20 de Março de 2026 · Mitos Desmentidos

O Mito da Segmentação de Precisão

Mitos Desmentidos

TL;DR

Em 2017, a P&G cortou US$200 milhoes em investimento de targeting digital. O alcance aumentou 10%. As vendas continuaram iguais. O Ehrenberg-Bass Institute provou, em 25 anos de dados e mais de 50 categorias, que os perfis de compradores de marcas concorrentes raramente diferem. A Adidas investiu 77% do budget em performance e quase destruiu seu brand equity. A ciencia e clara: segmentacao estreita sacrifica crescimento de marca. O mercado ainda faz o contrario.

US$200 milhoes jogados no targeting errado

Em 2016, a Procter & Gamble era a maior anunciante do mundo. Marc Pritchard, Chief Brand Officer da empresa, fez uma declaracao que reconfigurou a conversa sobre midia digital. A P&G tinha ido longe demais com a segmentacao de precisao. O targeting estava tao granular que os anuncios atingiam audiencias minusculas, com frequencia excessiva, enquanto milhoes de potenciais compradores ficavam fora do radar.

A empresa cortou US$200 milhoes do investimento em midia digital programatica. O resultado? O alcance cresceu 10%. As vendas se mantiveram estaveis. Nenhum impacto negativo mensuravel.

Pense nisso. A maior anunciante do planeta removeu duzentos milhoes de dolares de targeting digital e descobriu que o dinheiro estava sendo desperdicado em precisao inutil. O CFO da P&G, Jon Moeller, confirmou em call com investidores em 2018: o corte "did not have any negative impact on growth".

O que a P&G entendeu, e que boa parte do mercado ainda resiste a aceitar, e que segmentacao de precisao tem um ponto de retorno decrescente. Alem desse ponto, cada camada adicional de targeting reduz o universo de compradores potenciais sem compensar em taxa de conversao. Voce melhora a eficiencia do clique e degrada a eficacia da marca.

A crenca que o mercado defende

O argumento e sedutor. Dados permitem identificar o consumidor ideal. Plataformas digitais permitem entregar a mensagem certa, para a pessoa certa, no momento certo. Por que desperdicar dinheiro falando com quem nao vai comprar?

Essa logica domina salas de reuniao de marketing em todo o mundo. E o modelo mental padrao da era digital. E tem tres falhas estruturais que a ciencia ja documentou.

A primeira: a suposicao de que existe um "consumidor ideal" e que ele e diferente dos consumidores dos concorrentes. A segunda: a suposicao de que investir em alcance amplo e desperdicio. A terceira: a suposicao de que dados de comportamento passado predizem comportamento futuro com precisao suficiente para justificar restricoes de audiencia.

Todas as tres estao erradas. Os dados sao claros.

A lei que destrói a hipersegmentacao

O Ehrenberg-Bass Institute for Marketing Science, da Universidade do Sul da Australia, e o maior centro de pesquisa empirica sobre como marcas crescem. Andrew Ehrenberg comecou o trabalho nos anos 1960. Byron Sharp e Jenni Romaniuk continuaram nas decadas seguintes. O acervo de dados cobre mais de 50 categorias, dezenas de paises e mais de 25 anos de analise longitudinal.

Uma das descobertas mais replicadas do instituto e a Lei dos Perfis de Usuarios de Marca: os perfis demograficos e comportamentais dos compradores de marcas concorrentes dentro de uma mesma categoria raramente diferem.

50+ categorias analisadas pelo Ehrenberg-Bass em 25 anos confirmam:
perfis de compradores de marcas rivais sao praticamente identicos

O estudo de Kennedy, Ehrenberg e Long (2000), replicado por Sharp, Riebe e colegas em 2012, analisou marcas concorrentes em categorias de bens de consumo, servicos financeiros, telecomunicacoes e varejo. O resultado e consistente: compradores de Coca-Cola e compradores de Pepsi tem o mesmo perfil. Compradores de Nike e compradores de Adidas tem o mesmo perfil. Compradores de iPhone e compradores de Samsung tem o mesmo perfil.

A implicacao e devastadora para a logica da hipersegmentacao. Se os perfis de compradores sao iguais entre marcas rivais, entao o targeting baseado em perfil demografico ou comportamental nao identifica "seu" consumidor. Identifica o consumidor da categoria. E excluir partes desse universo e excluir compradores em potencial.

"The customer profiles of rival brands seldom differ. The implication is that sophisticated segmentation and tight targeting often just reduce your reach to people who would have bought anyway." Byron Sharp, Ehrenberg-Bass Institute, 2023

Os light buyers que o targeting ignora

O Ehrenberg-Bass documentou outra regularidade que reforca o argumento. A distribuicao de compras dentro de qualquer marca segue um padrao Pareto modificado: os 20% mais pesados de compradores geram cerca de 50% a 60% das vendas. O Pareto classico diria 80/20. A realidade e mais equilibrada.

Isso significa que os 80% de compradores mais leves, os light buyers, geram entre 40% e 50% das vendas de uma marca. Quase metade. E esses light buyers sao os que a segmentacao de precisao tende a excluir primeiro, porque seus sinais comportamentais sao fracos, esporadicos e dificeis de capturar por algoritmos de targeting.

O problema se agrava quando consideramos crescimento. Romaniuk e Sharp demonstraram em How Brands Grow: Part 2 (Oxford University Press, 2015) que o crescimento de marca vem predominantemente da aquisicao de novos compradores e do aumento de penetracao, e que esses novos compradores sao, quase por definicao, light buyers ou non-buyers da marca no momento anterior.

O targeting de precisao otimiza para heavy buyers existentes. Ou seja: para o segmento que ja compra, com a frequencia que ja compra. O crescimento incremental vem de outro lugar, e o targeting de precisao esta, por definicao, excluindo esse lugar.

Adidas: 77% em performance e a marca comecou a desaparecer

Em outubro de 2019, Simon Peel, entao diretor global de midia da Adidas, fez uma declaracao publica que sacudiu o mercado. A empresa tinha investido 77% do budget de marketing em performance e ativacao, contra 23% em construcao de marca. A recomendacao de Binet & Field, baseada em decadas de dados do IPA, era exatamente o inverso: 60% em marca, 40% em ativacao.

O resultado da Adidas foi o que a ciencia preveria. O excesso de performance marketing gerou vendas de curto prazo, otimizou metricas de clique e conversao, e degradou progressivamente o brand equity da empresa. Peel admitiu: "We had a problem. We focused too much on efficiency rather than effectiveness."

A Adidas nao fazia brand tracking. Sem medicao de longo prazo, os dados de curto prazo dominaram as decisoes. E os dados de curto prazo sempre favorecem performance sobre marca, porque os efeitos de marca sao lentos, cumulativos e dificeis de atribuir a uma peca especifica. Enquanto isso, cada real investido em performance gera um clique rastreavel.

A armadilha e simetrica. Quanto mais voce investe em targeting preciso e performance, mais dados de curto prazo voce gera que justificam investir ainda mais em targeting preciso e performance. A marca se degrada silenciosamente, ate que o custo de aquisicao sobe porque ninguem mais lembra da sua marca sem um anuncio pago na frente.

O que Kahneman explica sobre alcance e marca

Daniel Kahneman e Amos Tversky publicaram em 1973 o estudo seminal sobre a heuristica de disponibilidade. A conclusao: pessoas estimam a probabilidade de eventos com base na facilidade com que exemplos desses eventos vem a mente. Quanto mais facilmente voce lembra de algo, mais frequente e importante voce assume que aquilo e.

Aplicado a marcas, o mecanismo e direto. O consumidor em momento de compra nao avalia todas as opcoes disponiveis. Ele recorre ao repertorio de marcas que vem a mente com facilidade. Sharp chamou isso de mental availability, e o conceito e fundamentado pela pesquisa de Kahneman.

O que constroi mental availability? Exposicao repetida, ampla, em contextos variados. E o oposto do targeting restrito. Quando uma marca so aparece para um segmento estreito, ela deixa de construir estruturas de memoria em todos os outros consumidores da categoria. Esses consumidores, quando entram em modo de compra, nao lembram dela. Escolhem o que esta disponivel na memoria.

Tversky e Kahneman (1973) mostraram que a frequencia percebida de um evento e proporcional a facilidade de recuperacao na memoria. Para marcas, facilidade de recuperacao e funcao do numero de exposicoes, da variedade de contextos e do alcance da distribuicao de midia. Restringir o alcance e restringir a construcao de memoria. Restringir a construcao de memoria e restringir o crescimento futuro.

Binet & Field: a evidencia que o mercado ouviu e ignorou

Les Binet e Peter Field analisaram mais de 1.000 cases de eficacia do IPA (Institute of Practitioners in Advertising) em The Long and the Short of It (IPA, 2013). A conclusao central: campanhas que combinam construcao de marca com amplo alcance emocional produzem efeitos de negocio mais fortes e mais duradouros do que campanhas que focam em ativacao de curto prazo e segmentacao precisa.

A ratio otima, segundo os dados do IPA: 60% do budget em construcao de marca (emocional, amplo alcance) e 40% em ativacao (racional, targeting mais preciso). A industria opera, em media, exatamente ao contrario.

Em 2025, na IPA Effectiveness Conference, Binet atualizou o argumento. A apresentacao, intitulada "Go Big or Go Home", confrontou diretamente a obsessao da industria com eficiencia e targeting. O argumento de Binet: "our industry has become focused on efficiency, targeting, and short-term metrics at the expense of scale, reach, and true brand-building effectiveness."

60/40 A ratio otima de investimento segundo dados de 1.000+ cases do IPA:
60% construcao de marca (alcance amplo) / 40% ativacao (targeting preciso)

Os dados do IPA mostram que campanhas com alta cobertura de mercado (alcance amplo) geram, em media, 3x mais efeitos de negocios de longo prazo do que campanhas com targeting restrito. O mecanismo e coerente com Ehrenberg-Bass: alcance amplo constroi disponibilidade mental entre light buyers e non-buyers, que sao a fonte do crescimento futuro.

A neuroquimica da exposicao repetida

Robert Zajonc publicou em 1968 o estudo sobre o efeito de mera exposicao: a exposicao repetida a um estimulo, mesmo sem atencao consciente, aumenta a preferencia por esse estimulo. O efeito e robusto, replicado em centenas de estudos subsequentes, e opera abaixo do limiar de consciencia.

Antonio Damasio, em O Erro de Descartes (1994), demonstrou que decisoes, incluindo decisoes de compra, envolvem marcadores somaticos, associacoes emocionais inconscientes formadas pela experiencia. A formacao desses marcadores depende de exposicao variada em contextos diferentes.

O targeting restrito reduz o numero de exposicoes para a maioria dos consumidores da categoria. Reduz a variedade de contextos. E portanto reduz a formacao de marcadores somaticos favoraveis. O resultado: a marca perde relevancia emocional entre compradores em potencial que nunca veem seus anuncios porque o algoritmo decidiu que eles nao fazem parte do "target".

O paradoxo dos dados: quanto mais precisos, menos uteis

A industria de ad tech vendeu durante duas decadas a ideia de que dados de comportamento permitem prever quem vai comprar. A realidade, documentada por Sharp e por pesquisadores independentes, e outra.

Comportamento de compra dentro de uma categoria e estocastico. Compradores alternam entre marcas com regularidade que segue modelos probabilisticos (Dirichlet, documentado por Ehrenberg). A "lealdade" que os dados de CRM sugerem e, na maioria dos casos, uma ilusao estatistica: compradores frequentes de uma marca sao compradores frequentes da categoria, e compram outras marcas com frequencia semelhante.

Quando uma plataforma de targeting identifica um "consumidor leal" e concentra impressoes nele, esta gastando dinheiro com quem ja compraria de qualquer forma. E deixando de investir nos consumidores que poderia conquistar. O economista jargoniza isso como "subsidizing inframarginal consumers". Na pratica: voce paga para convencer quem ja esta convencido.

Mark Ritson, colunista do Marketing Week e professor de marketing, resumiu em 2019: "The more we target, the less we grow. It's not a paradox. It's just math."

O que funciona de verdade

A alternativa ao targeting de precisao extremo e o que Sharp chama de "sophisticated mass marketing". A estrategia segue tres principios:

Alcancar todos os compradores da categoria. O objetivo e maximizar o alcance entre todos os consumidores que compram ou podem vir a comprar na categoria. Segmentacao por categoria sim. Hipersegmentacao por perfil comportamental ou demografico, na maioria dos casos, nao.

Construir e refrescar estruturas de memoria. A comunicacao deve criar e manter Category Entry Points, os momentos e contextos em que o consumidor pensa na categoria. Quanto mais CEPs uma marca ocupa na memoria, maior a probabilidade de ser lembrada e comprada.

Investir em ativos distintivos, com alcance amplo. Cores, logos, sons, formatos, personagens: os distinctive assets precisam ser vistos por todos, repetidamente, para construir reconhecimento automatico. Restringir a audiencia de ativos distintivos e restringir sua eficacia.

Por que a P&G voltou ao alcance

Apos o corte de US$200 milhoes, Marc Pritchard articulou a nova estrategia como "mass reach with precision". O objetivo mudou: em vez de buscar a pessoa certa, buscar a maior cobertura possivel entre compradores da categoria, usando dados para controlar frequencia (nao para restringir audiencia). Em 2019, a P&G declarou que reach era a metrica primaria, substituindo metricas de targeting.

Jon Moeller, CFO da P&G, confirmou em julho de 2019: "A focus on reach is more important than a focus on spend. We're reaching more consumers, more often, more efficiently."

O que isso muda no planejamento

Auditoria anti-hipersegmentacao: 7 perguntas

O alvo que o mercado esta acertando

A ironia do targeting de precisao e que ele acerta. Acerta exatamente as pessoas que ja comprariam. E erra todas as pessoas que poderiam comprar pela primeira vez.

Marcas crescem quando entram na memoria de mais consumidores. Entram na memoria de mais consumidores quando sao vistas por mais consumidores. Sao vistas por mais consumidores quando o alcance e amplo. A cadeia e simples. E a cadeia esta quebrada toda vez que alguem adiciona mais uma camada de segmentacao ao plano de midia.

A P&G entendeu isso em 2016. A Adidas aprendeu da pior forma em 2019. Binet & Field documentaram com 1.000 cases. O Ehrenberg-Bass replicou em 50+ categorias. Kahneman explicou o mecanismo cognitivo.

A pergunta que resta e simples: quantos anos de evidencia sao necessarios para que uma sala de reuniao mude um PowerPoint?

Referencias

Livros

  • Sharp, Byron. How Brands Grow: What Marketers Don't Know. Oxford University Press, 2010.
  • Romaniuk, Jenni & Sharp, Byron. How Brands Grow: Part 2. Oxford University Press, 2015.
  • Binet, Les & Field, Peter. The Long and the Short of It: Balancing Short and Long-Term Marketing Strategies. IPA, 2013.
  • Kahneman, Daniel. Thinking, Fast and Slow. Farrar, Straus and Giroux, 2011.
  • Damasio, Antonio. Descartes' Error: Emotion, Reason, and the Human Brain. Putnam, 1994.
  • Ehrenberg, Andrew S.C. Repeat-Buying: Facts, Theory and Applications. Griffin, 1972.

Artigos e Estudos

  • Kennedy, R., Ehrenberg, A.S.C. & Long, S. "Competitive Brands' User Profiles Hardly Differ." Market Research Society Conference, 2000.
  • Sharp, B., Riebe, E. et al. "In 25 Years, Across 50 Categories, User Profiles for Directly Competing Brands Seldom Differ." Journal of Advertising Research, 52(2), 2012.
  • Tversky, A. & Kahneman, D. "Availability: A Heuristic for Judging Frequency and Probability." Cognitive Psychology, 5(2), 1973.
  • Zajonc, Robert. "Attitudinal Effects of Mere Exposure." Journal of Personality and Social Psychology, 9(2), 1968.
  • Binet, L. & Field, P. "Effectiveness in Context." IPA, 2018.
  • Ehrenberg, A.S.C., Uncles, M.D. & Goodhardt, G.J. "Understanding Brand Performance Measures: Using Dirichlet Benchmarks." Journal of Business Research, 57(12), 2004.

Fontes de Industria

  • AdWeek. "When Procter & Gamble Cut $200 Million in Digital Ad Spend, It Increased Its Reach 10%." Marco de 2018.
  • Marketing Week. "Adidas: We over-invested in digital advertising." Outubro de 2019.
  • Marketing Week. "P&G puts focus on reach: It's a more important measure than spend." Julho de 2019.
  • IPA. Les Binet & Will Davis. "Go Big or Go Home." IPA Effectiveness Conference, 2025.
  • Ehrenberg-Bass Institute. "The Law of Brand User Profiles." Research Briefs, 2023.
  • The Drum. "P&G's media chief doubles down on bid to better balance targeted ads with mass-reach strategies." Maio de 2017.

Diego Isaac

Estrategista de marcas com +10 anos na intersecao entre ciencia de marca, comportamento do consumidor e comunicacao.

diegoisaac.com.br
Myths Debunked

TL;DR

In 2017, P&G cut $200 million in digital targeting spend. Reach increased 10%. Sales held steady. The Ehrenberg-Bass Institute proved, across 25 years of data and 50+ categories, that buyer profiles of competing brands rarely differ. Adidas invested 77% of its budget in performance and nearly destroyed its brand equity. The science is clear: narrow targeting sacrifices brand growth. The market still does the opposite.

$200 million wasted on the wrong targeting

In 2016, Procter & Gamble was the world's largest advertiser. Marc Pritchard, the company's Chief Brand Officer, made a statement that reconfigured the conversation about digital media. P&G had gone too far with precision targeting. The targeting was so granular that ads reached tiny audiences at excessive frequency, while millions of potential buyers stayed off the radar.

The company cut $200 million from programmatic digital media investment. The result? Reach grew 10%. Sales held steady. No measurable negative impact.

Think about that. The planet's largest advertiser removed two hundred million dollars of digital targeting and found that the money had been wasted on useless precision. P&G's CFO, Jon Moeller, confirmed in a 2018 investor call: the cut "did not have any negative impact on growth."

What P&G understood, and what much of the market still resists accepting, is that precision targeting has a diminishing return threshold. Beyond that threshold, each additional layer of targeting shrinks the potential buyer universe without compensating in conversion rate. You improve click efficiency and degrade brand effectiveness.

The belief the market defends

The argument is seductive. Data can identify the ideal consumer. Digital platforms can deliver the right message, to the right person, at the right time. Why waste money talking to people who won't buy?

This logic dominates marketing boardrooms worldwide. It is the default mental model of the digital era. And it has three structural flaws that science has already documented.

The first: the assumption that an "ideal consumer" exists and that they differ from competitors' consumers. The second: the assumption that investing in broad reach is waste. The third: the assumption that past behavioral data predicts future behavior with enough precision to justify audience restrictions.

All three are wrong. The data is clear.

The law that destroys hyper-targeting

The Ehrenberg-Bass Institute for Marketing Science, at the University of South Australia, is the largest empirical research center on how brands grow. Andrew Ehrenberg started the work in the 1960s. Byron Sharp and Jenni Romaniuk continued in the following decades. The data covers more than 50 categories, dozens of countries, and over 25 years of longitudinal analysis.

One of the institute's most replicated findings is the Law of Brand User Profiles: the demographic and behavioral profiles of buyers of competing brands within the same category rarely differ.

50+ categories analyzed by Ehrenberg-Bass over 25 years confirm:
buyer profiles of rival brands are virtually identical

The study by Kennedy, Ehrenberg, and Long (2000), replicated by Sharp, Riebe, and colleagues in 2012, analyzed competing brands in consumer goods, financial services, telecommunications, and retail. The result is consistent: Coca-Cola buyers and Pepsi buyers share the same profile. Nike buyers and Adidas buyers share the same profile. iPhone buyers and Samsung buyers share the same profile.

The implication is devastating for hyper-targeting logic. If buyer profiles are identical across rival brands, then targeting based on demographic or behavioral profile doesn't identify "your" consumer. It identifies the category consumer. And excluding parts of that universe means excluding potential buyers.

"The customer profiles of rival brands seldom differ. The implication is that sophisticated segmentation and tight targeting often just reduce your reach to people who would have bought anyway." Byron Sharp, Ehrenberg-Bass Institute, 2023

The light buyers that targeting ignores

Ehrenberg-Bass documented another regularity that reinforces the argument. Purchase distribution within any brand follows a modified Pareto pattern: the heaviest 20% of buyers generate roughly 50% to 60% of sales. The classic Pareto would say 80/20. Reality is more balanced.

This means that the lightest 80% of buyers, the light buyers, generate between 40% and 50% of a brand's sales. Nearly half. And these light buyers are the first ones precision targeting tends to exclude, because their behavioral signals are weak, sporadic, and hard for targeting algorithms to capture.

The problem worsens when we consider growth. Romaniuk and Sharp demonstrated in How Brands Grow: Part 2 (Oxford University Press, 2015) that brand growth comes predominantly from acquiring new buyers and increasing penetration, and that these new buyers are, almost by definition, light buyers or non-buyers of the brand at the prior moment.

Precision targeting optimizes for existing heavy buyers. In other words: for the segment that already buys, at the frequency it already buys. Incremental growth comes from elsewhere, and precision targeting is, by definition, excluding that elsewhere.

Adidas: 77% in performance and the brand started to disappear

In October 2019, Simon Peel, then Adidas's global media director, made a public statement that shook the market. The company had invested 77% of its marketing budget in performance and activation, versus 23% in brand building. The recommendation from Binet & Field, based on decades of IPA data, was exactly the reverse: 60% brand, 40% activation.

The Adidas outcome was what science would predict. The excess of performance marketing generated short-term sales, optimized click and conversion metrics, and progressively degraded the company's brand equity. Peel admitted: "We had a problem. We focused too much on efficiency rather than effectiveness."

Adidas wasn't doing brand tracking. Without long-term measurement, short-term data dominated decisions. And short-term data always favors performance over brand, because brand effects are slow, cumulative, and hard to attribute to a specific piece. Meanwhile, every dollar invested in performance generates a trackable click.

The trap is symmetrical. The more you invest in precise targeting and performance, the more short-term data you generate that justifies investing even more in precise targeting and performance. The brand degrades silently, until acquisition cost rises because nobody remembers your brand without a paid ad in front of them.

What Kahneman explains about reach and brand

Daniel Kahneman and Amos Tversky published their seminal study on the availability heuristic in 1973. The conclusion: people estimate the probability of events based on how easily examples of those events come to mind. The more easily you remember something, the more frequent and important you assume it is.

Applied to brands, the mechanism is direct. The consumer at the moment of purchase doesn't evaluate all available options. They draw from the repertoire of brands that come to mind easily. Sharp called this mental availability, and the concept is grounded in Kahneman's research.

What builds mental availability? Repeated, broad exposure across varied contexts. The opposite of restricted targeting. When a brand only appears to a narrow segment, it stops building memory structures in all other category consumers. Those consumers, when they enter buying mode, don't remember it. They choose what's available in memory.

Tversky and Kahneman (1973) showed that perceived frequency of an event is proportional to the ease of memory retrieval. For brands, ease of retrieval is a function of exposure count, context variety, and media distribution reach. Restricting reach restricts memory building. Restricting memory building restricts future growth.

Binet & Field: the evidence the market heard and ignored

Les Binet and Peter Field analyzed over 1,000 effectiveness cases from the IPA (Institute of Practitioners in Advertising) in The Long and the Short of It (IPA, 2013). The central conclusion: campaigns combining brand building with broad emotional reach produce stronger and more lasting business effects than campaigns focused on short-term activation and precise targeting.

The optimal ratio, according to IPA data: 60% of budget on brand building (emotional, broad reach) and 40% on activation (rational, more precise targeting). The industry operates, on average, in exactly the reverse.

In 2025, at the IPA Effectiveness Conference, Binet updated the argument. The presentation, titled "Go Big or Go Home," directly confronted the industry's obsession with efficiency and targeting. Binet's argument: "our industry has become focused on efficiency, targeting, and short-term metrics at the expense of scale, reach, and true brand-building effectiveness."

60/40 The optimal investment ratio from 1,000+ IPA cases:
60% brand building (broad reach) / 40% activation (precise targeting)

IPA data shows that campaigns with high market coverage (broad reach) generate, on average, 3x more long-term business effects than campaigns with restricted targeting. The mechanism is consistent with Ehrenberg-Bass: broad reach builds mental availability among light buyers and non-buyers, who are the source of future growth.

The neurochemistry of repeated exposure

Robert Zajonc published his study on the mere exposure effect in 1968: repeated exposure to a stimulus, even without conscious attention, increases preference for that stimulus. The effect is robust, replicated across hundreds of subsequent studies, and operates below the threshold of consciousness.

Antonio Damasio, in Descartes' Error (1994), demonstrated that decisions, including purchase decisions, involve somatic markers, unconscious emotional associations formed through experience. The formation of these markers depends on varied exposure across different contexts.

Restricted targeting reduces the number of exposures for the majority of category consumers. Reduces context variety. And therefore reduces the formation of favorable somatic markers. The result: the brand loses emotional relevance among potential buyers who never see its ads because the algorithm decided they weren't part of the "target."

The data paradox: the more precise, the less useful

The ad tech industry sold for two decades the idea that behavioral data can predict who will buy. The reality, documented by Sharp and by independent researchers, is different.

Purchase behavior within a category is stochastic. Buyers alternate between brands with regularity that follows probabilistic models (Dirichlet, documented by Ehrenberg). The "loyalty" that CRM data suggests is, in most cases, a statistical illusion: frequent buyers of one brand are frequent buyers of the category, and they buy other brands with similar frequency.

When a targeting platform identifies a "loyal consumer" and concentrates impressions on them, it's spending money on someone who would have bought anyway. And failing to invest in consumers it could acquire. Economists call this "subsidizing inframarginal consumers." In practice: you're paying to convince the already convinced.

Mark Ritson, Marketing Week columnist and marketing professor, summarized in 2019: "The more we target, the less we grow. It's not a paradox. It's just math."

What actually works

The alternative to extreme precision targeting is what Sharp calls "sophisticated mass marketing." The strategy follows three principles:

Reach all category buyers. The goal is to maximize reach among all consumers who buy or may buy in the category. Category-level targeting, yes. Hyper-targeting by behavioral or demographic profile, in most cases, no.

Build and refresh memory structures. Communication should create and maintain Category Entry Points, the moments and contexts in which consumers think of the category. The more CEPs a brand occupies in memory, the higher the probability of being recalled and purchased.

Invest in distinctive assets with broad reach. Colors, logos, sounds, formats, characters: distinctive assets need to be seen by everyone, repeatedly, to build automatic recognition. Restricting the audience for distinctive assets restricts their effectiveness.

Why P&G went back to reach

After cutting $200 million, Marc Pritchard articulated the new strategy as "mass reach with precision." The goal shifted: instead of finding the right person, maximize coverage among category buyers, using data to control frequency (not to restrict audience). In 2019, P&G declared that reach was the primary metric, replacing targeting metrics.

Jon Moeller, P&G's CFO, confirmed in July 2019: "A focus on reach is more important than a focus on spend. We're reaching more consumers, more often, more efficiently."

What this changes in planning

Anti-hyper-targeting audit: 7 questions

The target the market keeps hitting

The irony of precision targeting is that it hits. It hits exactly the people who would have bought already. And misses everyone who could buy for the first time.

Brands grow when they enter the memory of more consumers. They enter the memory of more consumers when they're seen by more consumers. They're seen by more consumers when reach is broad. The chain is simple. And the chain is broken every time someone adds another layer of segmentation to the media plan.

P&G understood this in 2016. Adidas learned the hard way in 2019. Binet & Field documented it with 1,000 cases. Ehrenberg-Bass replicated it across 50+ categories. Kahneman explained the cognitive mechanism.

The remaining question is simple: how many years of evidence does it take for a boardroom to change a PowerPoint?

References

Books

  • Sharp, Byron. How Brands Grow: What Marketers Don't Know. Oxford University Press, 2010.
  • Romaniuk, Jenni & Sharp, Byron. How Brands Grow: Part 2. Oxford University Press, 2015.
  • Binet, Les & Field, Peter. The Long and the Short of It: Balancing Short and Long-Term Marketing Strategies. IPA, 2013.
  • Kahneman, Daniel. Thinking, Fast and Slow. Farrar, Straus and Giroux, 2011.
  • Damasio, Antonio. Descartes' Error: Emotion, Reason, and the Human Brain. Putnam, 1994.
  • Ehrenberg, Andrew S.C. Repeat-Buying: Facts, Theory and Applications. Griffin, 1972.

Academic Articles

  • Kennedy, R., Ehrenberg, A.S.C. & Long, S. "Competitive Brands' User Profiles Hardly Differ." Market Research Society Conference, 2000.
  • Sharp, B., Riebe, E. et al. "In 25 Years, Across 50 Categories, User Profiles for Directly Competing Brands Seldom Differ." Journal of Advertising Research, 52(2), 2012.
  • Tversky, A. & Kahneman, D. "Availability: A Heuristic for Judging Frequency and Probability." Cognitive Psychology, 5(2), 1973.
  • Zajonc, Robert. "Attitudinal Effects of Mere Exposure." Journal of Personality and Social Psychology, 9(2), 1968.
  • Binet, L. & Field, P. "Effectiveness in Context." IPA, 2018.
  • Ehrenberg, A.S.C., Uncles, M.D. & Goodhardt, G.J. "Understanding Brand Performance Measures: Using Dirichlet Benchmarks." Journal of Business Research, 57(12), 2004.

Industry Sources

  • AdWeek. "When Procter & Gamble Cut $200 Million in Digital Ad Spend, It Increased Its Reach 10%." March 2018.
  • Marketing Week. "Adidas: We over-invested in digital advertising." October 2019.
  • Marketing Week. "P&G puts focus on reach: It's a more important measure than spend." July 2019.
  • IPA. Les Binet & Will Davis. "Go Big or Go Home." IPA Effectiveness Conference, 2025.
  • Ehrenberg-Bass Institute. "The Law of Brand User Profiles." Research Briefs, 2023.
  • The Drum. "P&G's media chief doubles down on bid to better balance targeted ads with mass-reach strategies." May 2017.

Diego Isaac

Brand strategist with 10+ years at the intersection of brand science, consumer behaviour, and communication.

diegoisaac.com.br