Marketing to the Machine: Rethinking Global Marketing for the AI Era

Classified as probably not a substantial claim. May combine two claims.

Prior research has examined recommendation agents, algorithmic curation and personalization, platform governance, and the growing role of analytics in shaping firm strategy.

“Prior research has examined recommendation agents (Jiang & Benbasat, 2007), algorithmic curation and personalization (Grewal et al., 2017), platform governance (Van Alstyne et al., 2016), and the growing role of analytics in shaping firm strategy (Rust & Huang, 2014).”

• cites Jiang & Benbasat, 2007 sourceNot checked — no full text of the source
• cites Grewal et al., 2017 sourceSupported
• cites Van Alstyne et al., 2016Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing).

Digital mediation of consumption is not new.

“Digital mediation of consumption, as illustrated in the vignettes above, is not new.”

• cites Jiang & Benbasat, 2007 sourceNot checked — no full text of the source
• cites Grewal et al., 2017 sourceSupported
• cites Van Alstyne et al., 2016Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

May combine two claims.

Algorithmic systems often privilege dominant cultural narratives, altering the meanings consumers encounter and the products they see.

“Other work finds that algorithmic systems often privilege dominant cultural narratives, altering the meanings consumers encounter and the products they see (González-Cantera & Bonacchi, 2025).”

• cites González-Cantera & Bonacchi, 2025 sourcePartly supported

  The indexed text (section “1 Active”; no page given) reports that ChatGPT-4o’s interpretations of three contested memorial sites favoured dominant, capitalist, and government-aligned perspectives, supporting the point about privileged narratives. It does not address consumers or the products they see, and the finding is specific to ChatGPT’s heritage interpretations rather than algorithmic systems generally. (LLM, conf=0.85)

  Experimental — accuracy not yet measured

May combine two claims.

AI personalizes, but it also governs and mediates; in some markets it may even help create the representations through which consumption is understood.

“AI personalizes, but it also governs and mediates.In some markets it may even help create the representations through which consumption is understood (Airoldi & Rokka, 2022).”

• cites Airoldi & Rokka, 2022 sourceSupported

May combine two claims.

A brand may appear with hallucinated qualities that misinform consumers or introduce expectations the firm cannot meet.

“A brand may appear with hallucinated qualities that misinform consumers or introduce expectations the firm cannot meet (Campbell et al., 2022).”

• cites Campbell et al., 2022 sourcePartly supported

  The Perils section (page not available) warns that manipulated ads can spread authentic-appearing misinformation, confuse consumers, and alter brand meaning. It does not specifically say that brands may appear to have invented qualities or create expectations the firm cannot meet. (LLM, conf=0.82)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Consumer researchers have shown how digital environments extend the self and reshape identity construction.

“Consumer researchers have also shown how digital environments extend the self and reshape identity construction (Belk, 2013).”

• cites Belk, 2013 sourceNot checked — no full text of the source

May combine two claims.

Demand is increasingly coauthored by people and the systems that interpret them, relocating demand formation from the bounded consumer to a distributed socio-technical architecture.

“together these shifts suggest that demand is increasingly coauthored by people and the systems that interpret them, relocating demand formation from the bounded consumer to a distributed socio-technical architecture.”

May combine two claims.

The four shifts operate at distinct levels of the marketplace-actor, ecosystem, gatekeeping layer, and representation-yet interact recursively as behavior becomes training data that reshapes future exposure.

“these operate at distinct levels of the marketplace, the actor, the ecosystem, the gatekeeping layer, and representation, yet interact recursively as behavior becomes the training data that reshapes future exposure.”

Probably not a substantial claim: may not pass S4 (based). May combine two claims.

Four shifts drive AI-mediated change: the rise of the hybrid consumer, the emergence of algorithmic borders, the growth of intermediated demand, and the spread of generative consumption.

“We outline four shifts driving this change: the rise of the hybrid consumer, the emergence of algorithmic borders, the growth of intermediated demand, and the spread of generative consumption.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

Consumer-AI symbiosis is a persistent configuration in which human intention and machine inference are reciprocally adjusted so that preference and choice are jointly produced.

“We conceptualize this as consumer-ai symbiosis, a persistent configuration in which human intention and machine inference are reciprocally adjusted so that preference and choice are jointly produced.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based). May combine two claims.

In the age of generative consumption, firms must compete for the integrity of the representations through which consumers make sense of products and for influence over the generative mechanisms that define category logic.

“In this age of generative consumption, firms must compete not only for visibility or relevance but for the integrity of the representations through which consumers make sense of products and for influence over the generative mechanisms that increasingly define category logic itself.”

Probably not a substantial claim: may not pass S4 (based). May combine two claims.

In China, discovery flows through ecosystems like WeChat, Douyin, and Alipay, where visibility depends on alignment with local narratives and usage behaviors.

“Discovery flows through ecosystems like WeChat, Douyin, and Alipay, where visibility depends on alignment with local narratives and usage behaviors.”

May combine two claims.

Consumer-AI symbiosis shapes what becomes visible, redefines relevance, filters or constructs options, and can increasingly shape what counts as choice.

“It shapes what becomes visible, redefines relevance, filters or constructs options, and can increasingly shape what counts as choice (Campbell et al., 2020).”

• cites Campbell et al., 2020Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S2 (checkable). May combine two claims.

Generative consumption is analytically distinct from intermediated demand, even though the two operate together.

“Generative consumption is analytically distinct from intermediated demand, even though the two operate together.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

In these environments, familiar indicators such as reach and impressions may reveal little about a brand's true presence.

“In these environments, familiar indicators such as reach and impressions may reveal little about a brand's true presence.”

May combine two claims.

Algorithmic interpretations are not neutral; they reflect cultural and linguistic cues embedded in model training data, producing systematic variation in how needs are recognized and framed across regions.

“Research shows that these interpretations are not neutral.They reflect cultural and linguistic cues embedded in model training data, which produce systematic variation in how needs are recognized and framed across regions (Lu et al., 2025).”

• cites Lu et al., 2025Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

May combine two claims.

Consumer-AI symbiosis is neither frictionless nor neutral; it is structured by the logics embedded in the systems themselves.

“Consistent with emerging evidence that consumers must perform cognitive, practical, relational, and affective work to accommodate algorithmic limitations (Kaliyamurthy & Schau, 2025), symbiosis is neither frictionless nor neutral.It is structured by the logics embedded in the systems themselves.”

• cites Kaliyamurthy & Schau, 2025 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

To appear coherently across AI systems, brands must adapt their narratives, associations, and category cues so they are machine readable.

“To appear coherently across these systems, brands must adapt their narratives, associations, and category cues so they are machine readable.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

The ecosystem of consumption is no longer a neutral marketplace or shared media space; it is a model-governed field of visibility built from predictive rules, cultural priors, and institutional constraints.

“The ecosystem of consumption is no longer a neutral marketplace or a shared media space.It is a model-governed field of visibility built from predictive rules, cultural priors, and institutional constraints.”

Probably not a substantial claim: may not pass S4 (based). May combine two claims.

Markets with thin data, weak infrastructure, or limited linguistic coverage might face algorithmic marginalization within AI-mediated ecosystems, where agents perform less effectively and consumers receive systematically weaker recommendations.

“Markets with thin data, weak infrastructure, or limited linguistic coverage might face algorithmic marginalization within AI-mediated ecosystems, where agents perform less effectively and consumers receive systematically weaker recommendations.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based). May combine two claims.

Visibility alone does not determine outcomes; even within the right ecosystem, demand must still be routed and allocated.

“Yet visibility alone does not determine outcomes.Even within the right ecosystem, demand must still be routed and allocated.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

In developing markets, where product information is often sparse or available only in limited languages, generative systems fill representational gaps.

“In developing markets, where product information is often sparse or available only in limited languages, generative systems fill representational gaps.”

Probably not a substantial claim: may not pass S4 (based).

Segmentation and targeting rely on stable identities, not hybrid actors.

“Segmentation and targeting rely on stable identities, not hybrid actors.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based). May combine two claims.

In this age, success no longer hinges on entering a country; it might hinge on entering the model that defines what consumers in that country see.

“In this age, success no longer hinges on entering a country.It might hinge on entering the model that defines what the consumers in that country see.”

Probably not a substantial claim: may not pass S4 (based).

If models consistently privilege certain features in automated comparisons, firms that align with those features may gain disproportionate algorithmic shelf space in downstream recommendation systems.

“If models consistently privilege certain features in automated comparisons, firms that align with those features may gain disproportionate algorithmic shelf space in downstream recommendation systems.”

May combine two claims.

Intelligent systems presort, pre-shape, and sometimes pre-decide what appears.

“Intelligent systems presort, pre-shape, and sometimes pre-decide what appears.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

Hybrid consumers operate within fields of view that are shaped long before they begin to search.

“Hybrid consumers operate within fields of view that are shaped long before they begin to search.”

May combine two claims.

AI changes more than consumer behavior; it reconfigures how markets work and how firms compete.

“AI changes more than consumer behavior.It reconfigures how markets work and how firms compete.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

The hardening of algorithmic borders depends on infrastructural maturity; in markets with thin data or limited AI penetration they form only partially.

“The hardening of these borders depends on infrastructural maturity, however, and in markets with thin data or limited AI penetration they form only partially.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

Fragmentation across domestic AI ecosystems may complicate challenges for firms.

“Fragmentation across domestic AI ecosystems may complicate these challenges.”

Probably not a substantial claim: may not pass S4 (based).

Even well-managed brands can find themselves surrounded by synthetic content that shapes impressions in ways they did not anticipate.

“Even well-managed brands can find themselves surrounded by synthetic content that shapes impressions in ways they did not anticipate.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

Global strategy must account for fragmented AI systems that apply different priors, rules, and narratives.

“Global strategy must account for fragmented AI systems that apply different priors, rules, and narratives.”

May combine two claims.

Demand is no longer formed only between firms and consumers; it now emerges inside AI ecosystems.

“Demand is no longer formed only between firms and consumers.It now emerges inside AI ecosystems.”

Probably not a substantial claim: may not pass S4 (based).

Consumers no longer enter the marketplace as solitary decision makers.

“Consumers no longer enter the marketplace as solitary decision makers.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

Patterns of delegation likely vary with trust norms, uncertainty avoidance, beliefs about authority, and local scripts around convenience or expertise.

“Patterns of delegation likely vary with trust norms, uncertainty avoidance, beliefs about authority, and local scripts around convenience or expertise.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

Algorithmic borders arise from the coded and cultural materials embedded in AI systems.

“Algorithmic borders arise from the coded and cultural materials embedded in AI systems.”

Probably not a substantial claim: may not pass S4 (based).

The conceptualization focuses on consumption episodes in which AI systems materially shape the set of choices, the ordering of alternatives, the evaluative frame, or the execution of purchase.

“The conceptualization we develop focuses on consumption episodes in which AI systems materially shape the set of choices, the ordering of alternatives, the evaluative frame, or the execution of purchase.”

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based). May combine two claims.

At the societal level, AI-mediated commerce may introduce new forms of inequality, raise questions about sovereignty, and reshape long-standing governance debates.

“At the societal level, AI-mediated commerce may introduce new forms of inequality, raise questions about sovereignty, and reshape long-standing governance debates.”

May combine two claims.

The recursive architecture is both a representation mechanism and a governance mechanism; representation and allocation reinforce each other on each pass.

“It is therefore both a representation mechanism, in that it governs how products are depicted, and a governance mechanism, in that it governs how demand is allocated; the two are difficult to separate because representation and allocation reinforce each other on each pass.”

Probably not a substantial claim: may not pass S4 (based). May combine two claims.

Intelligent systems walk beside consumers, interpret their signals, and influence the earliest stages of preference formation.

“Intelligent systems walk beside them, interpret their signals, and influence the earliest stages of preference formation.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

The rise of the hybrid consumer challenges long-standing assumptions about how preferences form, how attention is allocated, and how identity evolves.

“The rise of the hybrid consumer challenges long-standing assumptions about how preferences form, how attention is allocated, and how identity evolves.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based). May combine two claims.

The hybrid entity is the foundation of the wider transformations; once the actor changes, everything else in the marketplace changes with it.

“This hybrid entity is the foundation of the wider transformations that follow, because once the actor changes, everything else in the marketplace changes with it.”

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing). May combine two claims.

Firms might require new roles such as a Chief Algorithm Officer, cross-functional teams dedicated to platform relations, or governance structures centered on model-centric markets.

“Firms might require new roles such as a Chief Algorithm Officer, cross-functional teams dedicated to platform relations, or governance structures centered on model-centric markets (Campbell et al., 2025).”

• cites Campbell et al., 2025 sourceNot checked — no full text of the source

May combine two claims.

Cultural context does not disappear; it becomes embedded in training data, regulatory regimes, and domestic AI stacks.

“Cultural context, long central to global marketing (Cayla & Eckhardt, 2008), does not disappear.It becomes embedded in training data, regulatory regimes, and domestic AI stacks, as recent work demonstrates in generative AI's culturally patterned outputs (Lu et al., 2025).”

• cites Cayla & Eckhardt, 2008 sourceNot checked — no full text of the source
• cites Lu et al., 2025Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

May combine two claims.

Resistors maintain strong human primacy, limit inference, avoid deep personalization, and favor transparency and control.

“A third group, resistors, maintains strong human primacy.They limit inference, avoid deep personalization, and favor transparency and control.”

Classified as probably not a substantial claim.

The four transformations correspond to four analytically distinct levels: the actor (the hybrid consumer), the ecosystem (algorithmic borders), the gatekeeping layer (intermediated demand), and representation (generative consumption).

“Applying these criteria, the four transformations correspond to four analytically distinct levels: the actor (the hybrid consumer), the ecosystem (algorithmic borders), the gatekeeping layer (intermediated demand), and representation (generative consumption).”

Probably not a substantial claim: may not pass S4 (based). May combine two claims.

The four transformations operate at different levels of the marketplace but form a connected system that now shapes global demand.

“They operate at different levels of the marketplace but form a connected system that now shapes global demand.”

Classified as probably not a substantial claim.

The four transformations were retained against explicit criteria: recurrence across more than one type of input, structural change in demand formation at a distinct marketplace level, inadequate treatment by existing marketing theory, and particular consequences for global marketing.

“From this synthesis we identified candidate transformations and retained four against explicit criteria.To be included, a transformation had to recur across more than one type of input; mark a structural change in how demand forms, located at a distinct level of the marketplace; be inadequately addressed by existing marketing theory; and carry particular consequences for global marketing.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

Collaborators treat AI as a thinking partner, not a proxy, exploring possibilities through dialogue, testing alternatives, and refining preferences while keeping final authority.

“Others act as collaborators, treating AI as a thinking partner not a proxy.They explore possibilities through dialogue, test alternatives, and refine preferences with the system's help while keeping final authority.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based). May combine two claims.

As AI becomes a persistent coauthor of preference and choice, some consumers might begin to treat agents as extensions of the self, potentially giving rise to new loyalty patterns directed toward systems not brands.

“As AI becomes a persistent coauthor of preference and choice, some consumers might begin to treat agents as extensions of the self, potentially giving rise to new loyalty patterns directed toward systems not brands.”

May combine two claims.

The layers of AI-mediated demand interact recursively: generative outputs influence ranking systems; ranking systems structure exposure; exposure shapes behavioral data; behavioral data retrain models.

“Crucially, these layers interact recursively.Generative outputs influence ranking systems.Ranking systems structure exposure.Exposure shapes behavioral data.Behavioral data retrain models.”

May combine two claims.

In Europe, strict rules around data use and AI deployment limit behavioral inference and constrain personalization, producing less tailored but more tightly governed representations.

“In Europe, strict rules around data use and AI deployment limit behavioral inference and constrain the degree of personalization that models can provide.The resulting representations are less intensely tailored, but also more tightly governed, which produces a different informational climate.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

AI-mediated consumption may require extending trust theory to models that personalize, agents that act autonomously, and platforms that filter visibility.

“AI-mediated consumption may require extending trust theory to models that personalize, agents that act autonomously, and platforms that filter visibility.”

May combine two claims.

Platform scholarship has shown how control over digital infrastructure reshapes competitive dynamics, yet the implications for global demand formation remain underdeveloped.

“Platform scholarship has shown how control over digital infrastructure reshapes competitive dynamics (Van Alstyne et al., 2016), yet the implications for global demand formation remain underdeveloped.”

• cites Van Alstyne et al., 2016Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

As habits form, systems become woven into routines and quietly assume more agency in shaping choices.

“As habits form, systems become woven into routines and quietly assume more agency in shaping choices.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

In a generative environment, brands must also translate for models, making themselves understandable and usable to the systems that explain them to consumers.

“But in a generative environment, brands must also translate for models.They must make themselves understandable, and usable, to the systems that explain them to consumers.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

In the age of intermediated demand, the most important audience is often not the consumer but the system interpreting the consumer.

“In the age of intermediated demand, the most important audience is often not the consumer but the system interpreting the consumer.”

The quote only partly states this claim.

Across individuals, consumer-AI symbiosis takes recognizable forms: delegators, collaborators, and resistors.

“Across individuals, this symbiosis takes recognizable forms, as summarized in Table 3.Some consumers behave as delegators, readily offloading search, evaluation, and even parts of the choice process to AI systems that manage complexity on their behalf.”

May combine two claims.

Firms become dependent on ecosystems they do not own and cannot meaningfully negotiate with, creating new tensions around control, accountability, and value capture.

“Firms become dependent on ecosystems they do not own and cannot meaningfully negotiate with, which creates new tensions around control, accountability, and value capture.”

Probably not a substantial claim: may not pass S4 (based).

Synthetic influencers and AI-generated spokespersons can redirect attention across borders through narratives crafted without managerial oversight.

“Synthetic influencers and AI-generated spokespersons can redirect attention across borders through narratives crafted without managerial oversight.”

May combine two claims.

Consumer-AI symbiosis is neither uniform nor frictionless and is increasingly foundational to how demand forms in global markets.

“This variation reinforces that consumer-AI symbiosis is neither uniform nor frictionless.It is increasingly foundational to how demand forms in global markets.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

Consumer expectations evolve faster than firms can monitor.

“Consumer expectations evolve faster than firms can monitor.”

Probably not a substantial claim: may not pass S4 (based).

Generative AI has become the primary architect of the informational environment consumers encounter.

“Generative AI has become the primary architect of this environment.”

Probably not a substantial claim: may not pass S4 (based).

Generative systems actively construct category logics by recombining attributes, ranking dimensions, and narrative tropes from training corpora, foregrounding certain features, normalizing comparison criteria, and defining relevance within a category.

“Generative systems do not merely summarize existing content.They actively construct category logics.By recombining attributes, ranking dimensions, and narrative tropes drawn from training corpora, they foreground certain product features, normalize comparison criteria, and implicitly define what counts as relevant within a category.”

May combine two claims.

Visibility becomes unevenly distributed, and the divide shifts from developed versus developing markets to high-versus low-visibility AI environments.

“Visibility becomes unevenly distributed, and the divide shifts from developed versus developing markets to high-versus low-visibility AI environments.”

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).

These dependencies could heighten exposure to subtle nudges that steer choices along predictable decision paths.

“These dependencies could heighten exposure to subtle nudges that steer choices along predictable decision paths.”

Probably not a substantial claim: may not pass S4 (based). May combine two claims.

Even the questions consumers ask are shaped by AI; systems propose prompts, suggest alternatives, and guide interaction toward decision frames they construct on the fly.

“Even the questions consumers ask are shaped by AI.Systems propose prompts, suggest alternatives, and guide the interaction toward decision frames they construct on the fly.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

Today, brands compete across ecosystems of models, platforms, and infrastructures, not only for space in the consumer's mind.

“Instead of competing only for space in the consumer's mind, brands now compete across ecosystems of models, platforms, and infrastructures.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

If models begin to describe a category as offering features or benefits that no brand supplies, the norm can move overnight.

“If models begin to describe a category as offering certain features or benefits that no brand supplies, the norm can move overnight.”

Probably not a substantial claim: may not pass S4 (based). May combine two claims.

Computational borders influence which products appear in early-stage discovery, which categories feel salient, and which brand meanings gain traction.

“They influence which products appear in early-stage discovery, which categories feel salient, and which brand meanings gain traction.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

Firms accustomed to managing meaning through controlled communications now face a generative landscape where meaning is continuously rewritten by external systems.

“Firms accustomed to managing meaning through controlled communications now face a generative landscape where meaning is continuously rewritten by external systems.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

In creator-driven markets, generative tools amplify established remix cultures.

“In creator-driven markets, generative tools amplify established remix cultures.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

The result is a new consuming entity: a pairing of human intention and machine inference that evaluates, anticipates, and chooses jointly.

“The result is a new consuming entity: a pairing of human intention and machine inference that evaluates, anticipates, and chooses jointly.”

May combine two claims.

Intermediation of demand is strongest in platform-mediated, information-rich categories, and weaker where discovery and comparison remain offline or relationship-based.

“Intermediation of this kind is strongest in platform-mediated, information-rich categories, and weaker where discovery and comparison remain offline or relationship-based.”

Probably not a substantial claim: may not pass S4 (based).

In some markets, computational borders determine whether a foreign brand enters the consideration set at all.

“In some markets, they determine whether a foreign brand enters the consideration set at all.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

As algorithmic borders harden, demand formation fragments across AI systems that carry different cultural priors and local behavioral norms.

“As these borders harden, the consequences for global marketing become clear.Demand formation fragments across AI systems that carry different cultural priors and local behavioral norms.”

May combine two claims.

Unlike earlier forms of digital intermediation, contemporary AI systems do not simply retrieve or rank existing options; they predict, generate, and adapt in real time, influencing which products are encountered and how categories are framed and compared.

“Unlike earlier forms of digital intermediation, contemporary AI systems do not simply retrieve or rank existing options.They predict, generate, and adapt in real time, influencing not only which products are encountered but how categories are framed and compared (Kietzmann et al., 2018).”

• cites Kietzmann et al., 2018 sourceNot checked — no full text of the source

May combine two claims.

Marketing science has documented the transformation of service systems through analytics and automation but has not fully theorized what happens when intelligent systems actively constrain legibility, visibility, and legitimacy in ways that structure consumer choice environments.

“Similarly, marketing science has documented the transformation of service systems through analytics and automation (Rust & Huang, 2014) but has not fully theorized what happens when intelligent systems actively constrain legibility, visibility, and legitimacy in ways that structure consumer choice environments (Kaliyamurthy & Schau, 2025).”

• cites Rust & Huang, 2014 sourceNot checked — no full text of the source
• cites Kaliyamurthy & Schau, 2025 sourceNot checked — no full text of the source

May combine two claims.

Data governance regimes dictate what systems may collect or infer, limiting the behavioral signals available in some regions and expanding them in others.

“Data governance regimes dictate what systems may collect or infer, limiting the behavioral signals available in some regions and expanding them in others.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

In India, commerce is routed through national digital rails such as UPI and WhatsApp Shopping, which privilege domestic platforms and are increasingly tied to India's own AI stacks.

“In India, commerce is routed through national digital rails such as UPI, the country's real-time government-backed payments infrastructure, and WhatsApp Shopping.These systems privilege domestic platforms and are increasingly tied to India's own AI stacks, which shape how products are classified, interpreted, and recommended.”

May combine two claims.

Platform and intermediation research explains how infrastructures allocate exposure and capture value but under-theorizes how those infrastructures shape preference formation upstream of choice.

“Platform and intermediation research explains how infrastructures allocate exposure and capture value (Van Alstyne et al., 2016) but under-theorizes how those same infrastructures shape preference formation upstream of choice.”

• cites Van Alstyne et al., 2016Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Classified as probably not a substantial claim.

Brand equity may become more fragile when meaning is coauthored by generative systems.

“Brand equity may also become more fragile when meaning is coauthored by generative systems.”

Probably not a substantial claim: may not pass S4 (based).

A growing divide between high-AI-visibility and low-AI-visibility regions could emerge as a new digital inequality.

“A growing divide between high-AI-visibility and low-AI-visibility regions could emerge as a new digital inequality.”

Probably not a substantial claim: may not pass S4 (based).

Attributes underrepresented in training data may fade from consumer consideration altogether.

“Conversely, attributes that are underrepresented in training data may fade from consumer consideration altogether.”

Probably not a substantial claim: may not pass S4 (based).

The informational and symbolic material through which consumers interpret products is itself increasingly generated rather than retrieved.

“The informational and symbolic material through which consumers interpret products is itself increasingly generated rather than retrieved.”

Classified as probably not a substantial claim. May combine two claims.

The regional cases are bounded illustrations; conditions vary within each region as well as between them, and no market reduces to a single mechanism.

“We offer these as bounded illustrations whereby conditions vary within each region as well as between them.We do not suggest that any market reduces to a single mechanism; the point is that algorithmic borders form through several distinct routes that a uniform comparison would obscure.”

Probably not a substantial claim: may not pass S4 (based). May combine two claims.

Generative systems construct the local texture of consumption, influencing how categories are understood and which meanings rise to the surface, and feed outputs back into the wider ecosystem as new training data, user prompts, and engagement signals that shape future model behavior.

“Across these settings, generative systems do more than translate.They construct the local texture of consumption, influencing how categories are understood and which meanings rise to the surface and feeding those outputs back into the wider ecosystem as new training data, user prompts, and engagement signals that subsequently shape future model behavior.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

Where cultural translation once aimed at people, model localization aims at the systems that represent the brand to people.

“Where cultural translation once aimed at people, model localization aims at the systems that represent the brand to people.”

May combine two claims.

Classic consumer decision-making research centers the bounded individual consumer and treats preferences as internally generated, leaving it with little to say once inference is distributed across human and machine.

“Classic consumer decision-making research centers the bounded individual consumer and treats preferences as internally generated (Bettman, 1979;Wedel & Kamakura, 2000) and so has little to say once inference is distributed across human and machine.”

• cites Bettman, 1979Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• cites Wedel & Kamakura, 2000 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S2 (checkable), S4 (based). May combine two claims.

AI-mediated consumption creates new forms of vulnerability and raises new questions about exposure and dependency.

“AI-mediated consumption creates new forms of vulnerability and raises new questions about exposure and dependency.”

Probably not a substantial claim: may not pass S4 (based).

The central theoretical claim of the article is that demand formation is relocating from the bounded consumer to a distributed human-AI architecture.

“the central theoretical claim of this article: that demand formation is relocating from the bounded consumer to a distributed human-AI architecture.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based). May combine two claims.

Hybrid consumers are not stable categories; they are configurations whose decisions emerge from the interplay of human intention, agent inference, platform defaults, and cultural norms around delegation.

“Yet hybrid consumers are not stable categories.They are configurations.Their decisions emerge from the interplay of human intention, agent inference, platform defaults, and cultural norms around delegation.”

May combine two claims.

Delegation is neither universal nor unconditional.

“Research on algorithm aversion and resistance (Castelo et al., 2019;Longoni et al., 2019) reminds us that delegation is neither universal nor unconditional.”

• cites Castelo et al., 2019Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• cites Longoni et al., 2019 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

Taken together, the four transformations reshape the structural conditions of global demand.

“Taken together, these four transformations reshape the structural conditions of global demand.”

Probably not a substantial claim: may not pass S4 (based).

In China, a Western brand may be invisible not because it is restricted but because domestic models trained on local cultural material deprioritize unfamiliar categories.

“In China, a Western brand may be invisible not because it is restricted but because domestic models trained on local cultural material deprioritize unfamiliar categories.”

Probably not a substantial claim: may not pass S2 (checkable).

Taken together, these shifts redefine the structural conditions under which global marketing is practiced.

“Taken together, these shifts redefine the structural conditions under which global marketing is practiced.”

Probably not a substantial claim: may not pass S2 (checkable).

Generative representation is a recursive layer that interacts with algorithmic borders and intermediated demand.

“In this way, generative representation is not a terminal layer but a recursive one that interacts with algorithmic borders and intermediated demand.”

Probably not a substantial claim: may not pass S4 (based).

Brand visibility now depends on a firm's ability to influence and integrate with algorithmic platform and model ranking logic, recommendation criteria, API access points, and data formats that models can ingest.

“Brand visibility now depends on a firm's ability to influence and integrate with algorithmic platform and model ranking logic, recommendation criteria, API access points, and the data formats that models can ingest.”

Probably not a substantial claim: may not pass S2 (checkable).

Competition becomes an ecosystem strategy.

“Competition becomes an ecosystem strategy.”

Probably not a substantial claim: may not pass S4 (based). May combine two claims.

In the European Union, GDPR and the AI Act restrict behavioral inference, forcing models to rely on narrower forms of personal data, producing different recommendation patterns and altering the balance between personalization and privacy.

“In the European Union, GDPR and the AI Act restrict behavioral inference, forcing models to rely on narrower forms of personal data.This constraint produces different recommendation patterns and alters the balance between personalization and privacy.”

May combine two claims.

Algorithms do not simply assist experience; they delimit and structure it.

“In this view, algorithms do not simply assist experience.They delimit and structure it.”

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).

Comparison tables may emphasize attributes the brand never intended to prioritize, shifting the dimensions on which it is judged.

“Comparison tables may emphasize attributes the brand never intended to prioritize, shifting the dimensions on which it is judged.”

May combine two claims.

The recursive loop operates at more than one site: it revises category meaning when retrained models change how a category is described and compared, shifts demand allocation when those revisions feed ranking and exposure, and shapes consumer learning when the resulting environment alters what consumers come to expect.

“This loop operates at more than one site: it revises category meaning when retrained models change how a category is described and compared, it shifts demand allocation when those revisions feed ranking and exposure, and it shapes consumer learning when the resulting environment alters what consumers come to expect.”

May combine two claims.

People resist algorithmic influence when choices feel identity-relevant, dignity-relevant, or when autonomous judgment is especially valued.

“Their stance often reflects concerns documented in consumer research showing that people resist algorithmic influence when choices feel identity-relevant, dignity-relevant, or when autonomous judgment is especially valued (Castelo et al., 2019;Longoni et al., 2019).”

• cites Castelo et al., 2019Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• cites Longoni et al., 2019 sourceNot checked — no full text of the source

Classified as probably not a substantial claim.

These disparities could distort representation, with some populations encountering narrower choice sets or less relevant information.

“These disparities could distort representation in meaningful ways, with some populations encountering narrower choice sets or less relevant information.”

Probably not a substantial claim: may not pass S4 (based). May combine two claims.

In regions with thin data, consumers receive weaker recommendations, narrower choice sets, and relevance signals that misfire.

“In regions with thin data, consumers receive weaker recommendations, narrower choice sets, and relevance signals that misfire.”

Probably not a substantial claim: may not pass S4 (based). May combine two claims.

Consumers gain convenience but trade away transparency and agency, often without knowing what has been surrendered.

“Consumers gain convenience but trade away transparency and agency, often without knowing what has been surrendered.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

In hybrid environments, preferences could emerge iteratively as each model output shapes the next intention.

“In hybrid environments, preferences could instead emerge iteratively as each model output shapes the next intention.”

Probably not a substantial claim: may not pass S4 (based). May combine two claims.

Domestic AI ecosystems embed local priors that can alter how brands are explained, sometimes aligning with strategy and sometimes quietly distorting it.

“Domestic AI ecosystems embed local priors that can alter how brands are explained, sometimes aligning with strategy and sometimes quietly distorting it.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

Cultural understanding is filtered through models with their own priors.

“Cultural understanding is filtered through models with their own priors.”

Classified as probably not a substantial claim. May combine two claims.

The four transformations are mutually constitutive; the actor is where the exposition begins, not a first cause from which the others are derived.

“We therefore treat the four transformations as mutually constitutive: the actor is where we begin the exposition, not a first cause from which the others are derived.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

The hybrid consumer pairing is most pronounced where consumers actively enlist AI in search, comparison, and choice; in habitual or low-involvement consumption it may be peripheral.

“This pairing is most pronounced where consumers actively enlist AI in search, comparison, and choice; in habitual or low-involvement consumption it may be peripheral.”

Probably not a substantial claim: may not pass S2 (checkable). May combine two claims.

Excellence might depend on the ability to design model-compatible content, manage data to strengthen a model's understanding of the brand, cultivate platform diplomacy, and develop API fluency.

“Excellence might depend on the ability to design model-compatible content, manage data in ways that strengthen a model's understanding of the brand, cultivate platform diplomacy, and develop API fluency (Campbell et al., 2020).”

• cites Campbell et al., 2020Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

May combine two claims.

Consumer culture theory takes meaning and identity seriously yet treats the machine as a backdrop to consumption, not as an active participant in the choice architecture.

“Consumer culture theory takes meaning and identity seriously (Arnould & Thompson, 2005;Belk, 2013) yet still treats the machine as a backdrop to consumption not as an active participant in the choice architecture.”

• cites Arnould & Thompson, 2005 sourceNot checked — no full text of the source
• cites Belk, 2013 sourceNot checked — no full text of the source

May combine two claims.

The hybrid consumer cannot be understood as a single global profile; it is a fluid identity structure shaped by human agency, model prediction, cultural norms, and platform design.

“The hybrid consumer therefore cannot be understood as a single global profile.It is a fluid identity structure shaped by the dance between human agency, model prediction, cultural norms, and platform design.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

These orientations appear across markets, though their prevalence varies by context.

“These orientations appear across markets, though their prevalence varies by context.”

May combine two claims.

The consumer remains agentic yet operates within boundaries partially set by infrastructural design.

“The consumer remains agentic yet operates within boundaries partially set by infrastructural design.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

Generative patterns vary widely across regions because they reflect local creative norms, infrastructural capacity, and regulatory constraints.

“These generative patterns vary widely across regions because they reflect local creative norms, infrastructural capacity, and regulatory constraints.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

In many cases, the system becomes the default navigator, which can create either comfort or subtle dependence.

“In many cases, the system becomes the default navigator, which can create either comfort or subtle dependence.”

May combine two claims.

Algorithmic borders are shaped by the data on which models are trained and by the institutional environments in which they operate.

“They are shaped by the data on which models are trained and by the institutional environments in which they operate.”

May combine two claims.

AI was marketed as a tool for personalization, but it is becoming a participant.

“AI was marketed as a tool for personalization, a gentle assistant designed to help consumers navigate abundance.Instead, it is becoming a participant (Mo et al., 2026).”

• cites Mo et al., 2026Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

These transformations apply most strongly in high-information, platform-mediated categories; their reach is conditioned by infrastructural maturity, regulatory variation, and sectoral differences.

“These transformations apply most strongly in high-information, platform-mediated categories; their reach is further conditioned by infrastructural maturity, regulatory variation, and sectoral differences, which we flag as each transformation is introduced and develop as boundary conditions in the conclusion.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

Cognitive offloading could gradually reduce competence in certain tasks, while reliance on agents might develop unevenly across socioeconomic groups.

“Cognitive offloading could gradually reduce competence in certain tasks, while reliance on agents might develop unevenly across socioeconomic groups.”

Probably not a substantial claim: may not pass S2 (checkable). May combine two claims.

The goal is not only to be meaningful in culture but also to be seen by the infrastructures that determine visibility and access.

“The goal is not only to be meaningful in culture but also to be seen by the infrastructures that determine visibility and access.”

Probably not a substantial claim: may not pass S2 (checkable).

The four transformations synthesize this shift into a systemic account of AI-mediated demand.

“Taken together, the four transformations developed in this article synthesize this shift into a systemic account of AI-mediated demand.”

Probably not a substantial claim: may not pass S4 (based).

The traveler scrolling through hotel options is encountering a field of visibility already shaped by model training data, ranking logics, and platform incentives.

“The traveler scrolling through hotel options is not simply choosing among listings.She is encountering a field of visibility already shaped by model training data, ranking logics, and platform incentives.”

Probably not a substantial claim: may not pass S4 (based).

The framework applies most strongly in high-information, platform-mediated categories where search, comparison, and evaluation are algorithmically shaped.

“First, the framework applies most strongly in high-information, platform-mediated categories where search, comparison, and evaluation are algorithmically shaped.”

Probably not a substantial claim: may not pass S4 (based).

In Jakarta, the commuter who reorders groceries delegates part of decision authority to a system that remembers, predicts, and prioritizes.

“In Jakarta, the commuter who reorders groceries delegates part of decision authority to a system that remembers, predicts, and prioritizes.”

May combine two claims.

AI symbiosis is structured, but not deterministic.

“AI symbiosis is structured, but not deterministic.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

In Mumbai, the shopper's personalized storefront does not merely reflect her preferences; it recalibrates them as she moves through it.

“In Mumbai, the shopper's personalized storefront does not merely reflect her preferences.It recalibrates them as she moves through it.”

Probably not a substantial claim: may not pass S4 (based).

In Berlin, the student comparing loan products receives not a neutral set of alternatives, but a machine-synthesized framing of risk and relevance.

“In Berlin, the student comparing loan products receives not a neutral set of alternatives, but a machine-synthesized framing of risk and relevance.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

Sectoral variation matters; the analysis centers on commercial consumption, and public services and highly regulated domains may follow distinct governance logics.

“Finally, sectoral variation matters.Our analysis centers on commercial consumption; public services and highly regulated domains may follow distinct governance logics.”

Probably not a substantial claim: may not pass S2 (checkable).

Marketplaces may increasingly be shaped not only by human intention but also by systems that learn, infer, and coauthor the conditions of visibility and access.

“Taken together, these dynamics suggest that marketplaces may increasingly be shaped not only by human intention but also by systems that learn, infer, and coauthor the conditions of visibility and access.”

Probably not a substantial claim: may not pass S4 (based).

Regulatory regimes constrain inference and generative deployment, shaping the degree to which representational drift or algorithmic allocation operate.

“Third, regulatory regimes constrain inference and generative deployment, shaping the degree to which representational drift or algorithmic allocation operate.”

Probably not a substantial claim: may not pass S4 (based).

Infrastructural maturity matters; markets characterized by thin data or limited AI penetration may experience partial forms of algorithmic border formation.

“Second, infrastructural maturity matters.Markets characterized by thin data or limited AI penetration may experience partial forms of algorithmic border formation.”

May combine two claims.

These scenes are not anecdotes; they illustrate a structural shift in how global demand forms.

“These scenes are not anecdotes.They illustrate a structural shift in how global demand forms.”