MARKETING EFFECTIVENESSEVIDENCE REPOSITORY
Most marketing decisions are made on intuition, convention, or short-term performance data. This repository exists to correct that. It is a structured, verified collection of the most rigorous evidence on how brands grow, how advertising works, and how investment compounds.
Brands Are Memory Structures, Not Rational Propositions.
Core thesis - running through every pillar
Growth comes from being mentally (and physically) present at the moment a buyer enters a category - not from persuading them in that moment. The evidence for this is not a theory or a school of thought. It is a set of empirical laws, replicated across categories, markets, and decades, that describe how buyers actually behave and how brands actually grow.
The following pillars are arranged in a deliberate sequence across eight strategic layers.
Twelve Pillars Across Eight Layers.
Behavioural Science & Decision Architecture.
The neurological and cognitive foundation that explains why marketing works. This pillar documents how human decisions are actually made - through fast, automatic, emotion-driven processes rather than slow rational deliberation - and what that means for how brands should communicate.
If you believe consumers evaluate your brand rationally at the point of purchase, your marketing strategy will be structurally wrong. The evidence from Kahneman, Damasio, Heath, and others is not speculative - it is derived from decades of controlled experiments, neurological case studies, and cross-cultural replication. Damasio's somatic marker hypothesis demonstrates that emotion is a neurological prerequisite for effective decision-making, not a bias to be overcome. Heath's low attention processing work shows that brand communications build associations and emotional affinity without conscious attention or active recall - a mechanism that operates across paid, earned, and owned touchpoints, not just advertising exposure.
The strategic tension here is not a debate about methodology - it is a debate about what branded communications is fundamentally for. One school holds that it should build emotional resonance, creating warm associations that make a brand feel right before a purchase moment arrives. The other holds that it should build distinctive memory cues - ensuring the brand is noticed and correctly attributed, regardless of emotional content. The direction of travel from both sides is convergent: rational persuasion is not the mechanism through which most branded communications work, and the marketing job is to build pre-conscious availability. The practical implication is not to choose between emotion and salience, but to ensure creative achieves both - processed because it engages, remembered because it is distinctively branded.
An important emerging dimension: Shaw and Nave (Wharton, 2026) formally extend Kahneman's dual-process framework by positing System 3 - artificial cognition operating outside the brain - through Tri-System Theory. Their key empirical finding is cognitive surrender: buyers adopting AI outputs with minimal scrutiny, bypassing both System 1 intuition and System 2 deliberation entirely. The marketing implication - if this holds at scale - is that brands absent from AI outputs face a structural retrieval problem that neither emotional resonance nor distinctive memory cues can overcome at the moment of AI-mediated decision-making. The paper is new and has not yet been independently replicated, so it should be treated as directional evidence rather than established law.
Stop briefing for rational persuasion. The goal of most brand communication - whether a TV spot, a piece of editorial coverage, a product interaction, or a consistently distinctive social presence - is to build the emotional memory structures that make a brand feel right before a buying situation arises, not to change minds in the moment. Cialdini's persuasion principles, Thaler's choice architecture, and Shotton's applied biases provide a practical toolkit for structuring brand communication and purchase environments in line with how decisions actually happen.
Mental Availability & Memory Structures.
Mental availability is the probability that a brand comes to mind when a buyer enters a buying situation - and it is determined not by awareness or preference, but by the breadth and strength of the memory structures linking the brand to category entry points.
Brands do not lose to competitors because they are worse. They lose because they are not thought of. The Ehrenberg-Bass research programme - spanning decades of longitudinal purchase panel data across hundreds of categories and markets - establishes that penetration, not loyalty, is the dominant source of brand growth. Most buyers buy any given brand infrequently, alongside competitors, without strong loyalty. The job of brand communications - paid, earned, and owned - is to build and refresh the memory structures that make retrieval more likely when those buyers next enter the category.
The live debate is not whether brands need to be thought of - that is settled - but whether being thought of is sufficient on its own. The Kantar evidence suggests that brands which are both mentally available and meaningfully different to category buyers grow penetration faster than those that are merely salient. The direction of travel implies a strategic hierarchy: salience is the floor, not the ceiling. Brands that achieve wide CEP coverage but attach no meaningful associations to those memory structures may generate awareness without preference and reach without conversion. The practical challenge is not to abandon the mental availability frame but to ensure the associations being built are ones that genuinely matter to buyers - not just that the brand comes to mind, but that what comes to mind has value.
Shaw and Nave's Tri-System Theory (Wharton, 2026) introduces a structural complication to the mental availability model: if AI increasingly mediates the moment of brand retrieval - summarising, recommending, and shortlisting on the buyer's behalf - then being present in human memory may be necessary but no longer sufficient. A brand can be strongly encoded in a buyer's mental availability network and still fail to surface in an AI-mediated response. The question of what determines AI-mediated salience - training data presence, authoritative citations, review volume, category signal strength - is materially different from what determines human memory retrieval. The two frameworks are not yet integrated, and the evidence base for AI-mediated mental availability is thin. But the direction of travel is clear enough to name: mental availability strategy may need to operate across two systems simultaneously - human memory and institutional AI retrieval - and the investment logic for each is different. At the operational level, this creates a dual audience problem: human buyers who need emotional memory structures that make a brand feel right before a purchase moment, and machine systems that need clean, extractable, verifiable claims they can parse and act on. A brand that addresses only one audience is either invisible to AI agents or inert to human buyers. The brands most likely to maintain mental availability across both systems are those whose communications are simultaneously emotionally resonant and concretely evidenced.
Build memory structures before you need sales. Identify the category entry points - the buying situations, occasions, and motivations through which buyers enter your category - and map your brand's coverage of them. Gaps in CEP coverage are gaps in future revenue. Invest consistently across paid, earned, and owned channels to reach all category buyers across all relevant CEPs - recognising that owned and earned channels tend to index toward existing audiences, and that broad reach to light and non-buyers typically requires paid media to achieve at scale.
And begin - directionally, not definitively - to think about AI-mediated availability as a parallel task. The brands most likely to survive System 3 cognitive surrender intact are those that are both strongly encoded in human memory and well-represented in the authoritative sources that AI systems draw on. Maintain that investment over time: mental availability decays non-linearly when advertising stops, and rebuilding is significantly more expensive than maintaining. A useful reframe for internal alignment: the job is not to be known. It is to be thought of, at the right time, by enough people.
Physical Availability & Distribution Reach.
The conversion prerequisite. Physical availability is being easy to find, choose, and buy - across every channel and context in which category buyers transact. It is the mechanism that converts mental availability into actual sales.
A brand that comes to mind in a buying situation but cannot be found, selected, or purchased in that moment has converted brand investment into consideration without revenue. Physical availability has three dimensions: presence (being in the channels where buying actually happens), prominence (being easy to find within those channels through distinctive assets and shelf or search positioning), and frictionlessness (removing barriers to completion of the purchase). The Ehrenberg-Bass dual market-based asset model establishes the multiplicative relationship: high mental availability where the brand is physically absent produces minimal sales, and vice versa.
The direction of travel in physical availability is toward greater complexity, not less. In traditional retail, physical availability was primarily a distribution and ranging problem. In digital categories it has become an algorithmic problem - search rank, platform prominence, checkout friction, and in-feed placement are the modern equivalents of shelf position and in-store visibility. The strategic question that is still being resolved is whether strong mental availability can override weak physical availability through the force of consumer preference, or whether in digitally mediated categories physical availability has become the primary determinant of whether consideration converts to purchase. The emerging evidence favours the latter: even strongly preferred brands lose meaningful revenue when they are difficult to find or slow to complete.
A materially important emerging dimension: AI-mediated search - Google AI Overviews, Perplexity, ChatGPT with search, and conversational commerce - is changing what physical availability means in practice. The traditional frame assumed a human navigating search results; in AI-mediated search the AI summarises and recommends before the human navigates anywhere. Being findable in a traditional SERP is no longer sufficient if the AI response does not surface the brand. The signals that drive AI recommendation appear to differ meaningfully from traditional SEO signals - they respond more to authoritative citations in trusted sources, brand name recognition as a category signal, review volume and sentiment, and presence in training data. Shaw and Nave's cognitive surrender finding adds a further dimension: in AI-mediated purchase contexts, the buyer may not navigate away from the AI response at all. If the AI recommends a brand and the buyer accepts that recommendation without further deliberation, then physical availability in the traditional sense becomes secondary to being selected before the buyer looks. This is a meaningful structural shift: presence in the AI response may be the new first shelf, and everything downstream of it is conversion, not discovery.
The evidence base here is thin and fast-moving; the direction of travel is clear enough to warrant attention now. The practical audit for AI-mediated physical availability is direct: query your category in ChatGPT, Perplexity, or whichever AI tool your category buyers use. Identify what gets cited or surfaced. Those URLs are the new shelf positions. Help centres, product pages, compatibility notes, and stated limitations - written as extractable evidence rather than brochure copy - become the raw material AI systems can lift and attribute. Broad web presence is insufficient; named presence in the specific sources models draw on is what determines AI shelf position.
Audit your physical availability before explaining a performance gap with brand metrics. Are you in every channel where your category buyers transact? Is your brand easy to find within those channels? In digital contexts, physical availability means search rank, platform visibility, and ease of completion - the digital equivalent of shelf positioning. Distinctive brand assets are the mechanism of physical prominence: they make the brand recognisable and selectable without requiring active search.
In AI-mediated categories, add a third audit question: does your brand surface in the AI responses your category buyers are most likely to encounter? No standard framework yet exists for measuring AI shelf presence the way Nielsen measures physical distribution - but the strategic logic is the same: if you are not present at the moment of selection, mental availability cannot save you.
Brand Equity, Distinctiveness & Meaningful Difference.
What a brand is worth in the minds of buyers - and the competing evidence on whether that value is driven by recognition, meaning, or both.
The most fundamental debate in marketing effectiveness is not whether brand investment pays - that is well-established - but why. The Ehrenberg-Bass school argues that brand growth is driven by salience and distinctiveness: being thought of and being recognised. The Keller/Aaker tradition argues that meaningful associations, perceived quality, and brand meaning create durable equity that translates to pricing power, loyalty, and resistance to competitive attack. The Kantar Blueprint frames this as Meaningful Difference - a composite of salience, meaningfulness, and differentiation - and argues it is five times more predictive of penetration growth than salience alone. These are not fully reconcilable positions; the evidence for each comes from different methodological traditions. Both must be engaged with honestly rather than one being selected as the canonical answer.
The unresolved strategic question is whether brand equity is better pursued as a salience and recognition challenge or a meaning and differentiation challenge - and whether the right answer differs by market position, category maturity, and competitive context. The direction of travel from the accumulated evidence points toward a synthesis: salience without meaning creates fragile equity that is vulnerable to commoditisation and promotional dependency; meaning without salience creates premium positioning that cannot scale because buyers are not retrieving the brand at the moment of choice. The most commercially durable position appears to belong to brands that are both easily thought of and associated with something genuinely valued. The open question for any given brand is which dimension to prioritise when resources are constrained - and the honest answer is that it depends entirely on where the brand currently sits on each.
Do not choose between the salience and meaning schools - the evidence is not strong enough to justify dismissing either. Invest in both across all brand touchpoints - paid, earned, owned, and the product experience itself: build the distinctive memory structures that make your brand come to mind, and ensure those memory structures are attached to associations that make the brand feel meaningfully different to enough people. Measure brand equity properly - revenue premium is a more rigorous outcome measure than brand awareness or brand image scores alone. A useful practitioner synthesis of the two schools: salience without substance is fragile; substance without salience is invisible.
Long vs Short-Term Investment.
The evidence base for the most commercially significant finding in marketing effectiveness research: that brand-building and sales activation are structurally different activities, operating on different time horizons, generating different types of commercial value, and requiring different measurement approaches to evaluate.
The dominant measurement infrastructure of modern marketing - digital attribution, short-term ROAS, last-click models - is calibrated to detect activation effects and is structurally blind to brand-building effects. The result is a systematic misallocation of investment toward the channels and activities that are most measurable rather than most effective. Binet and Field's IPA Databank analysis establishes the mechanism: activation drives immediate volume but decays within weeks; brand-building - through paid media, earned coverage, owned content, and consistent product experience - builds the memory structures and emotional associations that determine what volume is achievable in future, at what margin, with what resistance to competitive pressure. The long-term multiplier is that the majority of brand communications profit accrues beyond the first 13 weeks of a campaign. Most measurement systems capture none of it.
The direction of travel on this question has been unambiguous for over a decade: the industry has systematically over-invested in short-term activation and under-invested in brand-building, and the case for correcting this has only strengthened with each independent dataset. What remains genuinely contested is not the direction - rebalance toward brand - but the calibration: how much, in what ratio, and whether the optimal balance is shifting as media environments evolve. The practical risk is that marketers acknowledge the principle while deferring the rebalancing - using category context as justification for the status quo rather than as a tool for calibration. The most commercially dangerous version of this debate is the one that treats the 60/40 starting point as a ceiling rather than a floor for brand investment.
Protect brand investment - including paid media, but also the earned and owned activity that builds memory structures across all touchpoints - as a non-negotiable strategic budget line. The finance instinct to cut brand investment during a downturn is the single most evidence-refuted decision in marketing. Commission marketing mix modelling - not attribution - to measure total advertising ROI including the long-term component. Present the long-term ROI case to the board using total modelled return, not short-term ROAS: the gap between the two is the hidden cost of measurement myopia.
Share of Voice & Budget Setting.
The relationship between a brand's share of advertising investment and its share of market - and what that relationship means for how budgets should be set and defended.
Excess Share of Voice (ESOV = SOV minus market share) is the single most reliable predictor of future market share direction. Brands with positive ESOV grow; brands with negative ESOV decline. This relationship - first documented by John Philip Jones in 1990, replicated by Binet and Field across the IPA Databank, and independently corroborated by Analytic Partners across 1,000+ brands in 45 countries - is one of the most robustly replicated findings in the effectiveness literature. It transforms the budget-setting conversation from a negotiation based on historical precedent into a strategically grounded question: what level of investment is required to defend or grow our market position?
The direction of travel on ESOV is away from using it as a precise forecasting formula and toward using it as a strategic guardrail. The relationship between spend share and market share is robust enough to answer the most important budget question - are we above or below our fair share of voice, and what does that imply for our trajectory - without requiring mechanical precision about the conversion rate. The live strategic debate is whether traditional share of voice remains a meaningful concept in fragmented digital media environments, where attention quality, creative strength, and contextual relevance all mediate the relationship between money spent and minds reached. The direction of travel suggests that share of voice needs to be complemented by share of attention and share of search to remain strategically actionable.
Calculate your ESOV before setting or defending your media budget. If your SOV is below your market share, the evidence predicts decline - and that prediction belongs in every budget conversation. If you are a challenger brand, you need to maintain positive ESOV. Share of Search gives you a way to track this in real time without proprietary media data - and because it captures organic and earned signals as well as paid, it is a more channel-agnostic indicator of total brand momentum than traditional SOV measures alone.
Reach, Targeting & Contextual Relevance.
The evidence that broad reach - getting the brand message to as many category buyers as possible - is more effective for brand growth than narrow targeting of existing customers or high-probability converters. And that within broad reach, contextual relevance materially improves ROI without sacrificing the reach principle.
If brands grow through penetration - through acquiring more buyers from the full pool of category buyers - then the target audience for brand-building investment is, by definition, the broadest possible audience of category buyers. Any strategy that narrows that audience below the full category excludes some of the buyers the brand most needs to reach. Targeting strategies optimised for heavy users, existing customers, or lookalike audiences systematically underweight the people who represent the growth pool. When Brands Go Dark adds the decay dimension: mental availability erodes non-linearly when consistent broad-reach communication stops - whether that communication is paid, earned, or owned.
The direction of travel is clear but the practical destination is not yet settled: the industry is moving away from audience-based narrow targeting as the dominant planning logic and toward contextual relevance within broad reach as the emerging standard. The debate is no longer whether over-targeting damages brand growth - that is increasingly accepted - but how to operationalise broad reach with contextual intelligence at the scale modern media demands. The practical resolution emerging from the evidence is a tiered model: brand-building investment planned for broad category reach, with contextual placement optimising within that reach, and only sales activation investment targeted narrowly at high-intent moments. The structural risk in the current environment is that addressable media capabilities continue to pull investment toward targeting precision before the strategic case for reach has been accepted internally.
Resist the commercial pressure to over-target through paid media, but also audit whether your earned and owned activity is reaching beyond your existing audience - the reach imperative applies across all channels, and owned channels in particular tend to concentrate on those already engaged with the brand. Every time you narrow your audience below the full category, you are trading short-term efficiency for long-term reach deficiency. In digital channels, contextual placement within category-relevant environments is the evidence-based middle ground: broad reach by design, relevant by context.
Creative Effectiveness.
The evidence that creative quality is the largest single controllable variable in advertising effectiveness - and that the industry has been systematically producing less of it for two decades.
Given equivalent media investment, the difference in commercial effectiveness between outstanding creative and average creative is not marginal - it is decisive. The Effectiveness Code quantifies it: seriously creatively committed campaigns deliver ten times the efficiency of average campaigns. The Creative Dividend connects creative quality ratings directly to long-term brand contribution: high-rated creative delivers three times the long-term brand contribution of low-rated work. Orlando Wood's Lemon identifies the mechanism of the problem: brand communications have shifted structurally toward left-brain, rational, performance-style creative over two decades - driven by digital measurement incentives that apply across paid, owned, and earned channels alike.
The live strategic debate is not whether creative quality matters - the directional evidence from multiple independent sources is consistent enough to treat that as settled. The more important and genuinely unresolved question is whether the specific research underpinning the strongest claims in this pillar can bear the weight placed on it. Byron Sharp and others have raised substantive challenges to The Creative Dividend specifically: that System1's Star Rating measures emotional response in a test environment, not causal commercial outcomes in market; that the linkage to Effie Awards introduces compounded survivorship bias - creative that won a commercial effectiveness award was already pre-selected for exceptional performance; and that System1's commercial model creates a structural incentive to demonstrate that their test predicts success, regardless of whether it actually does.
The Effectiveness Code carries related risks - comparing seriously creatively committed campaigns to an average that includes low-investment and poorly-executed work will naturally produce large multipliers, but those multipliers tell you less about creative quality per se than about the full bundle of commitment, budget, time, and integration that award-winning campaigns represent. Sharp's alternative reading is that what looks like an emotional creative effect may actually be a distinctive asset effect - creative that works does so because it encodes the brand correctly into memory, not because it generates emotional arousal. This is a mechanistic dispute, not a directional one. Where the evidence does converge, across independent sources and across the contested methodologies, is on a simpler and more durable finding: brands that communicate consistently, with genuine creative commitment over time, outperform those that do not. The specific multipliers should be treated as directional. The underlying principle should not.
An important emerging implication of AI on creative effectiveness: when average execution becomes free - when any brand can produce competent, on-brief content at scale - the performance differential between genuinely creative work and average work grows rather than shrinks. Teams with taste can produce at scale; teams without taste produce mediocrity at scale. Under AI-driven production conditions, that case becomes stronger, not weaker. The CMO function increasingly becomes a taste function - protecting the standard of work that no algorithm will spontaneously produce and no optimisation system will select for.
A specific AI risk for creative effectiveness: AI tools are highly capable at short-term activation work - generating offers, sequences, variants, and retargeting logic at scale and speed. This creates a structural pull toward short-term execution that is more severe than the pre-AI version of the same problem, because the supply of short-term work is now effectively unlimited. If no one actively protects long-term brand investment, short-term activation will expand to fill the entire creative and budget conversation.
Brief for emotional resonance first, rational messaging second. The right creative question is not 'does this communicate the message?' but 'will this be remembered warmly, and will it be unmistakably ours?' Commission creative testing that measures emotional response and long-term brand potential, not just recall and message comprehension. Invest in creative talent and production quality across all brand communications - not just paid media. The same emotional resonance principles that make paid advertising effective apply to owned content, earned media pitches, and the product experience itself. A superior creative running on the same media plan will outperform an average creative by a factor that no targeting or optimisation decision can replicate.
On consistency: the industry instinct is to retire creative when the team is bored of it, which is almost always before the audience has fully processed it. The evidence on wear-in versus wear-out consistently shows that effective creative takes longer to reach its performance peak than most organisations allow, and that the point at which internal stakeholders declare creative stale is typically well before the point at which audiences have stopped responding to it. Refreshing too early destroys the compounding returns that creative consistency generates - both in memory encoding terms and in the distinctive asset recognition that builds over repeated exposure. The practical discipline is to distinguish between creative wear-out - where response genuinely declines - and internal fatigue, which is not the same thing and should not drive the same decision. Evolve the creative system when the evidence demands it, not when the team needs variety.
An important emerging implication of AI on creative effectiveness: when average execution becomes free - when any brand can produce competent, on-brief content at scale - the performance differential between genuinely creative work and average work grows rather than shrinks. Teams with taste can produce at scale; teams without taste produce mediocrity at scale. Under AI-driven production conditions, that case becomes stronger, not weaker. The CMO function increasingly becomes a taste function - protecting the standard of work that no algorithm will spontaneously produce and no optimisation system will select for.
A specific AI risk for creative effectiveness: AI tools are highly capable at short-term activation work - generating offers, sequences, variants, and retargeting logic at scale and speed. This creates a structural pull toward short-term execution that is more severe than the pre-AI version of the same problem, because the supply of short-term work is now effectively unlimited. If no one actively protects long-term brand investment, short-term activation will expand to fill the entire creative and budget conversation.
Attention & Memory Encoding.
The evidence that attention is the conversion mechanism between media exposure and memory encoding - and that the media industry's primary quality metric (viewability) is a poor proxy for attention, systematically overvaluing low-attention digital environments.
Reach measures the potential audience for an advertising message. Attention measures whether that audience actually processed it. The two are not the same. Karen Nelson-Field's passive attention measurement research - using eye-tracking and biometric data across tens of thousands of ad exposures - establishes that active attention is strongly predictive of brand memory encoding and downstream sales outcomes. Viewability is a poor proxy: an ad can be fully viewable and receive virtually no human attention. The platform-level hierarchy has significant commercial implications: TV and premium video consistently generate substantially higher active attention per exposure than social feeds and digital display.
The direction of travel on attention is toward it becoming a planning input rather than a channel preference. The debate is shifting from whether attention matters - settled - toward how much attention is required for different brand-building effects, and whether that threshold varies by creative quality and emotional content. This is strategically significant because it determines whether low-attention environments should be avoided entirely or deployed selectively for the right type of message. The emerging evidence suggests that emotionally resonant creative may achieve meaningful brand encoding at lower attention thresholds than rational or informational creative - which means attention planning and creative strategy are more interdependent than either discipline currently treats them, and that the right response to a low-attention environment may be a creative solution rather than a media one.
Attention-adjust your media planning. CPM is not the right currency for comparing channels if attention quality per impression varies by a factor of five or ten across those channels. Apply attention-adjusted reach as a planning lens alongside raw reach metrics. Challenge your media agency to demonstrate attention quality per channel, not just cost per viewable impression.
For CFO and board communication, the attention research supports a reframe that lands more effectively than abstract brand-building arguments: if category buyers cannot identify the brand within the two-to-three second attention window available on social platforms, the media spend is wasted regardless of targeting quality. The question is not whether brand building is important - it is whether the brand is recognisable enough for the media investment to function at all.
Pricing Power & Promotional Mechanics.
The evidence on how pricing power is built and how promotional dependency erodes it. This pillar is structured in two sub-sections: (A) Pricing Power - how brand investment creates pricing headroom and reduces price elasticity; and (B) Promotional Mechanics - when promotional activity helps or harms brand equity.
Price is the highest-leverage variable in most marketing econometric models. Yet most marketing investment conversations treat pricing and promotions as finance and trade decisions rather than brand strategy decisions. They are both. Ehrenberg, Hammond and Goodhardt's foundational research established the promotional mechanics finding: almost everyone who buys a brand during a promotional period has bought it before. Promotions do not necessarily acquire new buyers - they pull sales forward from existing buyers at lower margin. Separately, the academic econometric literature (Sethuraman and Tellis) establishes that brand communications investment reduces price elasticity - meaning strong brand investment directly creates the pricing headroom that appears as margin in P&L models.
The strategic debate that remains live is not whether promotional dependency damages long-term brand value - that is well established - but whether brands have a viable and realistic path back to pricing power once they have lost it, and how long that path takes. The direction of travel is clear: pricing power is rebuilt through sustained brand investment, not through pricing architecture decisions, and the timeline is measured in years not quarters. The commercially uncomfortable implication is that brands which have traded pricing headroom for promotional volume efficiency have made a decision that is not quickly reversed even with a correct strategy. The more immediate debate for most businesses is how to protect brand investment during inflationary and recessionary cycles, which are precisely the conditions that create the greatest internal pressure to promote - and the greatest long-term cost of doing so.
Treat pricing power as a brand equity outcome, not a pricing strategy decision. If your brand is losing pricing headroom - requiring more promotion to hold volume, facing retailer pressure on margin, seeing price elasticity increase - the diagnosis is almost certainly a brand equity deficit. The solution is investment in mental availability, long-term brand investment, and creative quality - not more sophisticated promotional mechanics. Every sustained promotional campaign is training your buyers to wait for the deal.
Measurement, Econometrics & Causal Inference.
The evidence on how to measure advertising effectiveness accurately - and why the most widely used measurement approaches systematically misrepresent what is actually driving commercial outcomes. This pillar covers three measurement approaches: marketing mix modelling, digital attribution, and incrementality testing / causal inference.
The measurement system tells you where to invest. If it is wrong about causation, it will direct investment toward the wrong activities - consistently, at scale, compounding over years. Last-click attribution is demonstrably wrong about how advertising works. It overstates the role of the final digital touchpoint by a factor of two to ten, cannot capture brand effects or offline channels, and systematically rewards the channels that harvest demand the brand built - at the expense of the channels that built it. Marketing Mix Modelling is significantly more accurate than attribution for the questions that matter most to marketing strategy - but it is not perfect. Incrementality testing (geo-experiments, randomised lift studies) provides the causal inference that MMM cannot: it answers "did this specific activity cause this specific outcome" rather than "what does the correlation pattern suggest?"
The direction of travel on measurement is toward triangulation rather than any single authoritative method. The industry is moving - slowly - away from the false precision of last-click attribution and toward a portfolio of approaches: MMM for strategic budget allocation, incrementality testing for tactical validation, and brand health tracking as the leading indicator of future revenue that short-term financial metrics will never surface. The live strategic debate is how to govern that portfolio in practice - whose model to trust when methods disagree, and how to communicate complex measurement outputs to boards that have been conditioned by the apparent simplicity of digital attribution dashboards. The most dangerous current failure mode is not using the wrong method - it is using measurement selectively to validate decisions already made on other grounds, which is what poor attribution systems have enabled and what better measurement must make harder to do.
If your primary marketing effectiveness measurement is digital attribution, you are misallocating budget. Commission marketing mix modelling as a structural investment - not a one-off project - and use it as the primary instrument for budget allocation decisions. Build an incrementality testing capability alongside MMM to validate causal claims. Supplement both with brand health tracking and Share of Search as real-time leading indicators. Present the long-term ROI case using total modelled return, not short-term ROAS.
Category Dynamics, Market Structure & Boundary Conditions.
The evidence that brand growth does not happen in a vacuum but within a category that has its own demand structure, growth trajectory, and competitive logic - and an explicit inventory of the boundary conditions under which the preceding eleven pillars apply less reliably.
A brand executing flawlessly on every preceding pillar will still be constrained by the growth dynamics of the category it competes in. Category Entry Points define the structural opportunity available to any brand. Brands with broader CEP coverage have higher mental availability and higher penetration because they are retrieved in more of the moments when buyers are choosing. Kantar's demand space analysis extends this to category-level opportunity mapping.
The strategic question this pillar leaves genuinely open is when brands should compete for existing category demand and when they should invest in expanding the category itself. The direction of travel from the large-scale evidence is that category penetration - growing the total pool of buyers - is typically the higher-return strategy for established brands, because winning share from competitors in a fixed pool is expensive, contested, and often temporary, while category growth lifts the whole market disproportionately in favour of brands with the strongest mental availability.
The live debate is whether this logic holds in declining or commoditising categories (fewer people buying, growing pool of buyers is harder because the pool is shrinking); whether it applies to challenger brands with limited reach (the question of spreading thin across the whole category or concentrating investment to build strength in a smaller segment first); and whether brands operating in winner-take-most digital markets (more people that use the category leaders, the more valuable they become to everyone) face a fundamentally different competitive structure in which penetration logic breaks down in favour of network-driven consolidation.
Much of the foundational evidence in Layers 1-7 - particularly the empirical case that 95% of category buyers are out of market at any given time, the dominance of penetration over loyalty, and the case for sustained mental availability investment to reach future buyers - rests on Ehrenberg's NBD-Dirichlet model of purchase frequency. The model has decades of cross-category validation. It also has a critical assumption that is rarely surfaced in the popular synthesis literature: it applies only to stationary markets, where the total pool of category buyers is broadly stable over time and where new entrants and departing buyers approximately cancel each other out.
In growing categories, the assumption breaks. Empirical analysis (Dale W. Harris, extending Ehrenberg's original work) suggests the in-market percentage is not 5%. It is materially higher - directionally 20-30%+ in categories growing at 50-100% per year, scaling with category growth rate. More importantly, the composition of the in-market pool shifts. In stationary categories, almost all in-market buyers are returning category buyers making replacement purchases. In high-growth categories, the majority of in-market buyers can be first-time category entrants - buyers who have never bought the category before and have no existing brand associations to draw on.
A rough diagnostic for any category: estimate annual category growth (in buyer count, not revenue, where possible - revenue growth can come from price or upsell rather than buyer expansion). Convert to quarterly compound rate using (1 + annual growth) to the power of one quarter, minus one. Add the ~5% baseline for stationary-market in-market percentage. The result approximates total in-market percentage. Quarterly growth divided by total in-market approximates the share of in-market buyers who are first-time category entrants.
Worked example: a category growing 13% annually has a quarterly compound rate of ~3.2%. Add the 5% baseline. Total in-market is approximately 8.2%. First-time entrants are approximately 39% of in-market buyers. At 100% annual growth: quarterly rate ~19%, total in-market ~24%, first-time entrants ~79% of in-market buyers.
The strategic consequences are direct. The brand/activation split is conditional. Binet and Field's 60/40 guidance assumes a category where future buyers compound slowly through long-term mental availability investment. In a growing category with a larger active in-market pool, more weight can sit on activation and category education - because more buyers are actively choosing now, and a meaningful share of them need to be taught what the category is before they can develop brand associations.
Category education becomes a strategic activity, not background noise. First-time category entrants cannot be mentally available to a brand they don't yet understand a category exists for. In growing categories, brand-building work happens after - or in parallel with - category-defining work. In stationary categories, the category is already understood; the brand-building work happens within an existing mental model.
Two distinct strategic postures emerge for businesses in growing or emerging categories. Category creation (the Salesforce/CRM model) - the brand defines a new buyer-shopped procurement category and becomes synonymous with it. Or operating frame across established adjacent categories (the Adobe Creative Cloud model) - the brand enters through existing sub-categories (Photoshop in photo editing, Premiere in video editing) and uses an internal operating frame to give itself coherence across them, without trying to make the operating frame itself a buyer-shopped category. Both are defensible strategic choices. The right one depends on capital, structural conditions, and competitive timing. The Salesforce playbook is expensive, slow, and depends on conditions a company cannot fully control. The Creative Cloud playbook works with how the market already operates and produces a coherent multi-vertical brand without requiring the market to adopt new procurement language.
What this means for application of the doctrine: before applying the standard penetration / mental availability / 60-40 guidance, test whether the category is growing materially. If it is, the constraint mix is likely different from a stationary category and the brand/activation calibration shifts. The framework still applies - the buying mechanics (memory and availability) are the same - but the weighting changes. This is the boundary condition Pillar 12 has always pointed to. The Harris work makes the quantitative case explicit.
Before setting brand growth targets, map category growth. Is the category expanding, stable, or declining? Which demand spaces within it are growing fastest, and where is your brand positioned relative to them? Explicitly test whether your category resembles the contexts from which most of this evidence derives - or whether you are operating in one of the boundary condition contexts where the rules apply differently. And test whether your brand's growth is primarily advertising-constrained, distribution-constrained, or product-constrained: the investment logic differs significantly across each, and the channel-agnostic framing of brand communications matters most when the binding constraint is not paid reach.
What to Do Now
