As part of the IPA Excellence Diploma in Brands, Dan Holt explains why in a world of abundant media and AI-mediated choice, passive visibility decays quickly. Brands without coherent human evidence risk being flattened; brands that build like bands and create scenes have an advantage.
I believe brands should build like bands: from seen to scene.
On a recent run of live shows, Florence Welch of Florence and the Machine paused before Dog Days Are Over. Facing a sea of phones, she said: “Everyone, I don’t want to see a single phone in this place.” Phones disappeared. The arena erupted. Someone filmed it anyway. The clip spread, “not a phone in sight, total 90s energy.”
Beneath that reaction is nostalgia for a less mediated kind of togetherness, when the moment came before the recording. Florence was restoring the right priority: the moment made the meaning, the machine helped it travel. Too often, brands reverse it.
Three ideas run through this essay. The band is what the brand does and stands for. The scene is the behaviours people build around it. The machine is the platforms, algorithms and AI that identify, connect, surface and scale meaning. The danger comes when brands design for what the machine can serve, filter and optimise, rather than what people can join, repeat or care about. That is how the proxy replaces the person. The machine can help meaning travel, but it should not be asked to create the meaning itself.
My first experience of marketing was for my band: making posters, selling tickets, touring, getting radio play and coverage by NME. We learned which songs people sang back and how a gig could turn strangers into a crowd. I came up through the MySpace era, when digital connected scenes rather than replacing them. Music had already lived through what brands are now sleepwalking into: distribution abundance, collapsing gatekeepers, mistaking visibility for value and attention for attachment.
That mistake now sits at the heart of modern marketing. Brands have become brilliant at being distributed and weaker at being desired. The importance people place on connection with brands has been rising year on year, yet 80% of brands could disappear tomorrow and people would not care (Havas, 2025). Nearly half of UK consumers say brands “don’t understand them as people” (SAP, 2026). That gap is the opportunity: people still want connection, but brands are failing to earn it.
A good scene works like music: call-and-response, shared rhythm and memory. In my primary research, 60% agreed that the brand experiences they enjoy most are the ones they actively take part in, not passively scroll past (Author, 2026a). People remember what they do more than what they view. Memory is built through action, sensation and context, not just information (Roberts et al., 2022; Barsalou, 2008).
This is where the machine has distorted the model. Modern marketing has been engineered around what can be served, tracked and optimised. Creativity flattens to fit the feed. Brands become fluent in the systems that distribute them and less distinctive in people’s lives. The systems are competent, but as musician Jacob Collier has said, “competence has never made me cry.”
Scenes matter more in the AI era. If people remember what they do, machines retrieve the trace: what people review, recommend, discuss and repeat. Brands without lived human evidence risk being flattened into category averages. Reddit accounts for nearly half of Perplexity’s top citations, while brand-owned domains account for less than 1% of ChatGPT citations (Profound, 2026). Earned presence is the strongest predictor of AI visibility, with third-party advocacy outperforming owned content. A disconnect between what a brand says and what people say creates the biggest negative impact on visibility (Roebuck, 2026). What people say around the brand matters more than what the brand says about itself. Participation now matters twice: first for people, because doing builds memory; then for machines, because doing leaves evidence. The same behaviour serves both.
By scene, I do not only mean visible fan culture or subcultures. A scene is any repeatable context where people participate in the brand with others, through rituals, roles, shared language, places or visible behaviours. Some scenes are loud, like gigs and festivals. Others are quiet, like a family food ritual, a run club, a WhatsApp group or a post-night-out McDonald’s order.
Digital should be a bridge, not the destination. It can make a scene visible, but meaning comes from what people do inside it. Not every brand needs a cultural fandom, but every growing brand needs something people can join, repeat or pass on.
Therefore: create the band, grow the scene, then the machine can scale the evidence without hollowing
out the source.
1. Create the Band
A band begins with taste. Taste as judgement: what does and doesn’t belong.
Bands rarely start by trying to please the market. They start with something they need to express: a sound, a frustration, a way of dressing, a group of friends deciding the world does not quite fit them, so they will make something that does.
The consumer didn’t want Jimi Hendrix, but they got him, and it changed the world. They didn’t want the Sex Pistols, but they got it. F**k the customer. The customer doesn’t know what he wants. You give it to him, and he likes it.
The point is not to ignore people. It is to stop confusing research validation with imagination. Data is essential, but it often describes what people already do, recognise or reject. It cannot always invent what people will love next. When every decision is made by looking backwards, brands become the easy-listening acts of the marketing world: efficient, familiar, dead behind the eyes. The KLF, a British electronic duo, made the same point in The Manual: formulas can be written down, but spirit cannot (Drummond and Cauty, 1988).
Taste gives a brand cultural edge. It is not just aesthetic; it shows up in product, people, collaborations and what a brand chooses not to do. Taste is social, a way people position themselves and recognise each other (Bourdieu, 1984). In brand terms, taste signals what kind of world a brand belongs to, and who might feel at home there.
Scenes often start at the edges. K-pop and Afrobeats were carried by fans translating lyrics, sharing clips, building codes and turning participation into distribution. The edges eventually became stadium-filling fame.
It works the same way for brands: growth is found by identifying emerging energy before it has been averaged into the mainstream (Holt and Cameron, 2010). Burberry has demonstrated this successfully.
By entering smaller scenes through platforms, from Grime Gran to Sylvanian Families and miniatureworld TikTok, it used sharper codes to find participation and less competition. Its retail sales grew 3%, with Gen Z growth accelerating (Burberry plc, 2026). Distinctive cultural codes can create attention, participation and relevance instead of generic luxury messaging, echoing Lysette Jones’s “Fragmented Fame”: using niches and adjacent communities to help brands break through when mass fame is harder to buy (Jones, IPA Excellence Diploma essay 2023).
If taste gives judgement, positioning gives definition. A good position is not just a claim. It is a refusal.
Punk is a music genre, but its force comes from rejection: polish, excess and the establishment. It tells people what it stands for, what it stands against and what kind of identity they can step into. Most bands are comfortable not being for everyone. Most brands are terrified of it. That fear creates blandness. A brand unwilling to be wrong for some will rarely be deeply right for anyone. Real positioning sharpens choice. It gives people a reason to recognise themselves in the brand and understand who it is not for.
Lush shows how refusal can become part of a brand world: product, smell, shop theatre and activism either pull people in or push them away, and that clarity is part of its commercial power.
If taste gives judgement and positioning gives definition, coherence creates distinctiveness.
Consistency repeats assets. Coherence builds a world. Charli XCX’s Brat shows the principle in a modern media environment. It was not just an album. It became a world of colour, typography, language, attitude, collaboration and fan performance. People could enter it, signal it, remix it and wear it (Vogue Business, 2024).
Coherence becomes mythology when it gives people something to learn and recognise. Bands create universes with characters, places and accumulated stories. Brands feel interchangeable when they have not built the texture that lets people step inside. Competitors can copy a product or colour palette. They cannot easily copy the accumulated weight of stories around a brand (Holt and Cameron, 2010).
That is the difference between an asset system and a living brand world. Jeremy Bullmore’s “scraps and straws” idea applies here. People do not build brand memory from one campaign. They assemble it from fragments over time (Bullmore, 1996).
No single asset carries the brand. Every encounter does. A collaboration, a complaint handled well, a shop’s scent, a creator, an unboxing. Coherence is what makes those fragments add up.
Taste, positioning and coherence were once treated as creative virtues. Now they are growth infrastructure. They make the brand easier to recognise, easier to talk about and easier to retrieve. Without them, there is nothing strong enough for a scene to gather around.
2. Grow the Scene
A scene is defined by activity rather than passive viewing. Belief and affinity are created when you have a role, when you’re co-creating the value.
A strong band creates something to gather around. A scene gives them something to do.
People can move between scenes fluidly, but the task for brands is the same: earn a role in the moments, rituals and spaces people care about. Activity matters more than visibility. Scenes turn audiences into participants. They give people roles, rituals, language, places and behaviours they can repeat.
Music lives beyond the song. It lives in anticipation, attendance, clothing and the moment after the lights come up, when experience becomes memory. The same song means more when sung by thousands because music is embodied and collective: rhythm, voice, movement and shared emotion.
People understand themselves through groups, and shared music engages reward and bonding systems. That is why a song sung together feels different from a song heard alone (Barsalou, 2008; Collins, 2004; Tajfel and Turner, 1979; Greenberg et al., 2021).
The Grateful Dead understood this. Every show was different, so the value began in being there. Fans traded tapes afterwards, becoming archivists, distributors and evangelists. The recording served the scene rather than replacing it. Fred again.. provides a modern version. Phone restrictions protected the show from becoming content first; the Dropbox collaboration gave fans material to share, revisit and remix afterwards (Dropbox, 2026). Content served the live experience instead of replacing it.
The five scene roles
Scenes have structure. Five roles hold them together:
| Scene role | Band example | Brand equivalent | Brand job |
| The band | Artists, songs, albums | Products, founders, creative direction | Define taste, positioning and coherence |
| Tastemakers | DJs, critics, labels, bookers | Editors, retailers, creators, experts | Transfer credibility and associations |
| Venues | Clubs, festivals, record shops | Stores, events, platforms, group chats | Create places to participate |
| Fans | Gig-goers, top listeners, forums | Advocates, community members | Repeat and spread rituals |
| Rivals | Opposing genres, conventions, establishments | Category norms, competitors | Sharpen what the brand stands against |
Figure 1: Five roles in a scene
The model is diagnostic: no band, no difference; no tastemakers, no credibility; no venues, no participation; no fans, no repeatable behaviour; no rivals, no tension. Together, these roles turn attention into attachment, memory, advocacy and growth.
The Band creates the reason to gather
Gymshark is a garage-band version of this. Its taste came from gym culture, not generic wellness. Its positioning spoke to people who treated training as identity, not occasional exercise. Its coherence came through creators, events, expos and flagship spaces that reinforced the same world (Tacticone, 2025). Thirteen consecutive years of growth followed (FashionNetwork, 2026).
Tastemakers give the scene credibility
Grocery shopping is not naturally a scene for most people, yet Ocado’s Buy Women Built aisle shows how curation can make discovery feel more collective. The aisle launched with more than 1,000 products from 130 female-founded brands and, six months later, reported a 20% average sales uplift (Ocado Retail, 2025). Like a good record label, the curator made the long tail easier to discover.
Venues provide a setting for participation
Music has long been part of Dr. Martens’ brand DNA. Its stores, gigs and creative spaces give people somewhere to live the brand, not just see it, creating the clips and creator proof that paid media can scale. As Michael Boaler told me: “Music is part of the brand DNA. Going quiet opens a gap for a competitor to steal that important space in the minds of consumers” (Boaler, primary interview, 2026). He noted that the brand sees festival-season sales spikes without sponsoring festivals, because Dr. Martens is already part of the festival scene. A boot worn at a gig carries emotion and associations that a product advert would struggle to achieve.
Fans create advocacy by turning rituals into repeatable behaviour
McDonald’s is full of fan-made rituals: fries dipped into milkshakes, gherkin swaps, post-night-out orders and secret combinations. Sometimes scene-building means creating the stage. Sometimes it means noticing that the crowd is already singing your songs back.
Rivals give the scene tension
Scenes define themselves against something: punk against the establishment, mods against rockers, Oasis v Blur. The enemy can be a category convention, an emotional tension, a behaviour people want to reject or a cultural default that needs changing.
BrewDog is a cautionary example. Its early energy came from a clear enemy: corporate beer. Equity for Punks turned customers into participants. But when the brand’s behaviour began to contradict its anti-establishment positioning, the scene weakened. After selling a 22% stake for £213m, the brand threatened legal action against a Leeds music promoter over a bar called Draft Punk. The promoter’s reply cut through the contradiction: “They can’t own punk; that’s the whole point” (The Guardian, 2017a). By 2026, BrewDog had posted a £36.6m loss, assets were sold, with 38 bars closing and 484 jobs lost (The Guardian, 2025a; The Guardian, 2026).
It is dangerous to build a scene around rebellion and then behave like the thing you once opposed.
Pepsi’s criticised Kendall Jenner protest ad showed the opposite failure: borrowing the visual codes of a scene from the outside without earning any role within it (The Guardian, 2017b). It took the energy of protest without the risk or legitimacy.
This is scene-washing: performing the codes of a scene without earning a role inside it. It erodes the
credibility that made the brand commercially valuable. Scenes cannot be looted. They must be hosted, supported or grown.
When the five roles work together, brands generate the added values, personality and distinction that make them harder to substitute (King, 1971).
They create more than communication. They create behaviour.
3. Then the Machine
For bands, sold-out shows are a sign of demand, but selling out is the risk of losing the thing that created that demand in the first place.
Brands face the same test. The machine is not the enemy. Digital media, platforms, search and AI can all help a brand grow. But they should scale what the scene creates, not replace the scene itself.
Scenes create commercial value because participation creates memory, memory creates attachment, attachment leaves evidence, and evidence feeds retrieval by both people and machines.
Attachment makes reach work harder
Distribution is not the hard part. Attachment is. Music learned this before marketing. When streaming made recorded music abundant, value moved into live shows, merch, direct fan relationships, scarcity and belonging.
As Harry Styles put it: “Don’t confuse visibility with success” (Styles, 2022).
Brands face the same shift. Paid reach can put a brand in front of people, but it cannot make them care. Sharp’s evidence that brands grow through mental and physical availability still holds (Sharp, 2010).
Scenes do not replace mental and physical availability. They make them work harder by creating memorable behaviours, social proof and human evidence that media, platforms, search and AI can retrieve, surface and scale.
Paid media is the promoter. A good promoter helps people find the show. But no band should be built just for the promoter.
Scale still matters. The question is where it comes from. As media decentralises, reach builds through aggregate moments of participation and recommendation.
Those behaviours influence brand consideration. Across 30 UK brands in 10 categories since 2018, human evidence metrics, including buzz, brand impression, reputation and word of mouth, outperformed awareness metrics in their relationship with consideration: 49% versus 29%, a 1.7x lift overall, rising to roughly 3x among younger audiences (Author, 2026b). This directional finding aligns with established IPA effectiveness evidence on long-term brand effects, and with IPA data showing creator-led content delivers almost twice the long-term ROI of brand-led paid social (Binet and Field, 2013; IPA, 2025). Awareness works harder when reinforced by what people think, say and share. These become KPIs in a scene-based approach.
Scenes compound like networks
A campaign launches, peaks, decays and gets replaced. A scene accumulates. People come back. They bring friends. They repeat the ritual, teach the habit and pass on the story. The brand becomes easier to remember because it has become something people do together.
Commercial value follows when paid media amplifies evidence that people already care. IPA effectiveness evidence shows paid, owned and earned media work hardest together, with owned and earned activity increasing the effectiveness of paid campaigns (Binet and Field, 2017). Scenes are how long-term brand building happens in practice: short-term activations spike and fade, while repeated behaviours, shared rituals and accumulated evidence build over time.
Brand community research points in the same direction, linking community membership to stronger word of mouth and purchase intent (Muñiz and O’Guinn, 2001; Algesheimer et al., 2005). Bain’s analysis strengthens the point: Net Promoter Score leaders outgrew competitors by more than 2x on average (Bain & Company). Advocacy is a growth mechanism.
Scale while protecting what people care about first
Once something starts working, marketers often flatten it for scale. They remove the tension, simplify the codes, generalise the audience and polish away the edges. In doing so, they destroy the thing that made people care.
The lesson is not “stay small.” It is build deep, then go wide. Radiohead used distribution to invite participation with In Rainbows, asking fans to pay what they wanted (Swash, 2007). U2 did the opposite when Songs of Innocence appeared uninvited in people’s iTunes libraries (Hern, 2014). Maximum reach without consent can feel like a violation of the relationship.
Restaurant chain Chicken Shop had built credibility through rap, queer comedy and London Gen Z culture, then used an AI-generated K-pop band to sell K-Pop Fries. “More like chickenslop,” one customer replied. As Adam Stone, author of The Stuff You Can’t Bottle, put it to me, the machine took a role that should have belonged to the scene (Stone, 2026). The machine impersonated the scene instead of amplifying it.
This becomes commercially urgent in the AI era. If machines increasingly mediate choice, they will favour brands surrounded by coherent human evidence: reviews, recommendations, creator explanations, community discussions and usage rituals. AI is trained on human outputs, so it is susceptible to the same behavioural biases as humans, including social proof (Shotton, 2026). A brand with a living scene gives both people and machines more reasons to discover and understand it. Create the band so there is difference. Grow the scene so there is participation. Then the machine can scale the evidence.
Sold out, not sell out.
Conclusion
Too much marketing starts with the machine: what can we target, optimise and instantly attribute? The first question should be: will anyone care enough to repeat this, recommend it, return to it or make it part of their life?
Scenes answer that question because they create participation before amplification. They give people roles, rituals and places. Participation creates memory and attachment. People then leave evidence through reviews, recommendations, remixes, posts, conversations and repeated behaviour. The machine can scale that evidence, but it cannot fake its source for long.
From seen to scene. In a world of abundant media and AI-mediated choice, passive visibility decays quickly. Brands without coherent human evidence risk being flattened; brands that create scenes have an advantage.
I have seen this taken to an extreme. A fan has my band’s artwork tattooed on her back.
No brand-tracking chart has ever made the point more clearly: does anyone care enough to make this
part of who they are?
Florence asked the audience to put their phones away so the moment could happen.
She did not rage against the machine. She put it in its place.
People before machines.
Florence, then the Machine.
Dan Holt is Strategy Partner at Havas Media UK. This essay was submitted as part of the IPA Excellence Diploma in Brands.
The opinions expressed here are those of the authors and were submitted in accordance with the IPA terms and conditions regarding the uploading and contribution of content to the IPA newsletters, IPA website, or other IPA media, and should not be interpreted as representing the opinion of the IPA.
Last updated 22 September 2026