. We’ve simply agreed on the contexts that make it acceptable.

The defence is usually ‘but it’s based on data.’ And broadly speaking, so is stereotyping — it’s rationalised through observed patterns used to predict individuals. Personas do exactly the same thing. The only real difference is framing.

The structural argument against personas is that the very simplification that makes them useful for internal alignment makes it useless for understanding the community dynamics that actually drive participation. You can’t have it both ways.

The Architecture That Closes the Gap

Customer profiles were built for a world where attention was scarce, media was linear and the goal was directing a message accurately to the right person. In that world, the model was fit for purpose.

Today, attention is abundant. What’s scarce is participation. The moment when someone not only sees a brand but chooses to engage with it, share it, carry it forward, make it part of their identity. Research on brand community identification consistently shows that it is this sense of belonging that predicts both loyalty and the willingness to advocate publicly.

That’s a completely different problem. And it requires a completely different model.

The Real LOL architecture starts with a real community that already has its own cultural logic. Its own sense of what’s funny, what’s theirs and what a brand would have to do to belong.

This doesn’t mean abandoning data. Rather, it means demanding better data. Modern tools such as clustering algorithms and semantic analysis of forums and social platforms can help to surface emergent identity communities with a precision that persona workshops never could. But the most powerful signals often come from the figureheads and native voices already living inside those communities. Influencers and creators embody and shape community culture. In doing so, they reveal the hidden permission structures, humour codes and unspoken rules of the group in real time. The architecture rejects the idea that a fictional composite is the highest resolution you can achieve. The highest resolution is the living, breathing dialogue of the community itself.

Creators and influencers also do something personas only attempt. They deliver the right audience. The right identity group, beyond demographic brackets. A creator embedded in a community doesn’t approximate cultural fit, they help to define it. The audience they bring is already self-selected and exists before the campaign does, so distribution comes with it.

The Reframe

The argument here is that personas are often solving the wrong problem.

If your goal is simply to shout into a void with slightly better aim, personas help you do that. But in a fragmented attention economy, accurate targeting is not the same as actual impact. The problem with a persona-led approach is that it assumes a message has inherent value. Without the architecture of identity, a campaign is just targeted noise. It produces an event that requires post-rationalised analysis to define success, rather than a moment that survives in the culture for free.

The industry built its entire strategic infrastructure around the question: who are we talking to?

These aren’t the same question — and they don’t produce the same answers. Conflating them is how campaigns get measured by reach when what was lost was meaning.

The answer isn’t found in research alone. It’s found in architecture. And it all starts with identity.

Marketing Mike doesn’t exist. Even if he did, his “enjoyment of craft beer” is an isolated data point that tells you almost nothing about what he does next. The key isn’t Mike; it’s the Subreddit he moderates and the Discord server where he shares with his friends. The persona sees a 34-year-old urbanite; the reality is a network of ten thousand people who just shared your campaign (or tore it apart) because you either spoke their language or felt like an intruder. The community that carries your story forward long after the spend stops exists. They were never in the brief, but they are the only reason the “Mikes” of the world ever show up.