From Integrated Communication to Algorithmic Orchestration: Marketing Communication, Advertising Research, and the Digital Transformation of Persuasion, 2000–2026

1. Introduction

Few domains of marketing thought have been transformed thoroughly by digital media as marketing communication. The discipline that matured around the mass advertisement—built on one-to-many transmission, television dominance, and the reach–frequency calculus—now operates in an environment in which roughly six billion people, close to three-quarters of humanity, use the internet (International Telecommunication Union [ITU], 2025; World Bank, 2026), and approximately 5.8 billion social media user identities are active worldwide (Kemp, 2026). In 2000, at the peak of the dot-com boom, internet advertising in the United States generated US$8.1 billion (Interactive Advertising Bureau [IAB], 2011); in 2025 it generated US$294.6 billion (IAB, 2026). Global advertising revenue reached an estimated US$1.14 trillion in 2025, with pure-play digital channels accounting for roughly 73 percent of the total and commerce media overtaking television for the first time (WPP Media, 2025a, 2025b). The infrastructure of persuasion has shifted decisively from the broadcast tower to the distributed network of platforms, algorithms, creators, and databases.

Yet the transformation is not merely quantitative. Digital media have altered the ontology of marketing communication: who speaks (professional communicators, consumers, influencers, and now generative machines), what constitutes a message (from fixed copy to dynamic, personalized, co-produced, and increasingly machine-generated content), how audiences are reached (from purchased reach to algorithmically mediated attention), and how effects are conceptualized (from linear hierarchies of response to recursive journeys across paid, owned, earned, and shared touchpoints; Kannan & Li, 2017; Lemon & Verhoef, 2016; Schweidel et al., 2022). Foundational constructs—integration, consistency, message control, even the sender–receiver dyad itself—require re-examination under these conditions (Finne & Grönroos, 2009; Vargo & Lusch, 2004), and the diffusion of generative artificial intelligence (AI) since late 2022 has added a further layer, in which the author of a persuasive message may be neither a person nor an organization but a model (Hermann & Puntoni, 2024; Kshetri et al., 2024).

This paper responds to the need for a systematic, analytically rigorous account of marketing communication in this new environment, with particular attention to the field's intellectual relationship with advertising research and practice. Advertising is not only the field's largest tool (Section 7); it is also the domain that supplied marketing communication's oldest research institutions, its measurement conventions, and most of its accumulated empirical generalizations—while simultaneously undergoing its own crisis of accountability in the digital era (Gordon et al., 2019, 2021; Lewis & Rao, 2015). Three features distinguish the present analysis from earlier agenda-setting reviews (Dwivedi et al., 2021; Kannan & Li, 2017; Lamberton & Stephen, 2016). First, it treats the 26-year period from 2000 to 2026 as a single analytic window and anchors its conceptual claims in harmonized statistical series rather than in isolated headline figures. Second, it reads the field through its relationship with advertising research, arguing that the discipline's measurement crisis and its theoretical fragmentation are two faces of the same problem. Third, it integrates the most recent evidence, published between 2024 and 2026, on generative AI, AI disclosure, influencer economics, retail media, attention measurement, and privacy regulation—evidence that did not exist when earlier agendas were written.

The paper is framed by four research questions. RQ1: How has digitalization transformed the conceptual foundations and operational architecture of marketing communication? RQ2: What is the evolving role of advertising within the expanded ecosystem of marketing communication tools? RQ3: How do advertising research and advertising practice relate to marketing communication research and practice, and what does that relationship reveal about the field's condition? RQ4: Which theoretical, measurement, methodological, contextual, technological, ethical, and translational gaps most urgently require scholarly investigation? Section 2 outlines the research design; Section 3 establishes conceptual foundations; Section 4 traces the historical emergence of the discipline; and Section 5 reviews the research frontier through 36 recent studies. Section 6 to 8 present the results: the digital transformation of the field between 2000 and 2026, the repositioning of advertising, and the relationship between advertising research and practice and the marketing communication ecosystem. Section 9 addresses privacy, ethics, and sustainability; Section 10 consolidates the gap analysis; and a concise Discussion (Section 11) and Recommendations (Section 12) precede the Conclusion (Section 13).

2. Research Design and Method

This study is a qualitative, interpretive research paper based on an integrative literature review combined with thematic synthesis design appropriate for generating new conceptual perspectives on a mature but rapidly evolving domain, because it permits simultaneous engagement with empirical studies, theoretical contributions, and practitioner evidence (Snyder, 2019). The analytic window spans the 26 years from 2000 to 2026—from the commercial internet's first advertising cycle to the diffusion of generative AI—while the corpus extends historically to canonical works that established the field's theoretical and measurement foundations (e.g., Borden, 1964; Ehrenberg, 1974/2000; Keller, 1993; Schultz et al., 1993).

The review followed a structured protocol. First, the leading marketing, advertising, and communication journals were designated as anchor outlets, supplemented by searches in Scopus, Web of Science, and Crossref using combinations of the terms "marketing communication," "integrated marketing communications," "advertising effectiveness," "digital marketing," "influencer," "programmatic," "retail media," "attention," "privacy," and "generative artificial intelligence." Second, from the assembled corpus, 20 studies published between 2015 and 2023 were purposively selected as foundation studies using three criteria: direct relevance to the research questions; methodological weight (large-scale field data, experimentation, meta-analysis, or agenda-setting synthesis); and balanced coverage of the four thematic domains that emerged from coding. Third, the same criteria, with the added requirement of publication in a dated issue between January 2024 and mid-2026, identified 16 studies representing the most recent frontier; the bibliographic details of every study were verified against publisher records. Fourth, in keeping with the interpretive tradition, which treats context and triangulation as essential to trustworthiness (Denzin & Lincoln, 2018), harmonized statistical and industry sources were integrated to anchor conceptual claims in current empirical facts: the ITU series on internet use as published by the World Bank; the IAB/PwC series on U.S. internet advertising revenue, which runs continuously from 1996; the WPP Media global advertising forecasts; Data Reporter’s global digital reports; the Association of National Advertisers (ANA) and Incorporated Society of British Advertisers (ISBA) programmatic supply-chain audits; the CMS GDPR Enforcement Tracker; The CMO Survey; and the Gartner CMO Spend Survey, together with official regulatory texts from the European Union and the U.S. Federal Trade Commission (FTC). Where industry figures are estimates, the text identifies their vintage.

The analysis followed the six phases of reflexive thematic analysis articulated by Braun and Clarke (2006): familiarization; initial coding; theme generation; theme review; theme definition; and analytic narrative. Coding was abductive, moving iteratively between the data and sensitizing concepts drawn from IMC theory, relationship communication, service-dominant logic, digital marketing frameworks, and the empirical-generalizations tradition of advertising research (Assmus et al., 1984; Hanssens, 2018; Sethuraman et al., 2011), with triangulation across academic, statistical, and industry sources as the principal validity strategy (Denzin & Lincoln, 2018; Snyder, 2019). Three limitations should be acknowledged: the design is interpretive rather than meta-analytic, so no pooled effect sizes are estimated; the evidence base, like the field itself, is concentrated in North American, European, and Chinese contexts; and industry statistics are periodically revised, as the ITU's 2025 upward revision of its own internet-use estimates illustrates (ITU, 2025). The resulting thematic structure, presented in Section 10 (Figure 5), organizes the paper's findings.

3. Conceptual Foundations: Defining Marketing Communication

Marketing communication may be defined as the strategically orchestrated set of messages, media, tools, and relational activities through which an organization conveys brand-relevant meaning to target audiences and stakeholders in order to inform, persuade, remind, and engage them in ways that advance explicit commercial and relational objectives. Three properties distinguish it from adjacent concepts. Unlike general communication, marketing communication is purposeful: every message is designed to contribute to measurable marketing outcomes. Unlike marketing in its entirety, which concerns the holistic creation, delivery, and exchange of value, it constitutes the voice of the brand: the managed interface between organizational strategy and audience perception. And, most distinctively, it is integrative: its value lies not in any single instrument but in the deliberate coordination of heterogeneous tools—advertising, sales promotion, public relations, direct marketing, personal selling, digital and interactive media, events, and sponsorships—so that they speak with a consistent voice and generate synergistic rather than redundant effects (Pickton & Broderick, 2005; Schultz et al., 1993).

The field's conceptual vocabulary has evolved under digital conditions. McCarthy's (1960) four Ps positioned promotion as one manageable variable within the marketing mix; Lauterborn's (1990) reformulation from the consumer's perspective recast promotion as communication, foregrounding dialogue and meaning negotiation; and Schultz, Tannenbaum, and Lauterborn's (1993) Integrated Marketing Communications named the emergent managerial paradigm. Cornelissen (2001, 2003) later subjected the construct to sustained critique, noting that integration functioned more as normative aspiration than as precisely specified theory. Subsequent work has partly answered that critique by operationalizing integration as a firm-wide organizational capability (Porcu et al., 2017) and by showing that consumers' perceived consistency of communications predicts relationship outcomes (Šerić et al., 2020), yet the under specification Cornelissen diagnosed retains its relevance in an omnichannel era in which touchpoints multiply faster than planning frameworks (Cui et al., 2021; Verhoef et al., 2015).

Three interlocking strategic functions anchor the field's conceptual core. Marketing communication constructs brand equity: Keller's (1993) customer-based brand equity framework demonstrates how communication builds awareness, shapes associations, and anchors perceived quality and loyalty. It fosters relationships: Grönroos (1994) and Gummesson (2002) reframed communication as ongoing dialogue with multiple publics, elaborated in Duncan and Moriarty's (1998) communication-based model of relationship management. And it shapes perception and behavior, drawing on persuasion theory from Lavidge and Steiner's (1961) hierarchy-of-effects model to Petty and Cacioppo's (1986) elaboration likelihood model. Under service-dominant logic (Vargo & Lusch, 2004), these functions are increasingly understood relationally: communication is not message delivery to passive receivers but value co-creation among brands, stakeholders, and—crucially in the digital era—platforms and algorithms (Finne & Grönroos, 2009; Kannan & Li, 2017). The generative turn extends the point: when models can produce messages, the co-creating "stakeholder" may itself be a machine whose outputs consumers evaluate through their own persuasion knowledge (Campbell et al., 2022; Wu, 2026).

4. The Historical Emergence of Marketing Communication as an Independent Discipline

Marketing communication's institutionalization as a distinct discipline resulted from the convergence of commercial practice, academic marketing, and communication science across six identifiable stages, summarized in Table 1. Significantly for the argument developed in Section 8, the measurement institutions of the field were built first by advertising research—recognition and recall testing, media syndication, and panel-based sales linkage—decades before marketing communication acquired a distinct academic identity.

The craft era's commercial antecedents lay in the nineteenth century, when industrial overcapacity and the mass-circulation press created an apparatus for paid persuasion; the early twentieth century professionalized it through figures such as Albert Lasker and Claude Hopkins, whose Scientific Advertising (1923) introduced testing and accountability. Starch's (1914) psychological treatment of advertising and the rise of readership research gave the young practice its first measurement instruments, and academic respectability arrived with the first university marketing courses, the Journal of Marketing (1936), and the American Marketing Association (1937).

The mid-century consolidation was decisive. Television inaugurated mass-mediated brand building, and the marketing mix codified by Borden (1964) and McCarthy (1960), popularized by Kotler (1967), positioned promotion as one manageable variable among four. Communication science supplied the theoretical apparatus—transmission models, the two-step flow, diffusion theory, the Yale persuasion program, and later the elaboration likelihood model (Petty & Cacioppo, 1986)—and dedicated outlets, the Journal of Advertising Research (1960) and the Journal of Advertising (1972), signaled maturing scholarship, though promotional activity remained a sub-function of marketing.

The decisive transition to disciplinary autonomy occurred in the 1980s and 1990s, propelled by structural upheaval: media de-massification fragmented audiences; database technology enabled addressable communication; sales promotion spending overtook media advertising in the United States by the mid-1980s; and agency commission structures yielded to accountability regimes. Conceptually, Ries and Trout (1981) relocated strategy to the mind of the consumer, while the brand equity movement (Aaker, 1991; Keller, 1993) justified managing all communication instruments as investments in a single asset. Schultz et al. (1993) named the emergent paradigm; the Journal of Marketing Communications (1995) and landmark textbooks consolidated curricular identity (Belch & Belch, 2021; Pickton & Broderick, 2005; Shimp, 1990); and Grönroos's (1994) manifesto for relationship marketing, Duncan and Moriarty's (1998) communication-based model, and Vargo and Lusch's (2004) service-dominant logic completed the migration from message delivery to dialogic value co-creation—preparing the field, at least in principle, for the networked conditions of the digital century.

The two most recent stages are the subject of Section 5 and Section 6. Their boundary is marked less by a technology than by a change in the locus of authorship: the algorithmic era (2015–2022) automated the delivery of persuasion through programmatic and recommendation systems, whereas the generative–agentic era that opened with the public release of ChatGPT in November 2022 automates its production (Hermann & Puntoni, 2024; Peres et al., 2023) and, prospectively, its negotiation through AI agents and answer engines (Kim, 2025; Wu, 2026).

5. The Research Frontier, 2015–2026

The conceptual foundations established in Section 3 acquire empirical substance only through engagement with the contemporary research record. This section reviews 36 studies in two cohorts. Table 2 synthesizes 20 foundation studies published between 2015 and 2023, selected according to the criteria described in Section 2; Table 3 synthesizes 16 studies published between 2024 and 2026 that extend, and in places overturn, the earlier record. The narrative develops the review thematically, because the studies' collective significance lies in the pattern they form rather than in any single result.

5.1. Foundation Studies, 2015–2023

The foundation cohort comprises 20 studies published between 2015 and 2023, the period in which programmatic delivery, social platforms, and data-driven attribution became the operating infrastructure of marketing communication. The studies were drawn from the review corpus described in Section 2 and selected for their theoretical reach, methodological rigor, and direct bearing on the four research questions. They span the field's leading outlets in marketing, economics, and advertising, and they combine conceptual syntheses and systematic reviews with large-scale field experiments, panel models, and multi-brand generalizations. Table 2 summarizes each study's domain, method and data, and core finding, keyed to the research question it addresses; Section 5.2 then organizes the cohort into four thematic clusters that structure the analysis in Section 6 to 9.

5.2. Four Clusters in the Foundation Cohort

The foundation studies form four clusters that Section 6 to 9 develop in context. The integration and journey cluster (studies 1–4) supplies what three decades of IMC scholarship, by Cornelissen's (2003) diagnosis, had failed to deliver—a specified mid-range theory of integration (Batra & Keller, 2016)—and effects a quiet paradigm displacement in which the unit of planning is no longer the campaign but the journey (Lemon & Verhoef, 2016), a displacement that Kuehnl et al. (2019) ground in consumers' own perceptions of journey design, Cui et al. (2021) complicate with the informational challenges of omnichannel integration, and Colicev et al. (2018) show to be financially material. The causal cluster (studies 5–10) constitutes the evidentiary core of the accountability crisis analyzed in Section 8: adequately powered advertising experiments often require tens of millions of observations (Lewis & Rao, 2015); observational attribution overstates effects relative to experimental ground truth (Blake et al., 2015; de Haan et al., 2016; Gordon et al., 2019, 2023); and the median television elasticity across 288 brands is approximately 0.01, an order of magnitude below the meta-analytic means of Assmus et al. (1984) and Sethuraman et al. (2011), with publication selection implicated in the earlier record (Shapiro et al., 2021). The infrastructure cluster (studies 11–14) shows that ad blocking and privacy regulation reshape markets, not merely messages—advantaging incumbents (Todri, 2022) and concentrating ad-tech (Johnson et al., 2023)—while trust and disclosure condition persuasion at the micro level (Bleier & Eisenbeiss, 2015; Boerman et al., 2017; Bleier et al., 2020). The new voices cluster (studies 15–20) establishes the influencer's triple identity as channel, endorser, and peer (Hudders et al., 2021; Leung, Gu, & Palmatier, 2022), the inverted-U contingencies that govern influencer effectiveness (De Veirman et al., 2017; Leung, Gu, Li, et al., 2022; Wies et al., 2023), the strategic trajectory and source perception of machine communicators (Campbell et al., 2022; Huang & Rust, 2021; Longoni & Cian, 2022), and the authenticity-alignment typology that distinguishes brand activism from "woke washing" (Vredenburg et al., 2020).

5.3. The Most Recent Record, 2024–2026

The 16 studies synthesized in Table 3 were published after the foundation cohort and address questions that the earlier record could only pose.

The studies fall into five streams. The first, generative creative, converts the speculative promise of Huang and Rust (2021) into measured effects: AI-generated images can exceed professionally made marketing images on quality, realism, and aesthetics (Hartmann et al., 2025); generative, purchase-history-personalized video advertisements lift engagement (Kapoor & Kumar, 2025); a human-in-the-loop large language model writes search advertisements that outperform both human copywriters and unassisted models (Reisenbichler et al., 2026); and evaluators rate AI-generated persuasive content favorably, the residual preference for human work reflecting human favoritism rather than AI aversion (Zhang & Gosline, 2023). The second, AI disclosure, complicates that optimism: Koning and Voorveld (2025), Baek et al. (2026), Gu et al. (2026), and Sands et al. (2026) converge on the finding that telling consumers an advertisement was generated by AI activates persuasion knowledge, reduces credibility and trust, and increases avoidance—particularly for emotional appeals and sincere brands—although perceived human-likeness of the AI, a brand's commitment to the greater good, and the novelty and awe that generative formats evoke (Hou et al., 2026) buffer the penalty. Because the EU AI Act makes marking and labeling of synthetic content mandatory from August 2026 (European Commission, 2026; European Parliament & Council of the European Union, 2024), this stream carries immediate regulatory weight.

The third stream, influencer economics and governance, reverses a decade of intuition: engagement rather than reach drives purchases (Beichert et al., 2024; Yang et al., 2025), stakeholders hold misaligned conceptions of authenticity (Duffek et al., 2025), virtual influencers persuade less on attitudes but not on behavior (Chu et al., 2026), and 96 percent of sponsored posts on Twitter were undisclosed (Ershov et al., 2025). The fourth, privacy and signal loss, quantifies the infrastructure turn through the cost of losing offsite tracking data (Wernerfelt et al., 2025), the revenue losses that Apple's App Tracking Transparency (ATT) imposed on Meta-dependent firms (Aridor et al., 2026), the publisher revenue endangered by cookie restrictions (Miller & Skiera, 2024), and the recorded traffic lost after the GDPR (Goldberg et al., 2024), while Zhu et al. (2025) and Ju et al. (2026) add theory and psychology: privacy restrictions can raise consumer welfare when platforms reset ad rates, and dark patterns in consent interfaces provoke reactance that spills over into behavioral advertising. The fifth, platforms, attention, and retail media, sharpens measurement: most conversions attributable to short-form video advertising occur off-platform (Liang et al., 2025); displayed social endorsements raise clicks mainly for lesser-known brands (Huang & Lin, 2025); viewport logging is validated against eye-tracking (Bruns et al., 2025); retail media networks receive their first systematic conceptualizations (Pauwels & Fagbola, 2025; Thomas et al., 2025); and Kim (2025) and Wu (2026) open the agentic frontier, arguing that persistent, adaptive AI agents challenge exposure-centered advertising theory itself.

5.4. Synthesis: What the Recent Record Shows

Read as a whole, the 36 studies exhibit four patterns that structure the remainder of this paper. First, a causal turn: the field's most influential recent work replaces attitudinal mediation and correlational attribution with experimental and quasi-experimental identification, and the results are systematically deflationary—advertising effects are smaller, rarer, and more conditional than either practitioner dashboards or the earlier academic record implied (Blake et al., 2015; Gordon et al., 2019, 2023; Shapiro et al., 2021). Second, a contingency turn: the strongest new findings take the form of inverted-U relationships and boundary conditions (Bleier & Eisenbeiss, 2015; Huang & Lin, 2025; Leung, Gu, Li, et al., 2022; Longoni & Cian, 2022; Wies et al., 2023) rather than main effects, suggesting the discipline is ready for the mid-range contingent theorizing that Section 12 identifies as its central task. Third, an infrastructure turn: the most consequential variables increasingly reside not in messages or audiences but in the regulatory and technical architecture—ad blockers, consent regimes, tracking identifiers, platform market structure—within which communication occurs (Aridor et al., 2026; Johnson et al., 2023; Todri, 2022; Wernerfelt et al., 2025). Fourth, and newest, a generative turn: machine-produced messages now measurably outperform human-produced ones on several creative and performance criteria, while disclosure of their provenance measurably erodes trust (Hartmann et al., 2025; Koning & Voorveld, 2025). Each pattern is taken up in the results, gap analysis, and recommendations that follow.

6. Results I: The Digital Transformation of Marketing Communication, 2000–2026

This section presents the first set of results, addressing RQ1: how digitalization has transformed the conceptual foundations and operational architecture of marketing communication. It reads the 26-year window from 2000 to 2026 as a cumulative structural transformation rather than a sequence of new channels, drawing on the historical account in Section 4 and the evidence base assembled in Section 5. The argument moves from macro to micro. Section 6.1 documents the quantitative shift in connectivity, datafication, and audience fragmentation that conditions everything else; Section 6.2 and Section 6.3 trace the resulting change in the field's communication model, from linear transmission to networked, data-feedback exchange and from paid media logic to the PESO ecosystem; Section 6.4 and Section 6.5 examine the delegation of brand voice to consumers and influencers; Section 6.6 analyzes the personalization and programmatic infrastructure, the rise of retail media, and the privacy reset that now constrains them; and Section 6.7 turns to generative artificial intelligence and the automation of persuasion itself. Together these subsections establish the transformed ecosystem within which Section 7 and Section 8 reposition advertising and advertising research.

6.1. The Macro-Structural Shift: Connectivity, Datafication, and Fragmentation

The quantitative scale of digital connectivity establishes the context for every subsequent development, and the 26-year window makes the magnitude of the shift legible. Figure 1 plots two harmonized series. In 2005, the first year of the ITU's harmonized global estimate, roughly one billion people—15.6 percent of the world's population—used the internet; by 2025 the figure had reached six billion, or 73.6 percent (ITU, 2025; Our World in Data, 2026; World Bank, 2026). Over the same period, U.S. internet advertising revenue, which had contracted from US$8.1 billion in 2000 to US$6.0 billion in 2002 after the dot-com collapse, grew to US$26.0 billion in 2010, US$139.8 billion in 2020, and US$294.6 billion in 2025 (IAB, 2011, 2026). Advertising revenue grew far faster than the user base after 2010, reflecting the intensification of monetization per user through targeting, video, and commerce media, and the pandemic year 2020 marked an inflection after which social, video, and retail media formats accelerated rather than plateaued.

Table 4 consolidates the key indicators at four points in the window and records the reallocation of advertising money toward platform-mediated formats: social media (US$117.7 billion) overtook search (US$114.2 billion) as the largest U.S. format in 2025, and the top ten companies captured 84.1 percent of U.S. internet advertising revenue (IAB, 2026), while globally pure-play digital media accounted for 73.2 percent of advertising revenue and commerce media (US$178.2 billion) surpassed television (US$167.5 billion) for the first time (WPP Media, 2025a, 2025b).

Three qualitative properties of this environment matter more than any single statistic. First, interactivity collapses the sender–receiver boundary: consumers search, comment, review, create, and share, producing electronic word of mouth whose persuasive dynamics differ fundamentally from paid messaging (Babić Rosario et al., 2020; Hennig-Thurau et al., 2004; Kietzmann et al., 2011). Second, datafication renders communication addressable and measurable at the individual level; budgets have followed, with paid media absorbing 30.6 percent of marketing budgets even as total budgets have flattened at 7.7 percent of company revenue (Gartner, 2025), and 57.6 percent of companies are adding channels, digital channels and retail media foremost among them (Moorman, 2026). Third, fragmentation and avoidance coexist with abundance: attention is scarce, ad blocking is a mass behavior whose welfare consequences are now quantified (Gordon et al., 2021; Todri, 2022), and trust concentrates in earned sources—88 percent of more than 40,000 consumers surveyed across five regions trust recommendations from people they know more than any paid form (Nielsen, 2021). Because ad-blocker adoption depresses spending chiefly on brands the consumer has not previously bought (Todri, 2022), attention scarcity systematically favors incumbents and suppresses discovery.

6.2. From Linear Transmission to Networked, Data-Feedback Communication

These properties amount to a paradigm shift in the architecture of communication, depicted in Figure 2. The classical model of the mass era ran in one direction—from brand, through agency and mass media, to audience—with high message control and slow, aggregate feedback. The digital–algorithmic model is recursive: brands, platforms and their algorithms, creators, owned media, and consumer communities exchange messages in every direction, and a continuous data loop of attribution and optimization re-programs subsequent communication on the basis of observed behavior (Finne & Grönroos, 2009; Kannan & Li, 2017; Kietzmann et al., 2011). Two features of the lower panel are absent from the classical model—algorithmic intermediaries that determine message visibility, and a feedback loop in which the audience's behavior becomes an input to the next message—and the generative era adds a third: the message itself may be produced by a model operating inside the loop (Hartmann et al., 2025; Reisenbichler et al., 2026).

6.3. From Paid Media Logic to the PESO Ecosystem

The classical discipline organized its tools by function (advertising, promotion, public relations, direct marketing). Digital conditions have rendered a media-type taxonomy—the paid, earned, shared, and owned (PESO) framework—analytically indispensable (Dietrich, 2014). Table 5 compares the four media types across control, cost, credibility, and measurement.

The PESO lens clarifies a central empirical finding: owned and earned assets do not merely substitute for paid reach but interact with it. Owned and earned social media exert distinct, sequential effects on consumer mindsets and shareholder value (Colicev et al., 2018), a substantial share of consumers requires exposure to multiple channels before responding (Danaher & Dagger, 2013; Naik & Peters, 2009), and users exposed to short-form video advertising on ByteDance were eight times more likely to convert off-platform than on it (Liang et al., 2025)—so integration across media types is a value-relevant strategic variable, and platform-bounded measurement understates the value of paid media as a trigger for owned and earned conversion.

6.4. Social Media and Electronic Word of Mouth: The Delegation of Voice

Social media restructured marketing communication's power geometry. Kietzmann et al.'s (2011) honeycomb framework—identity, conversations, sharing, presence, relationships, reputation, groups—remains the canonical map of what platforms functionally do; TikTok alone reached one billion monthly users by September 2021 (TikTok, 2021), and more than 30 percent of adult internet users now report discovering new brands through advertisements on social platforms (Kemp, 2025). For brands, the decisive novelty was the delegation of voice: meaning increasingly forms in consumer-to-consumer spaces the brand does not control. Electronic word of mouth—consumer statements made available to multitudes via the internet (Hennig-Thurau et al., 2004)—has become a primary persuasion mechanism with well-documented dynamics: brand-related user-generated content differs systematically across platforms (Smith et al., 2012); sharing follows predictable creative and emotional drivers rather than random virality (Nelson-Field, 2013); eWOM creation, exposure, and evaluation form an integrated process (Babić Rosario et al., 2020); and social endorsement signals persuade conditionally, helping lesser-known brands most (Huang & Lin, 2025).

The strategic consequences are visible in hallmark cases. The ALS Ice Bucket Challenge raised US$115 million in the summer of 2014—communication whose reach and credibility derived entirely from peer participation (ALS Association, 2019)—and Dove's long-running Real Beauty program illustrates the hybrid logic of the era: a paid creative platform architected to generate earned coverage, user-generated conversation, and cultural debate. Social commerce, estimated at US$492 billion in 2021 and projected to reach US$1.2 trillion by 2025 (Accenture, 2022), shows that the commercial infrastructure of shared media has matured accordingly; such cases substantiate, in practice, the claim that digital-era communication is co-produced (Vargo & Lusch, 2004).

6.5. Influencer Marketing: The Professionalization of the Hybrid Voice

Influencer marketing formalizes the delegation of voice into a commercial institution: creator advertising spending in the United States alone reached US$37 billion in 2025 and is projected at US$44 billion for 2026 (IAB, 2026). Influencers occupy a unique structural position, simultaneously media channel, content producer, and peer (Hudders et al., 2021; Leung, Gu, & Palmatier, 2022). Their persuasive advantage rests on perceived authenticity and parasocial relationships, yet their effectiveness is contingent: large followings can signal popularity or commercial co-optation (De Veirman et al., 2017); originality, follower size, and sponsor salience help while activity, fit, and positivity show inverted-U effects (Leung, Gu, Li, et al., 2022); midsize "Goldilocks" influencers are the engagement-optimal tier (Wies et al., 2023); and low-followership targeting outperforms high-followership targeting by an order of magnitude on return on investment, because engagement rather than reach drives purchases (Beichert et al., 2024; Yang et al., 2025). Authenticity, the currency of the channel, is itself contested (Duffek et al., 2025), and virtual influencers—customizable, automatable personas (Mouritzen et al., 2024)—test its limits, persuading less than humans on attitudes but comparably on behavior (Chu et al., 2026).

This hybridity produces a distinctive governance problem: disclosure. The FTC updated its Endorsement Guides in 2023 to address influencers and deceptive reviews explicitly and in 2024 issued a rule banning fake reviews and testimonials, including AI-generated ones (Federal Trade Commission, 2023, 2024). Disclosure activates advertising recognition and can reduce persuasion (Boerman et al., 2017), yet well-designed disclosure language can preserve or even enhance credibility (Evans et al., 2017); compliance, however, lags rules, with 96 percent of sponsored influencer content on Twitter carrying no disclosure (Ershov et al., 2025). Fenty Beauty's 2017 launch—leveraging Rihanna's celebrity and a distributed creator network to generate a reported US$100 million in sales within its first weeks (Robehmed, 2019)—exemplifies the influencer-mediated model at scale.

6.6. Personalization, Programmatic Infrastructure, Retail Media, and the Privacy Reset

Beneath the visible channels lies the algorithmic infrastructure of contemporary practice. Programmatic buying mediates roughly 90 percent of U.S. digital display transactions (eMarketer, 2023) and reached US$162.4 billion in 2025 (IAB, 2026); search and social platforms auction attention in real time; dynamic creative optimization assembles individualized messages; and retail media networks monetize retailer first-party data through closed-loop reporting (Thomas et al., 2025), reaching US$178.2 billion globally in 2025 and projected at US$252.1 billion, or 18 percent of all advertising revenue, by 2030 (WPP Media, 2025a, 2025b). Retail media inverts the buyer–seller relationship between retailers and brands and raises unresolved questions of incrementality, competition, and data transparency (Pauwels & Fagbola, 2025), even as marketplaces can use advertising signals to improve product discovery (Long et al., 2022).

Personalization, in turn, is double-edged: its effectiveness depends critically on trust, and under low trust tailoring amplifies resistance (Bleier & Eisenbeiss, 2015); granting consumers control over their data increases ad effectiveness (Tucker, 2014); and there are boundaries beyond which targeting is experienced as overstepping (John et al., 2018). The institutional environment has responded accordingly. The EU's General Data Protection Regulation established the global template for consent-based data processing (European Parliament & Council of the European Union, 2016). Apple's ATT framework (2021) degraded mobile identifiers—Meta's chief financial officer put the 2022 revenue impact "on the order of $10 billion" (CNBC, 2022), and conversion-optimized Meta advertisements lost 37 percent of their click-through rates while Meta-dependent e-commerce firms suffered revenue declines of 8 to 40 percent (Aridor et al., 2026). Google's decade-long attempt to deprecate third-party cookies in Chrome ended in retreat: a "new path" preserving cookies in July 2024, the abandonment of a standalone user-choice prompt in April 2025, and the retirement of the Topics, Protected Audience, and Attribution Reporting APIs in October 2025 (Chavez, 2024, 2025a, 2025b). The economics of identifiers are now quantified (Miller & Skiera, 2024; Wernerfelt et al., 2025), and the GDPR's combination of reduced vendor use, higher ad-tech concentration, and lost recorded revenue (Goldberg et al., 2024; Johnson et al., 2023) shows that privacy architecture and market structure are jointly determined. Contextual targeting, first-party data strategies, clean rooms, and a renaissance of media mix modeling constitute the industry's adaptive response, with deep implications for how communication effects can be studied (Johnson et al., 2022; Martin & Murphy, 2017).

6.7. Generative Artificial Intelligence and the Automation of Persuasion

Generative AI constitutes the newest and potentially most disruptive layer. The diffusion speed is unprecedented: ChatGPT reached an estimated 100 million monthly users within two months of its November 2022 launch (Hu, 2023) and 900 million weekly users by February 2026 (Malik, 2026). Nearly nine in ten organizations now report regular use of AI in at least one business function, 40 percent of large organizations report scaling AI agents, and marketing and sales is the function in which revenue gains are most commonly reported (McKinsey & Company, 2026); generative AI's potential value has been estimated at US$2.6–4.4 trillion annually, with marketing productivity gains equivalent to 5–15 percent of marketing spending (McKinsey & Company, 2023); and among marketers, AI use has more than tripled since 2022, with AI projected to account for more than half of all marketing activities within three years (Moorman, 2026). Davenport et al. (2020) provide the canonical account of AI's marketing functions; Huang and Rust (2021, 2024) offer a strategic framework in which AI progressively absorbs mechanical, thinking, and feeling tasks—implying that creative strategy itself, long advertising's distinctive human preserve, is increasingly automatable; and Hermann and Puntoni (2024) distinguish the predictive AI of the algorithmic era from generative AI, a distinction that maps onto the shift from optimizing delivery to producing content.

The evidence reviewed in Section 5.3 shows that production has already crossed a quality threshold (Hartmann et al., 2025; Kapoor & Kumar, 2025; Reisenbichler et al., 2026), while consumer-side evidence is more ambivalent: AI-sourced recommendations are trusted more for utilitarian than hedonic judgments (Longoni & Cian, 2022), people rate AI-generated persuasive content highly (Zhang & Gosline, 2023), but disclosure of machine authorship activates persuasion knowledge, reduces trust, and increases avoidance (Baek et al., 2026; Gu et al., 2026; Koning & Voorveld, 2025; Sands et al., 2026). Early brand deployments illustrate the trajectory: Coca-Cola's "Create Real Magic" platform invited consumers to co-create brand imagery using generative tools (Coca-Cola Company, 2023), while Heinz's "A.I. Ketchup" campaign used image-generation outputs as brand-equity evidence (Rethink, 2022).

Conversely, generative AI industrializes deception at scale. Advertisers lost approximately US$84 billion to fraud in 2023—22 percent of digital advertising spending—with losses projected to reach US$172 billion by 2028 (Juniper Research, 2023), and the FTC's 2024 rule explicitly targets AI-generated fake reviews (Federal Trade Commission, 2024). Regulation of provenance is now concrete: Article 50 of the EU AI Act, applicable from 2 August 2026, requires providers to mark synthetic audio, image, video, and text outputs in machine-readable form and deployers to label deepfakes and AI-generated public-interest text (European Commission, 2026; European Parliament & Council of the European Union, 2024). Scholarly frameworks for the ethics of AI-mediated persuasion are emerging (Hermann, 2022; Kshetri et al., 2024), and the agentic frontier—advertising conducted through persistent AI agents and answer engines rather than exposures—has been opened conceptually (Kim, 2025; Wu, 2026), but, as Section 10 argues, theory substantially lags practice.

7. Results II: The Evolving Relationship Between Marketing Communication and Advertising

Advertising—paid, non-personal communication from an identified sponsor—remains the largest single instrument within marketing communication, but its role has been fundamentally repositioned. Historically, advertising was the field's public identity: its message control, scale, repetition, and long-run equity accumulation made it the default strategic instrument throughout the mix era (Borden, 1964; Kotler, 1967). That hegemony eroded under exactly the conditions that created the autonomous discipline of marketing communication: audience fragmentation inflated the cost of mass reach; promotion, direct, and public relations spending grew; databases enabled addressable persuasion; and clients demanded measurable returns that exposure metrics could not supply. IMC (Schultz et al., 1993) formalized the reallocation of authority—the advertiser's "big idea" became one voice in a coordinated portfolio planned outside-in from customer data rather than from creative intuition. Table 6 summarizes the comparative position of advertising relative to the instruments with which it must now be choreographed.

Two empirical regularities now discipline the advertising-versus-other-tools debate. First, synergy is real and quantifiable: hierarchical modeling demonstrates online–offline interactions that exceed the sum of isolated effects (Naik & Peters, 2009), multichannel studies show response concentrated among consumers reached across several channels (Danaher & Dagger, 2013), television advertising causally shifts online search and purchase within minutes of airing (Liaukonyte et al., 2015), and its retail effects vary spatially in ways location-based planning can exploit (Luo & Ranjan, 2025)—mass media functioning empirically as the anchor that digital channels harvest, with Batra and Keller (2016) supplying the corresponding planning theory. Second, time horizon matters: analysis of the IPA databank shows that communication weighted toward long-term brand building (the canonical "60/40" brand-to-activation split) outperforms short-term-only strategies on both revenue and profit over multiyear horizons (Binet & Field, 2013), and the marketing–finance literature documents that brand investments influence firm value through channels that short-window metrics do not capture (Edeling et al., 2021). The dominant contemporary tension—brand building versus performance harvesting—is therefore an allocation problem with measurable stakes, intensified by attribution systems that favor last-click measurability over equity accumulation (Kannan & Li, 2017).

Convergence and divergence coexist. Convergence appears where advertising dissolves into other forms: native advertising, branded content, sponsorship, and influencer collaborations are simultaneously advertising, public relations, and word of mouth—generating synergistic hybridity but also regulatory anxiety over disclosure and deception (Boerman et al., 2017; Evans et al., 2017; Federal Trade Commission, 2023; Sahni & Nair, 2020). Divergence appears where communication separates from brand meaning entirely: programmatic performance campaigns optimized solely for conversion represent advertising reduced to distribution mechanics, and retail media, now larger than television, embeds advertising inside the transaction itself (WPP Media, 2025b). The strategic synthesis emerging in practice (Lemon & Verhoef, 2016; Verhoef et al., 2015) assigns advertising the role not of hegemon but of anchor: supplier of the durable narrative that faster, smaller, more personalized instruments extend across the journey.

8. Results III: Advertising Research and Practice in the Marketing Communication Ecosystem

Because advertising historically dominated both the practice and the study of promotional activity, an adequate analysis of marketing communication's current condition requires examining advertising not only as a tool (Section 7) but as a research tradition and a professional practice system. Advertising research built the measurement conventions and accumulated the empirical generalizations upon which communication effectiveness claims still rest, while advertising practice built the institutional machinery—agencies, media auditors, testing laboratories—through which marketing communication is executed. Digital media have simultaneously fused and strained this relationship.

8.1. The Intellectual Lineage of Advertising Research

Advertising research predates marketing communication's academic identity by decades and supplied the field's earliest measurement infrastructure: Starch's recognition testing from the 1930s, Gallup's day-after recall in the 1940s, the Advertising Research Foundation's (founded 1936) measurement conventions, and A. C. Nielsen's retail audits linking advertising weight to sales; by the 1980s, split-cable experiments and single-source panels had established the gold standard for causal claims. Two enduring theoretical positions structured this tradition. Strong theory—implicit in hierarchy-of-effects models—held that advertising persuades by moving consumers through cognitive, affective, and conative stages (Lavidge & Steiner, 1961; Rossiter & Percy, 1997). Weak theory, articulated most influentially by Ehrenberg (1974/2000) and extended into the view of brand advertising as "creative publicity" (Ehrenberg et al., 2002), held that advertising's role is predominantly reinforcement and salience maintenance among existing buyers rather than persuasion of new ones.

The empirical record assembled under this tradition is remarkably sober. Meta-analysis yields mean short-term advertising elasticities of approximately 0.22 (Assmus et al., 1984), revised downward to roughly 0.12 in later generalizations that double with carryover (Sethuraman et al., 2011), and Shapiro et al. (2021) now report a median television elasticity near 0.01 for 288 brands while implicating publication selection in the earlier record—so the field's proudest cumulative achievement is itself implicated in the accountability crisis, and any restored program of empirical generalization (Hanssens, 2018) must build in publication-bias defenses. Roughly half of 389 split-cable-tested commercials produced no measurable short-term sales effect (Lodish et al., 1995); one-week purchase lifts appeared for only a minority of campaigns, though with genuine carryover (Jones, 1995) consistent with Broadbent's (1979) adstock logic; and Vakratsas and Ambler's (1999) landmark review found little support for intermediate hierarchy constructs. Meanwhile, the Ehrenberg-Bass program systematized weak-theory implications into the modern doctrine of mass reach, mental availability, and distinctive assets (Romaniuk, 2018; Romaniuk & Sharp, 2004; Sharp, 2010)—extended to social platforms by Nelson-Field et al. (2013)—which stands in direct theoretical tension with digital targeting orthodoxy.

8.2. Three Research Traditions and Their Convergence

The digital era has produced not one but three partially overlapping research traditions bearing on marketing communication effectiveness, compared in Table 7 and depicted in Figure 3. The classical message–response tradition examines individual advertisements and campaigns through psychology-rooted experiments and survey measures. The integrative marketing communication tradition examines brand-level communication programs through managerial, strategic, and relational constructs. The digital–econometric tradition—largely developed inside platforms and marketing science departments—examines impressions, interactions, and conversion paths through large-scale field experimentation and causal inference. Each possesses characteristic strengths and blind spots; each has claimed authority over the "effectiveness" question; and none alone is adequate to the orchestration problems the field now faces (Hanssens & Pauwels, 2016; Stewart, 2009).

The three traditions are converging, as Figure 3 depicts, not because any one has prevailed but because two catalytic developments require all three at once. Attention measurement—eye-tracking in the laboratory (Pieters & Wedel, 2004; Segijn, 2025) and viewport logging and passive panels at scale (Bruns et al., 2025; Nelson-Field, 2020)—supplies a construct that the message–response tradition theorized, the integrative tradition needs for planning, and the econometric tradition can now measure at the impression level; machine-generated content, likewise, is at once a creative variable, a governance problem, and an experimentally manipulable treatment (Hartmann et al., 2025; Kapoor & Kumar, 2025). Convergence, however, remains incomplete: the traditions are still published in separate journals, train separate doctoral cohorts, and speak to separate practitioner constituencies.

8.3. The Accountability Crisis: Evidence from Advertising Practice

If advertising research-built marketing communication's measurement foundations, contemporary advertising practice is exposing their limits: spending stands at record levels (WPP Media, 2025b), yet confidence in measurement has deteriorated on four fronts.

The first front is supply-chain transparency. The ISBA and PwC (2020) programmatic audit found that only 51 percent of advertiser spend reached publishers as working media, with a 15 percent "unknown delta" that could not be traced at all; a second audit in 2022 reduced the unknown delta to 3 percent through log-level data sharing, but the share of spend reaching publishers remained at roughly half (51 percent; ISBA & PwC, 2023). The larger ANA (2023) study, covering 21 advertisers, US$123 million of spend, and 35.5 billion impressions, concluded that of every dollar entering a demand-side platform, 29 cents is absorbed by transaction costs, a further 35 cents is lost to non-viewable, invalid, non-measurable, and made-for-advertising (MFA) inventory—MFA sites alone captured 21 percent of impressions and 15 percent of spend—and only 36 cents reaches the consumer effectively, as Figure 4 depicts. Layered atop this, estimated global ad fraud losses reached US$84 billion in 2023 (Juniper Research, 2023), and advertisers directed an estimated US$336 million between 2021 and 2024 to eleven news websites that publish health misinformation—US$35.7 million of it from government and health organizations (Patel et al., 2026). These findings transform waste from a rumor into a documented structural property of the digital advertising system (Gordon et al., 2021).

The second front is micro-effectiveness. Because individual-level advertising elasticities are tiny, randomized controlled trials capable of detecting typical returns require samples in the millions (Lewis & Rao, 2015). Gordon et al. (2019) found negligible short-term sales effects for most of 15 large Facebook campaigns and systematic overstatement of effects by observational attribution models; Gordon et al. (2023) confirmed at far larger scale that non-experimental methods are seldom "close enough"; Blake et al. (2015) demonstrated with eBay that brand-keyword search advertisements generated no measurable incremental sales; and Liang et al. (2025) showed that platform-bounded measurement can err in the opposite direction by missing off-platform conversions. Such findings do not show that digital advertising is ineffective; they show that its effectiveness is rare, conditional, and far harder to detect than attribution dashboards suggest—and signal loss compounds the problem by making attribution data sparser (Aridor et al., 2026; Wernerfelt et al., 2025).

The third front is attention. Exposure-based metrics presuppose that exposure implies processing, yet fewer than half of internet users actually fixate on banner advertisements even when measurably exposed (Drèze & Hussherr, 2003), and attention capture and attention transfer to the brand are analytically distinct (Pieters & Wedel, 2004). Television advertisements are far more likely than online video advertisements to be viewed beyond the 75 percent threshold at which exposure approaches full effectiveness, because online ad avoidance truncates viewing time (Bellman et al., 2020), and attention pays off in recall chiefly among light and non-users of a brand (Simmonds et al., 2020). The institutional response—attention metric standardization and the validation of scalable proxies such as viewport logging against eye-tracking (Bruns et al., 2025)—represents the first genuinely new effectiveness construct of the digital era, but its linkage to sales outcomes remains empirically immature.

The fourth front is managerial observability. Marketing budgets have flattened at 7.7 percent of company revenue and training budgets have fallen to 3.8 percent of marketing spending (Gartner, 2025; Moorman, 2026); fewer than half of companies report that marketing and finance work together on growth (Moorman, 2026), despite two decades of accountability rhetoric (Rust et al., 2004; Stewart, 2009). Part of this is construct inflation: "engagement," the era's signature metric, is not a single construct, and its dimensions relate differently to advertising outcomes (Calder et al., 2009). Table 8 compares the principal measurement approaches available to practice, their origins, strengths, and digital-era status.

8.4. Frictions and Synergies Between the Traditions

The reach-versus-targeting controversy crystallizes the friction between traditions. Targeting orthodoxy holds that addressability raises efficiency by concentrating spend on likely buyers; the weak-theory counterargument holds that because most buyers are out of market at any given time (Sharp, 2010), excessive targeting sacrifices the reach on which long-run growth depends (Romaniuk & Sharp, 2004). The evidence is usefully conditional—retargeting works best for new customers exposed to generic content (Lambrecht & Tucker, 2013), obtrusiveness moderates the payoff to targeting (Goldfarb & Tucker, 2011), and the value of targeting data is real but unevenly distributed (Wernerfelt et al., 2025)—suggesting that both traditions capture part of the truth within boundary conditions the field has yet to formalize. A parallel friction runs between creative and data: Binet and Field's (2013) long-run evidence privileges emotional brand building that resists short-cycle optimization, while algorithmic systems optimize toward whatever proxies are measurable—a textbook Goodhart dynamic in which the metric colonizes the objective—and generative AI sharpens the friction by making creative variation cheap enough to be optimized like a media variable (Reisenbichler et al., 2026). There is also a long-standing academic–practitioner gap: advertising practitioners rely principally on trade press and personal experience rather than academic research (Nyilasy & Reid, 2007), and platform-authored white papers and proprietary dashboards now compete with—and often displace—peer-reviewed evidence as the practice's epistemic authority. This translational problem, together with the accountability crisis, motivates the consolidated gap analysis that follows.

9. Critical Issues: Privacy, Ethics, and Sustainability

The analytical themes developed above converge on normative challenges the field can no longer treat as peripheral. Privacy is the central case. The personalization–privacy paradox is documented descriptively but poorly explained theoretically (Martin & Murphy, 2017), and experimental evidence complicates simple trade-off narratives: privacy controls can increase advertising effectiveness (Tucker, 2014), consumers exhibit boundaries beyond which targeting becomes counterproductive (John et al., 2018), and dark patterns in consent interfaces provoke reactance that attaches to subsequent behavioral advertising (Ju et al., 2026). Regulation has hardened: European data protection authorities had imposed 2,685 fines totaling €6.11 billion by March 2026, including the record €1.2 billion penalty against Meta in 2023 and €530 million against TikTok in 2025 (CMS, 2026), while the ATT episode shows that platform-level privacy changes can redistribute billions in communication budgets and measurably reduce the revenue of dependent firms (Aridor et al., 2026; CNBC, 2022). The regulatory response itself carries unintended structural consequences: the GDPR increased ad-tech market concentration (Johnson et al., 2023), so privacy governance and competition policy cannot be treated as separate domains, and whether consumers gain from privacy restrictions depends on how platforms reset advertising prices (Zhu et al., 2025).

Ethics of persuasion technology raises supply-side concerns. Dark patterns, targeting of vulnerable populations, undisclosed sponsorship, and undisclosed machine-generated persuasion have attracted more public debate than rigorous evidence on actual consumer harm—yet the evidence is accumulating: the overwhelming majority of sponsored influencer posts go undisclosed (Ershov et al., 2025), consumers are not deceived by clearly labeled native advertising (Sahni & Nair, 2020), and transparency about machine authorship carries a measurable persuasion cost (Baek et al., 2026; Koning & Voorveld, 2025). Hermann (2022) argues that AI in marketing should be evaluated within an explicit social-good frame of transparency, accountability, and human oversight, and the disclosure research stream (Boerman et al., 2017; Evans et al., 2017) offers a template for quantifying institutional interventions such as the FTC's 2023 Endorsement Guides and 2024 fake-reviews rule and the EU AI Act's 2026 labeling obligations (European Commission, 2026; Federal Trade Commission, 2023, 2024). The documented programmatic leakage to MFA sites, invalid traffic, and misinformation publishers (ANA, 2023; ISBA & PwC, 2020; Patel et al., 2026) is itself an ethical problem of the communication system, not merely an operational one: brands routinely fund misinformation-adjacent inventory without intent or knowledge.

Sustainability communication completes the triad. As firms escalate environmental claims, greenwashing has become a focal construct (de Freitas Netto et al., 2020; Lyon & Montgomery, 2015), and digital channels amplify both green claims and their detection. Brand activism raises adjacent questions: taking a sociopolitical stand can raise or lower firm value depending on its fit with stakeholder values (Bhagwat et al., 2020; Mukherjee & Althuizen, 2020), and consumers distinguish authentic value advocacy from "woke washing" (Vredenburg et al., 2020), whose alignment typology—purpose, values, messaging, and practice must cohere—supplies the field's clearest current instrument for evaluating this boundary. For a discipline whose core function is shaping perception, the boundary between legitimate advocacy and manipulative polarization remains an unresolved theoretical and empirical problem.

10. Current Gaps in Marketing Communication Research

The thematic analysis yields four overarching themes characterizing marketing communication in the digital age—dialogic orchestration, algorithmic mediation, trust and ethics, and convergence-within-fragmentation—each shadowed by research deficiencies that flow from five root causes: fragmentation of the three research traditions documented in Section 8; proprietary enclosure of evidence within platforms; the speed of technological change, now measured in months rather than years; concentration of the evidence base in WEIRD and B2C contexts; and misaligned incentives in both academia (publication over cumulation) and industry (optimization over verification). Table 9 consolidates the resulting taxonomy, extending the framework with two previously under-recognized domains: the measurement gap exposed by the accountability crisis and the translational gap between academic research and practice.

Figure 5 maps the four themes and the cross-cutting foundation on which they rest. Three gaps deserve elaboration because they bind the rest together. First, the theoretical gap has a distinctive historical character: advertising research accumulated genuine empirical generalizations—elasticity ranges, adstock decay, weak-theory regularities—that gave the mass era a cumulative science, whereas digital-era research has produced abundant findings but few stabilized generalizations, and even the mass-era record is now understood to have been inflated by publication selection (Sethuraman et al., 2011; Shapiro et al., 2021). Rebuilding such generalizations under conditions of algorithmic mediation and machine authorship is arguably the discipline's core theoretical task. Second, the measurement gap is not a technical nuisance but a substantive phenomenon: as Section 11 argues, measurement regimes allocate budgets and therefore shape market reality; studying them is studying the field itself. Third, the translational gap completes the circuit: the academic–practitioner divide documented long before digitalization (Nyilasy & Reid, 2007) has been aggravated by platform enclosure of data, meaning that even rigorous findings often cannot be tested, replicated, or applied outside proprietary walls.

Figure 5. Thematic Map of Marketing Communication in the Age of Digital Media. Note. Themes and sub-themes derived through reflexive thematic analysis of the reviewed corpus (Braun & Clarke, 2006); the cross-cutting foundation is the advertising research tradition now strained by the accountability crisis documented in Section 8.3. Figure prepared by the author.

Figure 5. Thematic Map of Marketing Communication in the Age of Digital Media. Note. Themes and sub-themes derived through reflexive thematic analysis of the reviewed corpus (Braun & Clarke, 2006); the cross-cutting foundation is the advertising research tradition now strained by the accountability crisis documented in Section 8.3. Figure prepared by the author.

11. Discussion

Five theses synthesize the analysis. First, the relationship between advertising and marketing communication is best understood as epistemic symbiosis. Advertising research supplies the field's most developed causal methods and its only substantial stock of empirical generalizations—elasticity magnitudes, carryover dynamics, attention mechanics—while marketing communication research supplies the integrative, relational, and strategic frame that single-medium advertising research structurally lacks. The three traditions documented in Section 8 are complementary rather than competing (Figure 3): the orchestration problems that define contemporary practice—when to reach, when to target, when to build, when to harvest—cannot be answered by any single tradition (Hanssens & Pauwels, 2016), and the four turns identified in Section 5.4 each originate in one tradition yet acquire meaning only within the frame the others supply.

Second, the field confronts a measurement trilemma: micro-internal validity (randomized experiments), macro-ecological validity (media mix modeling), and construct validity (attention and tracking research) each capture part of the effectiveness picture, and none captures all (Table 8). Experiments are costly, context-bound, and blind to what happens off-platform (Gordon et al., 2019; Lewis & Rao, 2015; Liang et al., 2025); models rest on stationarity assumptions the digital era violates; and attention metrics remain immaturely linked to outcomes (Bruns et al., 2025; Pieters & Wedel, 2004). Practice over-indexes on micro methods because they are fast and platform-supplied, mechanically shifting budgets toward short-term, measurable, bottom-funnel channels—an observational analogue of Goodhart's law at the level of the entire discipline, visible in the 70.6 percent of marketers reporting a tilt toward short-term impact (Moorman, 2026).

Third, the field exhibits a productivity paradox: communication spending at record levels (WPP Media, 2025b) coexists with documented supply-chain leakage (ANA, 2023; ISBA & PwC, 2020, 2023), truncated attention (Bellman et al., 2020), pervasive fraud (Juniper Research, 2023), signal loss (Aridor et al., 2026; Wernerfelt et al., 2025), and managerial inability to demonstrate long-term impact (Moorman, 2026). Effectiveness estimation has thus become as consequential as effectiveness creation. Measurement regimes are not neutral instruments; they allocate budgets, and budgets shape market structure—privacy rules concentrate ad-tech markets (Johnson et al., 2023), retail media relocates advertising inside the point of sale (WPP Media, 2025b), and identifier loss penalizes small advertisers (Wernerfelt et al., 2025). Marketing communication research should therefore treat measurement itself as a substantive object of study—reflexively, since the field's own metrics participate in the phenomena they purport to describe.

Fourth, ethics has become constitutive rather than peripheral. Privacy architecture (Tucker, 2014; Zhu et al., 2025), disclosure regimes (Boerman et al., 2017; Ershov et al., 2025), AI provenance (Hermann, 2022; Koning & Voorveld, 2025), and the ethics of funding MFA and misinformation inventory (ANA, 2023; Patel et al., 2026) jointly define the terms on which persuasion occurs. The AI-disclosure findings are the sharpest illustration: transparency mandated by the EU AI Act from August 2026 carries a measurable persuasion cost (Baek et al., 2026; European Commission, 2026), so the normative and the positive questions—what brands should disclose and what disclosure does—can no longer be studied apart.

Fifth, the analysis supports a continuity thesis: each prior media rupture in the field's history was resolved not by incremental repair but by construct reconstruction—promotion became communication, communication became integration, integration is becoming orchestration (Table 1). The current rupture—machine-generated, algorithmically mediated, co-produced, and prospectively agent-negotiated communication—demands an analogous reconstruction around attention, authorship, and trust. That machine-made messages can outperform human-made ones (Hartmann et al., 2025) while disclosure of their origin erodes trust (Gu et al., 2026) captures the tension precisely: the founding sender–receiver dyad no longer specifies who is speaking, and the gaps cataloged in Section 10 are the shape of that reconstruction still pending. The contextual gap in Table 9 has a constructive corollary: markets such as Saudi Arabia and the wider Gulf region, in which digital advertising, influencer, and retail media formats are expanding rapidly, are conspicuously absent from the evidence base reviewed here (Dwivedi et al., 2021), and they therefore offer an opportunity to build and test theory outside the WEIRD frame rather than merely to replicate it.

12. Recommendations

The recommendations that follow convert the gap analysis of Section 10 and the interpretation offered in Section 11 into an actionable agenda for the three constituencies of the paper addresses. They rest on a single premise: the causal, contingency, infrastructure, and generative turns identified in Section 5 have outrun the theory, measurement standards, and institutional arrangements that the field inherited from the integrated marketing communication era, so that progress now depends as much on how research is organized as on which questions are asked. Section 12.1 sets research priorities, using a prioritization matrix (Figure 6) to sort the agenda by theoretical urgency and near-term feasibility, specifying ten research programs (Table 10), and naming the institutional reforms on which the most ambitious of them depend. Section 12.2 draws out the implications for managers, who must now treat measurement strategy as inseparable from communication strategy, and for policymakers and regulators, who face the task of designing disclosure, privacy, and competition rules as a single coherent problem.

12.1. Priorities for Future Research

Research investment should follow the dual logic of theoretical urgency and near-term feasibility, mapped in Figure 6: "quick wins" (AI-disclosure and authorship-perception experiments; natural experiments around ATT and the AI Act), "targeted programs" (retail media incrementality, attention–outcome linkage, B2B digital journeys, cross-cultural disclosure protocols), and "grand challenges" (contingent orchestration theory, machine-era persuasion theory, cumulative digital generalizations), alongside long-horizon societal work on sustainability, misinformation, and the macro effects of advertising.

The matrix makes two points that the tabular agenda cannot. First, the studies with the highest theoretical stakes—orchestration theory, machine-era persuasion theory, and cumulative digital generalizations—are also those that depend most on data the field does not yet control, which is why the institutional reforms discussed below are preconditions rather than afterthoughts. Second, the quick wins are quick because regulation is creating natural experiments faster than scholars are exploiting them: the AI Act's labeling obligations, the FTC's fake-reviews rule, and the retention of third-party cookies in Chrome each date a treatment that the disclosure and privacy literatures are already equipped to evaluate. Table 10 translates the framework into ten research programs grounded in the findings of Section 5 to 10.

Beyond individual studies, institutional reforms are prerequisites: preregistered and registered-report formats to combat selective reporting; industry–academic data trusts operating through privacy-preserving clean rooms; standardized, publicly documented attention and cross-channel metrics developed jointly with industry measurement bodies; practitioner-facing synthesis vehicles in the manner of medical Cochrane collaborations to close the translational gap (Nyilasy & Reid, 2007); and doctoral training that combines causal inference, computational methods, and normative theory, since the field's next generation must work across the three traditions rather than within one.

12.2. Implications for Practice and Policy

For managers, integration has become harder, not easier, and measurement strategy is now inseparable from communication strategy. The evidence supports rebalancing toward long-term brand investment (Binet & Field, 2013; Edeling et al., 2021); architecting paid media as anchors for owned and earned systems (Colicev et al., 2018; Liang et al., 2025); selecting influencers on engagement rather than reach (Beichert et al., 2024; Wies et al., 2023); auditing programmatic supply chains against documented leakage benchmarks (ANA, 2023; ISBA & PwC, 2023); triangulating attribution with experiments and mix models (Gordon et al., 2019); adopting validated attention metrics as leading indicators (Bruns et al., 2025); and governing generative AI with disclosure strategies calibrated to the documented trust penalty (Koning & Voorveld, 2025; Sands et al., 2026) and to fraud externalities (Juniper Research, 2023).

For policymakers and regulators, the disclosure literature demonstrates that transparency interventions measurably shape persuasion processes (Boerman et al., 2017; Evans et al., 2017), supporting evidence-based rule-making for influencer and AI-generated content (European Commission, 2026; Federal Trade Commission, 2023, 2024)—but the 96 percent non-disclosure rate documented by Ershov et al. (2025) shows that rules without computational enforcement are largely symbolic. The programmatic transparency findings justify supply-chain disclosure obligations analogous to financial auditing, and the attention research agenda suggests that "exposure" itself may require redefinition in future media standards. Regulators should also weigh the competitive side effects documented by Johnson et al. (2023), Goldberg et al. (2024), and Aridor et al. (2026): privacy rules can concentrate ad-tech market power and burden small firms even as they protect consumer data. For regulators in rapidly digitalizing markets such as Saudi Arabia and other Gulf states, which are building data-protection and advertising-governance regimes concurrently, these findings argue for treating disclosure, privacy, and competition rules as a single design problem rather than as three separate ones.

13. Conclusions

Marketing communication's trajectory—from the orphaned techniques of the promotion mix to a self-conscious integrative discipline, and now into an algorithmically mediated, generative, co-creative condition—reflects the dialectic of economic necessity, technological disruption, and theoretical ambition that has shaped marketing thought. Across the 26 years from 2000 to 2026, the internet moved from a minority medium to the infrastructure of three-quarters of humanity, U.S. internet advertising grew from US$8 billion to nearly US$300 billion, commerce media overtook television, and the author of a persuasive message became, for the first time, potentially a machine. This analysis has argued that the field's condition cannot be fully understood without its relationship to advertising research and practice: advertising supplied the measurement institutions, the empirical generalizations, and the professional infrastructure upon which marketing communication stands, even as digitalization has demoted advertising from hegemon to anchor within the tool ecosystem and thrown its accountability arrangements into crisis. The 36 studies reviewed in Section 5 confirm that this crisis is empirically documented across the causal, contingency, infrastructure, and generative dimensions of contemporary research.

The discipline's founding problem has inverted. Where communicators once struggled to reach mass audiences with consistent messages, they now struggle to orchestrate hyper-abundant, personalized, machine-generated touchpoints into coherent and trustworthy brand meaning—under measurement regimes that systematically distort what is known about what works, and under disclosure regimes that make the origin of a message part of its meaning. Closing the gaps identified here—through integrative orchestration theory, a restored program of empirical generalization, triangulated measurement architectures, cross-cultural evidence, and normative seriousness about persuasion's societal consequences, supported by the institutional reforms recommended in Section 12—will determine whether marketing communication consolidates its status as a rigorous academic discipline for the algorithmic age, or remains, in part, a practice outrunning its own explanation.