Analyzing total rewards perceptions of marketing professionals through linkedin discourse: a natural language processing approach Authors: Sezai Tunca, Yavuz Selim Balcioglu, Ceren Cubukcu Cerasi, Ayse Bilgen, Ayse Ilgun-Kamanli Published: 2026 ABSTRACT AbstractThis study examines how marketing professionals publicly articulate total rewards perceptions through LinkedIn discourse, extending natural language processing methodologies via domain-specific model development. Analyzing 35,691 comments from 2023, we identify significant shifts toward digital skill-based compensation discourse (+ 31.2%) and strong positive sentiment associations with flexible work arrangements (r = .48, p < .001), particularly among technology sector professionals. Hierarchical regression analysis confirms that digital skill discourse orientation and industry sector interact to predict sentiment variation (ΔR² = 0.04, p < .001), with digital competency emphasis demonstrating differential effects across technology, retail, and consumer goods contexts. Drawing on Skill-Based Total Rewards Theory, we provide preliminary empirical support for three propositions: Digital Skill-Reward Congruence, Market Value of Skills, and Skills-Based Flexibility principles. The sentiment analysis model achieved an F1-score of 0.89, while industry-specific analysis revealed significant variations (F(2,35688) = 27.32, p < .001) across sectors. Recognizing that LinkedIn discourse reflects publicly constructed professional identity positions rather than unfiltered latent attitudes, we interpret these patterns as expressions of signaled reward preferences within professional networking contexts. Methodologically, the study demonstrates NLP applications for large-scale professional sentiment analysis. Practically, findings inform industry-specific compensation strategies recognizing digital competencies and workplace flexibility as critical retention factors. This research contributes to compensation theory through quantitative operationalization of SBTRT propositions while establishing methodological frameworks for discourse-based organizational research.Keywords: Marketing Compensation, Digital Skills Recognition, Total Rewards Analytics, Social Media Discourse Analysis, Professional Benefits EvolutionSubject terms: Psychology, Environmental social sciences Introduction LinkedIn discourse reveals critical gaps between stated compensation preferences and authentic professional sentiment about total rewards packages. Marketing professionals discuss rewards differently in organic peer conversations than in formal survey contexts, creating opportunities for understanding genuine preference evolution that traditional methodologies miss. This distinction matters for three reasons. First, authentic discourse captures emerging compensation trends before they appear in formal organizational policies or structured survey instruments. Second, sector-specific discourse patterns enable targeted compensation strategy development aligned with industry characteristics. Third, temporal discourse analysis tracks real-time preference evolution during periods of rapid workplace transformation. Existing research relies predominantly on traditional survey methodologies with significant limitations. Terera and Ngirande (2014) surveyed 200 marketing professionals about rewards and retention, while McIlveen and colleagues (2021) interviewed 450 marketing professionals about workplace satisfaction. WorldatWork (2022) conducted the largest survey-based examination with 2,847 marketing professionals across multiple industries. Despite methodological rigor, these approaches share common constraints. Structured questionnaires limit respondents to predetermined categories, potentially missing emerging trends. Social desirability bias influences formal survey responses, particularly regarding sensitive compensation topics1,2. Martinez and colleagues3 demonstrated that professionals express 34% more negative sentiment about workplace policies through social media compared to formal surveys, suggesting traditional methods underestimate authentic dissatisfaction. This research addresses three critical gaps through LinkedIn discourse analysis of 35,691 marketing professional comments collected throughout 2023, providing empirical validation of Skill-Based Total Rewards Theory while establishing methodological frameworks for understanding authentic professional sentiment about evolving compensation structures. Definitional framework: total rewards in organizational context This study examines total rewards as comprehensive compensation packages provided by employing organizations to their marketing professional workforce, encompassing both monetary compensation elements including salary, bonuses, and performance incentives and non-monetary benefits such as flexible work arrangements, professional development opportunities, digital tool access, and career advancement pathways. The research specifically focuses on rewards offered by current employers rather than external recognition systems or inter-organizational reward mechanisms. Accordingly, the scope of analysis includes employer-provided benefits, workplace flexibility arrangements, skills development programs, recognition systems, and career progression opportunities, while excluding external rewards such as industry recognition awards, professional association honors, or compensation offered by potential alternative employers during recruitment processes. The scope of rewards analysis includes employer-provided benefits, workplace flexibility arrangements, skills development programs, recognition systems, and career progression opportunities that organizations offer to attract, retain, and motivate marketing professionals. The study excludes external rewards such as industry recognition awards, professional association honors, or compensation offered by potential alternative employers during recruitment processes. The COVID-19 pandemic accelerated many of these digital transformation trends, fundamentally reshaping workplace expectations and normalizing remote work arrangements across industries4. The pandemic served as a catalyst that compressed years of gradual workplace evolution into months of rapid adaptation, creating a unique context where digital transformation and pandemic-induced changes became intertwined forces shaping employee reward preferences5. Our study recognizes that the 2023 data collection period represents a post-pandemic workplace environment where remote work expectations had already been normalized, creating a complex landscape where digitalization-driven changes intersect with pandemic-induced structural shifts in employment relationships. Theoretical positioning and contribution framework This research contributes to theoretical development by providing systematic empirical testing of the emerging Skill-Based Total Rewards Theory through innovative methodological approaches. Rather than assuming SBTRT represents a fully mature theoretical construct, the study positions theoretical testing and refinement as primary research objectives that advance theoretical understanding through systematic investigation. The theoretical contribution extends established compensation theory by integrating digital transformation considerations into traditional skill-based pay frameworks. The research provides empirical evidence for theoretical propositions while identifying areas requiring continued theoretical development and refinement through future research applications. The study’s theoretical positioning acknowledges SBTRT’s emerging status while demonstrating its legitimate foundation in established theoretical constructs and its potential for addressing contemporary workplace challenges not adequately served by existing theoretical frameworks. Rationale for marketing professional focus The study’s specific focus on marketing professionals represents a strategic methodological choice based on three empirical and theoretical considerations that distinguish this occupational group from other professional categories in their relationship with total rewards structures. Marketing professionals operate at the intersection of traditional business functions and emerging digital technologies, creating unique compensation challenges that reflect broader workplace transformation trends. The marketing profession has experienced particularly rapid digital transformation, with traditional advertising and brand management roles evolving into data-driven, technology-enabled positions requiring new skill sets and competencies. This transformation creates distinctive total rewards considerations that may not be present in other professional contexts with more stable technological requirements. The marketing profession demonstrates exceptional heterogeneity in skill requirements across industry sectors, with technology sector marketing professionals requiring different competencies than those in retail or consumer goods contexts. This sectoral variation enables systematic examination of how industry characteristics influence total rewards preferences, providing insights that extend beyond occupation-specific findings to broader questions about industry-compensation relationships. Marketing professionals maintain particularly active LinkedIn presence and engagement patterns, creating rich data sources for social media discourse analysis. The professional networking requirements inherent in marketing roles generate authentic professional discourse about career development, compensation expectations, and workplace satisfaction that may be less accessible in other occupational contexts with different professional communication patterns. The academic literature examining marketing professionals’ perceptions of total rewards through digital platforms remains underdeveloped, with existing research predominantly relying on traditional survey methodologies6. A systematic review of compensation research in marketing reveals that 78% of studies published between 2020 and 2023 employed cross-sectional surveys or structured interviews, limiting the capture of authentic, unfiltered professional discourse7. Traditional survey approaches, while methodologically rigorous, present inherent limitations when examining sensitive topics such as compensation satisfaction. The social desirability bias phenomenon, extensively documented in organizational research, suggests that respondents may provide socially acceptable responses rather than authentic perspectives when directly questioned about workplace rewards1,8. Furthermore, structured survey instruments may constrain respondents’ ability to express nuanced viewpoints about emerging compensation trends, particularly those related to digital transformation and flexible work arrangements. LinkedIn discourse analysis offers methodological advantages by capturing organic, peer-to-peer professional conversations that occur without researcher intervention. This approach provides access to authentic expressions of professional sentiment that may differ significantly from responses provided in formal survey contexts9. The platform’s professional networking environment encourages candid discussions about career advancement, compensation expectations, and workplace satisfaction, creating a rich data source for understanding contemporary professional perspectives on total rewards. Recent methodological studies have demonstrated significant differences between social media-expressed sentiments and survey-reported attitudes in professional contexts. Research by Martinez and colleagues3 found that professionals expressed 34% more negative sentiment about workplace policies in social media contexts compared to formal survey responses, suggesting that traditional methods may underestimate employee dissatisfaction levels. Survey methodology represents a cornerstone of organizational research, with well-established protocols for validity, reliability, and generalizability that have advanced our understanding of workplace phenomena for decades. However, specific empirical evidence suggests that social desirability bias presents particular challenges when examining sensitive topics such as compensation satisfaction and workplace reward preferences. Social desirability bias, defined as the tendency of respondents to provide answers they perceive as socially acceptable rather than authentic, has been documented extensively in compensation research contexts. Krumpal’s (2013) comprehensive meta-analysis of 72 studies examining sensitive survey topics found that compensation-related questions exhibited significantly higher social desirability response rates compared to general workplace satisfaction measures, with effect sizes ranging from 0.34 to 0.67 across different organizational contexts. Specifically relevant to compensation research, Podsakoff and colleagues2 demonstrated that employees systematically underreport compensation dissatisfaction in formal survey contexts when questions are administered by organizational representatives or when respondent anonymity cannot be guaranteed. Their experimental study with 1,247 employees across multiple organizations revealed that compensation satisfaction ratings were 23% higher in employer-administered surveys compared to independent third-party administered surveys, suggesting significant response bias effects. More directly applicable to marketing professional contexts, Van de Mortel’s (2008) research on response bias in workplace surveys found that professionals in client-facing roles, including marketing positions, exhibited heightened social desirability tendencies when discussing workplace satisfaction topics. This pattern reflected concerns about professional reputation and career advancement implications of expressing dissatisfaction with compensation packages. Contemporary research by Martinez and colleagues3 provides compelling evidence for the specific context of compensation research. Their comparative study examined 892 marketing professionals’ responses to identical compensation satisfaction questions administered through both formal surveys and analyzed through social media discourse analysis. Results revealed that professionals expressed 34% more negative sentiment about compensation packages in social media contexts compared to formal survey responses, with particularly pronounced differences in discussions about work-life balance and flexible work arrangements. The theoretical foundation for these bias effects stems from impression management theory, which suggests that individuals modify their responses in formal evaluation contexts to maintain positive professional images. In compensation research, this manifests as reluctance to express dissatisfaction with rewards packages, particularly when such responses might be perceived as ungrateful or professionally damaging. These empirical findings do not diminish the value of survey methodology but rather highlight the importance of methodological triangulation in compensation research. Social media discourse analysis provides complementary insights that can validate and extend survey-based findings by capturing authentic professional expressions that may differ from formal response contexts. The literature review reveals three critical research gaps that necessitate empirical investigation through innovative methodological approaches. First, the predominant reliance on survey-based methodologies has limited our understanding of how marketing professionals authentically express and discuss their total rewards preferences in organic professional contexts. The documented social desirability bias effects in compensation research suggest that traditional approaches may not capture the full spectrum of professional sentiment about evolving reward structures, particularly those related to digital transformation trends. Second, existing research has insufficiently examined how industry-specific contexts influence marketing professionals’ reward preferences and sentiment patterns. While studies by Terera and Ngirande (2014) and McIlveen et al. (2021) provide valuable insights into general marketing professional satisfaction, they lack the granular industry-specific analysis necessary to understand how sectoral characteristics shape reward valuations. The technology, retail, and consumer goods sectors present distinct operational environments that likely influence how marketing professionals perceive and prioritize different components of total rewards packages. Third, the temporal evolution of marketing professionals’ reward preferences remains underexplored in the context of accelerated digital transformation. The COVID-19 pandemic and subsequent digital acceleration created unprecedented conditions where traditional reward structures evolved rapidly, yet existing research has not systematically examined how these changes manifested in professional discourse over time. The Skill-Based Total Rewards Theory provides a theoretical framework for understanding these changes, but empirical validation through real-time professional discourse analysis remains limited. These research gaps collectively indicate the need for methodological innovation that can capture authentic professional sentiment across industry sectors and temporal periods. LinkedIn discourse analysis offers the methodological precision necessary to address these gaps by providing access to organic professional conversations that reveal unfiltered perspectives on total rewards evolution. Furthermore, the integration of natural language processing techniques with social media analytics presents opportunities to scale qualitative insights about professional discourse while maintaining analytical rigor. This methodological approach can generate sector-specific insights that inform both theoretical understanding and practical compensation strategy development. Given these theoretical and methodological considerations, our research addresses the following questions that emerge directly from the identified literature gaps: Research Questions. (RQ1) How do marketing professionals express and discuss their perceptions of total rewards packages through LinkedIn discourse, particularly regarding digital transformation and skill development? RQ2 How do sentiments and perceptions of total rewards vary across different industry sectors (Technology, Retail, and Consumer Goods) among marketing professionals? RQ2a What are the patterns of sentiment across industry sectors? RQ2b How do industry characteristics influence the valuation of different benefit types? RQ3 How have marketing professionals’ preferences for different components of total rewards evolved, particularly concerning digital skills compensation and flexible work arrangements? RQ3a What is the temporal evolution of discourse themes related to digital marketing skills compensation? RQ3b How has the relative importance of traditional versus digital marketing benefits changed over time? Answering these questions will make various contributions to both academic research and the business world. In terms of application, it shows how NLP methods can be applied to professional discourse analysis. In terms of academics, it helps to understand marketing professionals’ perceptions of rewards in more depth. In terms of business, these findings will help employers develop their total reward strategies and help marketers better understand their expectations, thus increasing employee loyalty. Because understanding how employee expectations are shaped in the digitalizing business world has become a critical factor for organizations to maintain their competitive advantage. This research contributes to both theoretical understanding and practical application of marketing professional compensation management in the digital transformation era. Theoretically, the study validates and extends the Skill-Based Total Rewards Theory through empirical analysis of authentic professional discourse, providing evidence for the increasing importance of digital competency recognition in compensation structures. The integration of natural language processing techniques with professional discourse analysis establishes a methodological framework for understanding employee sentiment and preference evolution that complements traditional survey-based approaches, demonstrating how NLP methods can be applied to professional discourse at scale. Practically, the research provides evidence-based guidance for organizations seeking to optimize their marketing talent management strategies through industry-specific compensation approaches, digital skills recognition frameworks, and flexible work arrangement implementation. By illuminating how employee expectations are shaped in the digitalizing business world, the findings enable human resource professionals to develop total rewards packages that align with contemporary marketing professional expectations while supporting organizational digital transformation objectives. The methodological approach demonstrates how social media analytics can inform real-time compensation strategy adjustments based on evolving professional discourse patterns, providing organizations with competitive advantages in talent acquisition and retention. Literature review Total rewards are an important research area in terms of employee motivation, commitment and job satisfaction10. This concept includes both traditional compensation approaches and contemporary flexible benefit models11. Especially for individuals working in highly competitive sectors such as marketing professionals, total rewards packages are not just about salary and financial incentives12, but also include factors such as career development, flexible working opportunities and work-life balance (Getguru,13. Recent research shows that the expectations of marketing professionals towards rewards have changed significantly with the digital transformation and the widespread use of remote working models14. In this section, basic concepts of total rewards literature, current research, social media analytics and professional discourse analyses will be discussed. Total rewards and employee satisfaction Contemporary research demonstrates that organizational total rewards significantly influence employee satisfaction, retention, and organizational commitment across diverse industry contexts and generational cohorts. Total rewards represent the comprehensive value proposition that employing organizations provide to marketing professionals, encompassing both tangible compensation elements and intangible workplace benefits that contribute to overall employment satisfaction. The organizational total rewards framework distinguishes between employer-provided compensation and external recognition systems, focusing specifically on the employment relationship between marketing professionals and their current organizations. This approach enables systematic examination of how employing organizations structure their total rewards packages to attract, retain, and motivate marketing talent within specific industry contexts. Recent evidence from Malaysia’s food and beverage industry reveals that job satisfaction, organizational commitment, work-life balance, and competitive pay structures collectively reduce employee turnover intentions, with organizations that cultivate positive work environments and provide comprehensive development opportunities achieving superior talent retention outcomes (Yook Pei et al., 2025). The relationship between total rewards and employee satisfaction exhibits distinct generational patterns, with Generation Z employees demonstrating complex preference structures that prioritize both extrinsic rewards such as compensation and recognition alongside intrinsic motivators including personal growth opportunities and meaningful work purpose (Mabaso, 2025). Similarly, Millennials in post-COVID workplace environments emphasize five critical reward categories encompassing remuneration, well-being, development, recognition, and benefits, with non-monetary rewards such as flexible and remote work arrangements becoming essential components for achieving work-life balance satisfaction (Mabaso & Mathebula, 2025). The effectiveness of total rewards systems depends significantly on mediating factors, with research on Generation Z employees revealing that perceived supervisor support and comprehensive total rewards packages mediate the relationship between flexible work arrangements and turnover intentions, suggesting that reward system success requires integrated management approaches rather than isolated benefit provision (Protsiuk, 2025). Cross-cultural validation from Central European contexts confirms that employee satisfaction and motivation remain fundamentally connected to wages and benefits, professional development opportunities, working conditions, and leadership quality, indicating that total rewards frameworks must address both financial and environmental factors to achieve sustained employee commitment (Karácsony et al., 2025). The impact of digital transformation on marketing professionals’ perception of awards The push of technology has led to radical changes in the business world. It has affected many areas, especially the marketing sector15. In this context, the use of artificial intelligence models causes businesses to move away from traditional marketing approaches and switch to new methods such as data analytics and artificial intelligence-based customer relationship management. The competence expectations of marketing professionals, who are important actors in this field, and their perceptions of rewards are changing in parallel16. In a study conducted by Forbes Human Resources Council (2023), it was determined that marketing professionals with digital skills receive 30% higher salaries and have more career development opportunities compared to traditional marketing professionals17. The main reason for this is the need for employers to expand their total rewards package to retain employees who can keep up with digital transformation (Korn Ferry,18. In particular, it reveals that flexible working arrangements and access to digital tools are a determining factors in marketing professionals’ job satisfaction19. Sectoral differences and marketing professionals’ views on awards Marketing professionals’ perceptions of total rewards vary across sectors20. Various studies on marketing professionals working in the technology, retail and consumer goods sectors have revealed that the total rewards offered by the sectors have different priorities21. Marketing professionals working in technology firms are more likely to prioritize rewards such as access to digital tools, flexible working, skill-based bonuses, and stock options22. In the retail sector, performance-based bonuses and work-life balance are more important for marketing professionals23. When we analyze the consumer goods sector, marketing professionals working in this sector see rewards such as brand prestige, career development, and employee discounts as more valuable24. Research shows that these differences between sectors are determinants in shaping marketing professionals’ feelings and perceptions toward total rewards25. This research analyses the current studies by analyzing the posts of professionals who discourse on social media platforms. Social media analytics and professional discourse analysis: existing methodological landscape The application of natural language processing techniques to professional social media platforms has emerged as an established research methodology with substantial prior work examining employee sentiment, engagement patterns, and workplace attitudes. Recent studies have demonstrated the viability of LinkedIn data analysis for understanding professional discourse patterns, establishing a methodological foundation upon which this research builds. Schweitzer and colleagues (2019) pioneered the application of sentiment analysis to LinkedIn professional content, examining how professionals express career satisfaction and workplace engagement through platform discourse. Their work established fundamental methodological approaches for extracting meaningful insights from professional social media content while addressing privacy and ethical considerations inherent in social media research applications. Building upon this foundation, Chen and Zhang (2021) developed sophisticated natural language processing approaches specifically calibrated for professional discourse analysis, demonstrating how domain-specific lexicon development enhances analytical accuracy when examining workplace-related social media content. Their methodology established protocols for distinguishing between authentic professional expression and performative professional communication, providing crucial methodological guidance for subsequent research applications. Recent methodological studies have corroborated significant differences between social media-expressed sentiments and survey-reported attitudes in professional contexts, with comparative analyses revealing that professionals consistently express more candid and frequently more negative assessments of workplace policies through social media discourse than through formal survey instruments3; see Sect. 1 for detailed findings). Their comprehensive analysis of 50,000 professional posts demonstrated the scalability of natural language processing approaches while identifying industry-specific patterns in professional discourse about workplace satisfaction and compensation expectations. The application of named entity recognition to professional social media analysis has been demonstrated by Rodriguez and Thompson (2022), who developed custom entity recognition models for identifying specific organizational benefits, compensation structures, and workplace policies mentioned in professional discourse. Their work established technical frameworks for systematic entity extraction that enable quantitative analysis of qualitative professional discourse patterns. Recent methodological advances by Park and colleagues (2024) have integrated multiple natural language processing techniques to examine professional discourse evolution over time, tracking how workplace preference expressions change in response to external economic and technological factors. Their longitudinal analysis approach provided methodological guidance for temporal pattern identification in professional social media discourse. Skill-based total rewards theory: theoretical boundaries, novel contributions, and derived hypotheses SBTRT integrates established compensation theory with digital transformation dynamics to explain contemporary professional reward preferences. The theory builds upon four foundational frameworks while contributing three propositions that extend existing knowledge to address workplace challenges inadequately served by traditional theories. Rather than claiming wholesale theoretical novelty, SBTRT addresses specific explanatory limitations in existing compensation frameworks that become salient under conditions of rapid technological change limitations that the foundational theories were not designed to accommodate and that contextual adaptation alone cannot resolve. Theoretical foundations Skill-Based Pay Theory (Lawler, 1990; Ledford, 1995) established that compensation systems rewarding specific competencies outperform position-based structures. Human Capital Theory (Becker, 1964; Schultz, 1961) demonstrated the economic value of knowledge and skills as productive assets. Social Exchange Theory (Blau, 1964) explained how skill recognition influences employee-organization relationships through reciprocal exchange dynamics. Self-Determination Theory (Deci & Ryan, 1985) showed that reward systems supporting psychological needs generate superior motivational outcomes compared to purely extrinsic approaches. Explanatory gaps in existing frameworks While these foundational theories provide robust explanations for compensation dynamics in stable technological environments, they encounter specific explanatory limitations when applied to digitally transformed professional contexts limitations that constitute genuine theoretical gaps rather than mere contextual variations. The first gap concerns skill valuation temporality. Traditional Human Capital Theory assumes that competency market values adjust gradually, with educational credentials and accumulated experience serving as reliable proxies for productive capacity (Becker, 1964). This assumption becomes problematic in digital professional contexts where specific technological competencies can appreciate or depreciate rapidly a phenomenon illustrated by the swift market valuation shifts surrounding skills in marketing automation, AI-driven analytics, and programmatic advertising. Existing theory lacks mechanisms for explaining how compensation systems should respond when the temporal dynamics of skill valuation fundamentally change, creating a gap between theoretical predictions of gradual adjustment and observed market behavior requiring continuous recalibration. The second gap involves the skill-flexibility nexus. Self-Determination Theory explains why autonomy-supportive work environments enhance motivation (Deci & Ryan, 1985), and Skill-Based Pay Theory explains why competency recognition improves compensation satisfaction (Lawler, 1990). However, neither framework nor their combination explains why specific categories of competencies generate distinctive autonomy expectations that existing theories treat as unrelated to skill type. Digital competencies, by their nature, enable location-independent work execution and asynchronous collaboration in ways that traditional marketing competencies (e.g., in-person client management, trade show coordination) do not. This creates a theoretically underspecified relationship between competency type and flexibility expectations that existing frameworks address inadequately. The third gap relates to cross-sectoral skill valuation heterogeneity. Traditional compensation theory treats industry context as a moderating variable affecting the magnitude of established relationships (Milkovich et al., 2014). However, the relationship between digital competencies and reward structures may differ qualitatively not merely quantitatively across sectors with different digital intensity levels. In technology-intensive contexts, digital competencies function as foundational professional requirements, whereas in sectors such as consumer goods, they may operate as differentiating capabilities. This qualitative distinction in how competencies relate to compensation structures represents an explanatory gap that moderation-based approaches within existing frameworks cannot fully capture. SBTRT addresses these three gaps through propositions that specify mechanisms operating under conditions of digital transformation mechanisms that are not derivable from the foundational theories individually or in combination, yet remain conceptually continuous with their core constructs. Distinctive theoretical boundaries SBTRT differs from antecedent theories in three critical dimensions that correspond to the explanatory gaps identified above. First, traditional Skill-Based Pay Theory treats technological competencies as specialized expertise within broader skill portfolios, while SBTRT positions digital competencies as foundational professional requirements whose valuation dynamics differ qualitatively from traditional skill categories. This distinction reflects digital transformation fundamentally altering skill valuation hierarchies rather than adding new specialized categories to existing frameworks. Second, traditional Human Capital Theory assumes relatively stable skill market valuations changing gradually over extended periods, while SBTRT recognizes that digital competencies possess distinctive economic characteristics including rapid obsolescence potential and continuously evolving market valuation patterns. This temporal distinction requires adaptive compensation systems capable of rapid adjustment rather than static salary structures based on historical job evaluations. Third, traditional compensation frameworks treat workplace flexibility as universal employee benefits independent of specific competencies, while SBTRT positions flexibility as skill-contingent rewards that digital competency mastery functionally enables and legitimately requires. This conceptual shift recognizes that digital skills create professional autonomy expectations that organizations must address through comprehensive reward structures rather than discretionary benefits. Novel theoretical propositions and derived hypotheses While the research questions guide exploratory investigation of marketing professional discourse patterns, the three SBTRT propositions generate specific testable predictions about relationships between digital competencies, reward components, and professional satisfaction. Below, each proposition is presented alongside the formal hypotheses it logically yields. We emphasize that the current study provides a first empirical examination of these predictions through discourse analysis; the results should be understood as preliminary evidence informing continued theoretical development rather than definitive confirmation. The Digital Skill-Reward Congruence Principle states that organizations in digitally intensive environments must provide comprehensive reward packages recognizing digital competencies as primary value determinants rather than supplementary capabilities. This extends Skill-Based Pay Theory by arguing that digital transformation fundamentally alters which competencies command premium compensation, elevating technological proficiency from specialized expertise to foundational professional requirements. If digital competency recognition constitutes a primary driver of reward satisfaction as this principle proposes, then professionals who articulate digital competencies should evaluate their total rewards more favorably than those emphasizing traditional capabilities. Furthermore, this congruence effect should be most pronounced in digitally intensive industry contexts where the alignment between digital skills and organizational reward structures is strongest. These expectations yield the following hypotheses: H1 Marketing professionals expressing digital competency discourse will demonstrate significantly higher positive sentiment toward total rewards packages compared to professionals emphasizing traditional marketing capabilities. H2 The relationship between digital competency discourse and positive sentiment will vary systematically across industry sectors, with technology sector professionals demonstrating stronger associations than retail or consumer goods sectors. The Market Value of Skills Principle asserts that digital competency market premiums increase proportionally to technological change velocity within specific industry contexts, creating dynamic compensation adjustment requirements that traditional pay structures cannot accommodate. This extends Human Capital Theory by incorporating temporal dynamics and market responsiveness factors, recognizing that digital skill valuation operates on fundamentally different temporal scales compared to traditional competencies. If digital competencies command increasingly dynamic market premiums as this principle suggests, then professional discourse should reflect accelerating emphasis on digital skill compensation over time. Moreover, professionals possessing these high-market-value competencies should express stronger preferences for skill-based compensation mechanisms that reflect their evolving market worth rather than static traditional pay structures. These expectations generate the following hypotheses: H3 The frequency of digital skills compensation discourse will increase significantly throughout the 2023 study period, reflecting accelerating market valuation of technological competencies. H4 Marketing professionals discussing digital competencies will express stronger preferences for skill-based compensation mechanisms (premium salaries, performance bonuses, certification reimbursement) compared to traditional compensation structures. The Skills-Based Flexibility Principle posits that digital competencies create professional autonomy expectations necessitating flexible work arrangements as essential reward components rather than discretionary benefits. This integrates Self-Determination Theory with Skill-Based Pay Theory, arguing that digital skill mastery generates psychological ownership over work processes that employees expect organizations to recognize through autonomy-supporting workplace arrangements. If digital competency mastery generates autonomy expectations as this principle proposes, then flexible work arrangements should elicit stronger positive sentiment among digitally oriented professionals than among those emphasizing traditional capabilities. Additionally, the value placed on digital tool access a concrete expression of technology-mediated work autonomy should vary across sectors in proportion to industry digital intensity. These expectations yield the following hypotheses: H5 Flexible work arrangement discourse will demonstrate stronger positive sentiment correlation among marketing professionals expressing digital competency preferences compared to those emphasizing traditional capabilities. H6 Digital tool access discourse will correlate more strongly with positive sentiment in technology sector contexts compared to retail or consumer goods sectors, reflecting industry-specific digital intensity moderating effects. These six hypotheses, derived progressively from the three SBTRT propositions, generate empirically testable predictions regarding relational dynamics that extant theoretical frameworks do not explicitly anticipate. In doing so, they delineate SBTRT’s distinctive explanatory contribution while preserving conceptual continuity with established theoretical foundations, as illustrated in Fig. 1. Skill-Based Total Rewards Theory (SBTRT): Conceptual Framework. Methodology This study employs a comprehensive methodological framework to analyze marketing professionals’ perceptions of total rewards through social media discourse. Our data collection and analysis process focused on LinkedIn as the primary data source, encompassing 35,691 comments collected during the 2023 calendar. Research design and occupational focus The research design integrates three complementary computational approaches sentiment analysis, topic modeling, and named entity recognition to examine marketing professionals’ perceptions of total rewards through systematic analysis of LinkedIn discourse collected throughout the complete 2023 calendar year. Each analytical component captures a distinct dimension of professional discourse about compensation and benefits, enabling comprehensive examination of sentiment patterns, thematic evolution, and entity-level relationships within the dataset. Rationale for marketing professional focus The specific focus on marketing professionals represents a strategic methodological choice grounded in three empirical and theoretical considerations that distinguish this occupational group in their relationship with total rewards structures. First, marketing professionals operate at the intersection of traditional business functions and emerging digital technologies, creating unique compensation challenges that reflect broader workplace transformation trends. The marketing profession has experienced particularly rapid digital transformation, with traditional advertising and brand management roles evolving into data-driven, technology-enabled positions requiring fundamentally new skill sets and competencies26,27. This transformation creates distinctive total rewards considerations including digital tool access, skills certification support, and technology-enabled performance incentives that may not be present in professional contexts with more stable technological requirements. Second, the marketing profession demonstrates exceptional heterogeneity in skill requirements across industry sectors, with technology sector marketing professionals requiring different competencies than those in retail or consumer goods contexts20,21. This sectoral variation enables systematic examination of how industry characteristics influence total rewards preferences, providing insights that extend beyond occupation-specific findings to broader questions about industry-compensation relationships. The structured variation across sectors creates natural comparative conditions supporting theoretical proposition testing. Third, marketing professionals maintain particularly active LinkedIn presence and engagement patterns, creating rich data sources for social media discourse analysis28. Workplace Learning Report identified marketing as among the most active professional categories on the platform in terms of content creation and professional networking engagement. The professional networking requirements inherent in marketing roles generate authentic discourse about career development, compensation expectations, and workplace satisfaction that may be less accessible in other occupational contexts with different professional communication patterns. Definitional framework As established in the Introduction, this study examines total rewards as comprehensive compensation packages provided by employing organizations to their marketing professional workforce, encompassing both monetary and non-monetary elements within the current employment relationship (see Sect. 1, Definitional Framework, for complete specification). Contextual framework: digital transformation and post-pandemic workplace dynamics The data collection period throughout 2023 represents a unique historical moment where digital transformation trends intersected with pandemic-induced workplace changes, creating a dual-transformation context that fundamentally shapes professional discourse about total rewards. The COVID-19 pandemic accelerated many digital transformation trends, fundamentally reshaping workplace expectations and normalizing remote work arrangements across industries4. The pandemic served as a catalyst that compressed years of gradual workplace evolution into months of rapid adaptation, creating conditions where digital transformation and pandemic-induced changes became intertwined forces shaping employee reward preferences5. The 2023 data collection period represents a post-pandemic workplace environment where remote work expectations had already been normalized, creating a complex landscape where digitalization-driven changes intersect with pandemic-induced structural shifts in employment relationships. Rather than attempting to isolate these effects, this research examines marketing professionals’ total rewards preferences within this dual-transformation context, recognizing that contemporary professional discourse reflects the combined influence of technological change and pandemic-accelerated workplace evolution. Data collection protocol and sample characteristics The data collection process employed a systematic approach implemented through LinkedIn’s official API using Python’s linkedin-api library version 2.0.1, ensuring full adherence to platform terms of service and rate limiting requirements. The collection spanned the complete 2023 calendar year from January 1, 2023, to December 31, 2023, enabling comprehensive capture of marketing professionals’ total rewards discourse across multiple temporal contexts including seasonal variations in professional activity, quarterly business cycles, and annual compensation review periods. Search strategy and keyword selection The initial data identification phase utilized a targeted search strategy focusing on posts containing compensation and benefits-related terminology. The complete search algorithm employed two keyword tiers representing the exhaustive search strategy applied throughout data collection. Primary keywords included “total rewards,” “compensation,” “benefits,” “salary,” “work-life balance,” and “professional development,” representing core compensation discourse terminology identified through preliminary qualitative analysis of marketing professional LinkedIn content. Secondary keywords included “remote work,” “flexible hours,” “digital skills,” “career advancement,” and “employee satisfaction,” representing emerging themes in contemporary professional discourse about workplace rewards. The keyword selection process was grounded in established total rewards literature and validated through preliminary content analysis of 500 marketing professional posts collected during pilot testing in December 2022. Related terms such as “wage,” “income,” and “earnings” were explicitly considered during keyword development but excluded from the final search strategy because preliminary analysis revealed these terms appeared predominantly in economic commentary and labor market discussions rather than authentic professional discourse about individual compensation satisfaction. The term “salary” was retained as it appeared consistently in personal career development discussions, while “wage” appeared primarily in policy discussions and economic analysis contexts not relevant to professional reward perceptions. User identification and industry classification The user identification protocol specifically targeted marketing professionals through profile-based filtering criteria implementing multiple verification stages. The system identified eligible participants based on job title indicators including “Marketing Manager,” “Digital Marketing Specialist,” “Marketing Director,” “Brand Manager,” “Marketing Coordinator,” and related variations representing primary marketing professional designations. Additional verification occurred through industry sector classification, with participants categorized into technology, retail, and consumer goods sectors based on their current employer information and professional background indicators. The industry classification process identified these three sectors as the dominant employer categories for marketing professionals within the dataset, collectively representing 87% of identified marketing professional profiles. Healthcare sector marketing professionals represented 6% of initially identified profiles, education sector represented 4%, and nonprofit sector represented 3%. The decision to focus analytical attention on technology, retail, and consumer goods sectors reflects both their numerical dominance in the dataset and their theoretical relevance for examining digital transformation impacts on total rewards preferences. Healthcare, education, and nonprofit sector professionals were excluded from the final analytical sample to maintain sectoral coherence and enable meaningful comparative analysis across industries with comparable organizational structures and market dynamics. This exclusion represents a limitation acknowledged in the discussion and conclusion sections, with future research recommended to examine total rewards discourse patterns in these additional sectors. Geographic scope and linguistic parameters The geographic filtering protocol focused on English-language content from professionals based in North America, Europe, and Australia to ensure linguistic consistency and cultural relevance. This geographic scope encompassed the primary markets where the technology, retail, and consumer goods sectors demonstrate significant LinkedIn professional activity, ensuring representative coverage of marketing professional discourse patterns within these regions. The focus on English-language content represents a pragmatic methodological decision enabling sophisticated natural language processing analysis while acknowledging geographic and cultural limitations discussed further in Sect. 6.1. Temporal data collection strategy The temporal data collection strategy implemented systematic daily extraction procedures throughout the 2023 calendar year, with API queries distributed across 24-hour periods using rate-limiting protocols maintaining compliance with LinkedIn’s 1,000 requests per day limitation. This systematic daily extraction approach ensured consistent data capture across all twelve months, with automated collection procedures implemented at regular intervals to maintain temporal representativeness. The extended collection period enables robust temporal analysis of discourse evolution, particularly regarding the development of digital transformation themes and flexible work arrangement discussions throughout the year. Ethical compliance and privacy protection The data collection process implemented comprehensive privacy protection measures addressing ethical considerations in professional discourse analysis. All collected data underwent anonymization procedures removing personally identifiable information including names, specific employer identifiers, and geographic location details beyond regional classification. The research methodology specifically focused on aggregate pattern identification rather than individual identification, ensuring that findings contribute to theoretical understanding rather than enabling individual surveillance capabilities. Data preprocessing and quality assurance The data preprocessing phase implemented a systematic five-stage cleaning protocol designed to ensure dataset integrity and analytical reliability, with each stage addressing distinct data quality challenges commonly encountered in social media discourse analysis, as summarized in Table 1. The language detection stage utilized the langdetect library’s probabilistic language identification algorithm, processing each comment through Google’s language detection model with probability scores across 55 supported languages. Comments receiving English language confidence scores below 0.85 were excluded, representing 8% of the initial dataset distributed across 12 other languages. The deduplication process employed exact string matching identifying 1.2% direct duplicates, followed by fuzzy matching using Levenshtein distance calculations at 0.90 similarity threshold identifying an additional 1.6% near-duplicate content. Table 1Data Screening and Refinement Process.Processing StageInitial SampleExcludedRemainingExclusion CriteriaRaw data collection42,847-42,847-Language detection42,8473,42839,419English confidence < 0.85Exact deduplication39,41947338,946Identical character sequencesFuzzy deduplication38,94663138,315Levenshtein similarity > 0.90Content validation38,3151,89136,424Relevance score < 0.6Structural cleaning36,42439836,026Technical artifacts/spamAuthentication validation36,02633535,691Non-authentic content indicatorsFinal analytical sample35,691-35,691- Data Screening and Refinement Process. The content validation stage implemented automated relevance scoring using term frequency-inverse document frequency algorithms calibrated for total rewards discourse, evaluating each comment against a professionally curated lexicon of 347 compensation-related terms, as detailed in Table 1. Comments scoring below 0.6 relevance threshold underwent manual review by domain experts, resulting in 1,891 exclusions representing discussions mentioning compensation terminology in contexts unrelated to professional reward perceptions. The structural cleaning phase removed LinkedIn-specific formatting elements while preserving semantic content integrity, with 398 comments excluded due to excessive technical artifacts or encoding errors preventing reliable natural language processing. The authentication validation stage employed automated quality assessment algorithms analyzing linguistic complexity metrics, professional terminology density, and response pattern characteristics to distinguish authentic professional discourse from automated content generation. Comments demonstrating characteristics inconsistent with genuine professional communication patterns were flagged for manual review, resulting in 335 exclusions representing potential spam content or automated posting patterns, with the full screening and refinement outcomes presented in Table 1. Conceptual definitions and operational frameworks The analytical framework employs precise conceptual definitions with systematic operational frameworks ensuring methodological rigor in examining marketing professional discourse about total rewards components. These definitions establish clear boundaries for discourse categorization while enabling reliable quantitative analysis of qualitative professional expression patterns. Digital skills: conceptual definition and operational framework Digital skills are defined as technology-enabled competencies enabling marketing professionals to execute data-driven marketing strategies, utilize digital platforms for customer engagement, and leverage analytical tools for performance optimization. The operational framework identifies digital skills through systematic recognition of specific technological competencies mentioned in professional discourse, categorized into five distinct competency domains based on industry-standard marketing technology requirements, as detailed in Table 2. Table 2Digital Skills Operational Framework and Recognition Criteria.Competency DomainRepresentative KeywordsExample Discourse PatternsData Analytics SkillsGoogle Analytics, customer segmentation, conversion tracking, performance metrics, predictive modeling“My Google Analytics certification led to 15% salary increase”Social Media MarketingContent management systems, social media advertising, influencer platforms, community management“Companies value social media expertise over traditional advertising”Marketing AutomationCRM systems, email marketing platforms, lead nurturing, campaign automation“Marketing automation expertise opened new career opportunities”Search Engine OptimizationKeyword research, organic search optimization, technical SEO, search advertising“SEO expertise qualified me for senior positions”Customer Experience TechnologyPersonalization platforms, journey mapping tools, user experience analytics“Customer analytics capabilities command premium compensation” Digital Skills Operational Framework and Recognition Criteria. The operational identification process utilized named entity recognition algorithms trained on a comprehensive lexicon of 247 digital marketing technology terms and 156 data analytics competency indicators. Comments containing three or more digital skill terminology instances were classified as digital skills discourse, achieving 0.88 precision and 0.86 recall for digital skills terminology identification through validation analysis. Analytical framework Our analysis incorporated three primary computational approaches designed to comprehensively analyze marketing professionals’ discourse about total rewards. The first approach centered on sentiment analysis, implementing a BERT-based model29 specifically fine-tuned for our research context. Using a carefully curated subset of 1,000 marketing professionals’ comments for manual labeling, we achieved exceptional model performance with a validation accuracy of 0.91 and an F1-score of 0.89 on our test set. To enhance the accuracy of sentiment classification, we developed a custom lexicon incorporating marketing-specific compensation terminology, ensuring precise capture of industry-specific nuances in the discourse. The second component of our framework focused on topic modeling, employing both Latent Dirichlet Allocation (LDA) (Putri & Kusumaningrum,30 and BERTopic31 methodologies for comprehensive thematic analysis. Our LDA implementation was optimized through iterative testing, ultimately utilizing 15 topics based on coherence score optimization. The BERTopic approach provided valuable insights into the dynamic evolution of topics over time, while expert review by marketing compensation specialists ensured the validity and relevance of identified themes. For our third analytical component, we developed a custom Named Entity Recognition (NER) (Shaip,32 model using spaCy’s modeling framework. This model was trained on 2,000 manually annotated marketing-specific comments, carefully selected to represent the diversity of our dataset. The model was designed to identify and classify four key entity categories: benefit types, company names, job roles, and industry sectors. Our NER implementation achieved a robust F1-score of 0.85 on the validation set, demonstrating strong performance in entity identification and classification. Validation framework Our validation strategy incorporated multiple layers of verification to ensure the reliability and robustness of our findings. We implemented comprehensive 5-fold cross-validation across all models, calculating precision, recall, and F1-scores for each fold to ensure consistent performance across different data subsets. This systematic approach to validation provided strong evidence for the generalizability of our results. The inter-rater reliability assessment involved three independent coders with substantial marketing expertise who manually classified a random sample of 1,000 comments. This process achieved a Cohen’s Kappa coefficient of 0.84, indicating strong agreement among raters and validating the reliability of our coding framework. The technical infrastructure of our study was built entirely in Python, leveraging several specialized libraries to support different aspects of our analysis. We utilized Transformers33 for the implementation of our BERT-based models, spaCy for our custom NER development, Gensim for topic modeling operations, and Scikit-learn for comprehensive validation metrics. This integrated technical approach ensured efficient processing of our large dataset while maintaining high standards of analytical rigor. Comprehensive validation framework: computational, qualitative, and theoretical integration The study implements a multi-layered validation framework integrating computational performance assessment, qualitative expert review, and theoretical coherence evaluation to ensure that analytical outputs reflect meaningful professional discourse patterns rather than superficial statistical associations or methodological artifacts. Computational model validation The validation strategy incorporates comprehensive 5-fold cross-validation across all computational models, calculating precision, recall, and F1-scores for each fold to ensure consistent performance across different data subsets. The sentiment analysis model achieved validation accuracy of 0.91 and F1-score of 0.89 on the test set, demonstrating robust performance across diverse comment types and industry contexts. The custom Named Entity Recognition model achieved F1-scores of 0.88 for digital marketing tools, 0.86 for marketing job roles, 0.83 for marketing benefits, and 0.81 for industry sectors, providing strong evidence for reliable entity identification and classification across critical analytical categories. The cross-validation analysis revealed consistent model performance across temporal periods, with no significant performance degradation when applying models trained on early-year data to late-year comments, suggesting that the analytical frameworks maintain stability across the complete study period. Qualitative expert validation and semantic coherence assessment The study implemented qualitative validation through systematic review by three independent marketing compensation experts possessing minimum ten years of industry experience across technology, retail, and consumer goods sectors. The expert validation process involved detailed examination of representative comment samples from each thematic category, with evaluators assessing whether computational theme assignments aligned with human interpretation of discourse meaning and professional relevance. The validation protocol required experts to categorize 500 randomly selected comments using the identical thematic framework generated through computational analysis, achieving inter-rater reliability coefficient of 0.82 with the automated classification system. The inter-rater reliability assessment among the three expert coders achieved Cohen’s Kappa coefficient of 0.84, indicating strong agreement among independent evaluators and validating the reliability of the coding framework. The validation procedure further incorporated semantic depth analysis, in which expert reviewers assessed whether computationally classified digital skills discourse contained substantive discussions about skill-compensation relationships or represented peripheral terminology mentions. Similarly, the protocol examined whether positive sentiment expressions genuinely co-occurred with specific benefit category discussions or represented general professional satisfaction patterns. The results of these validation analyses are reported in Sect. 4.5, and their theoretical implications are discussed in Sect. 5.6. Theoretical coherence assessment protocol The validation framework included a dedicated theoretical coherence assessment examining whether empirical outputs aligned with SBTRT predictions or revealed patterns requiring theoretical refinement. This assessment evaluated the consistency between computational findings and each of the three SBTRT principles, with expert reviewers examining whether identified correlations and thematic patterns corresponded to theoretically predicted relationships. The substantive findings from this theoretical coherence assessment are integrated into the Discussion (Sect. 5.6), where they inform evaluation of SBTRT’s empirical support and identification of areas requiring continued theoretical development. Multivariate analytical procedures To complement the bivariate analyses described above, the study implemented hierarchical regression modeling to examine the simultaneous and interactive effects of industry sector, benefit type, and digital skill discourse orientation on sentiment scores. This multivariate approach addresses the possibility that bivariate correlations between individual predictors and sentiment may be confounded by shared variance across predictors, and enables formal testing of theoretically predicted moderation effects. Hierarchical regression specification The dependent variable in all regression models was the continuous sentiment score assigned to each comment through the BERT-based sentiment classification model (range: -1.0 to + 1.0). The hierarchical regression proceeded in three steps. In Step 1, control variables were entered including comment length (log-transformed word count), temporal quarter (Q1–Q4, dummy-coded with Q1 as reference), and geographic region (North America, Europe, Australia, dummy-coded with North America as reference). In Step 2, the primary predictors were entered: industry sector (technology, retail, consumer goods, dummy-coded with consumer goods as reference), digital skill discourse orientation (binary: classified as digital skill discourse versus non-digital skill discourse based on the operational framework described in Sect. 3.4), and benefit type discussed (flexible work, digital tool allowance, skill development, creative freedom, dummy-coded with creative freedom as reference). In Step 3, interaction terms were entered to test theoretically predicted moderation effects: the interaction between digital skill discourse orientation and industry sector, testing whether the sentiment premium associated with digital competency discourse varies across industries as predicted by H2 and the Digital Skill-Reward Congruence Principle. Moderated regression for sector-specific effects Separate moderated regression analyses examined whether the relationship between specific benefit types and sentiment was conditioned by industry sector, directly testing H6’s prediction that digital tool access discourse correlates more strongly with positive sentiment in technology sector contexts. These analyses employed benefit type frequency as a continuous predictor, industry sector as a categorical moderator, and the benefit-by-sector interaction term as the focal test of moderation. Robustness procedures To ensure that regression findings were not driven by distributional assumptions or influential observations, the analyses employed heteroscedasticity-consistent standard errors (HC3 estimator) and conducted sensitivity analyses excluding the top and bottom 1% of sentiment score distributions. Variance inflation factor diagnostics were computed for all models to assess multicollinearity among predictors, with VIF values below 3.0 confirming acceptable collinearity levels across all model specifications. Results Our analysis of 35,691 LinkedIn comments from marketing professionals revealed significant patterns in total rewards perceptions, presented through sentiment analysis, thematic evolution, and entity relationships. The analysis of marketing professionals’ LinkedIn discourse reveals patterns in how employees discuss and evaluate the total rewards packages provided by their current employing organizations. The findings reflect employee perspectives about their existing employment relationships rather than comparative assessments across multiple organizations or external reward systems. The sentiment analysis, thematic evolution patterns, and correlation findings specifically address marketing professionals’ attitudes toward the comprehensive compensation packages offered by their current employers. This organizational focus enables systematic examination of how different industry sectors structure their total rewards offerings and how marketing professionals respond to these employer-provided benefit packages. Sentiment analysis by industry Our sentiment analysis revealed distinct patterns in how marketing professionals across different industry sectors discuss total rewards (see Table 3). The technology sector demonstrated the highest positive sentiment (M = 0.45, SD = 0.28), followed by retail (M = 0.38, SD = 0.31) and consumer goods (M = 0.33, SD = 0.29). A one-way ANOVA confirmed statistically significant differences between industries, F(2,35688) = 27.32, p < .001. Table 3Sentiment Analysis by Industry for Marketing Professionals.Industry SectorMean SentimentSDSample Size (n)F-valuep-valueTechnology0.450.2812,84527.32< 0.001Retail0.380.3111,923Consumer Goods0.330.2910,923Overall0.390.2935,691 Sentiment Analysis by Industry for Marketing Professionals. Thematic Evolution Our topic modeling analysis identified significant shifts in discussion themes throughout the study period (see Table 4). Most notably, discussions about digital marketing skills compensation showed the highest increase (+ 31.2%), while traditional marketing role discussions decreased (-15.3%). The chi-square analysis confirmed the statistical significance of these thematic shifts (χ²(4) = 182.45, p < .001). Table 4Thematic Distribution and Temporal Evolution in Marketing Roles.Theme CategoryQ1 2023 (%)Q4 2023 (%)% Changeχ²p-valueDigital Marketing Skills28.459.6+ 31.2182.45< 0.001Remote Work Benefits18.738.9+ 20.2Traditional Marketing35.620.3-15.3Creative Benefits12.422.8+ 10.4Professional Development14.928.4+ 13.5 Thematic Distribution and Temporal Evolution in Marketing Roles. Digital Marketing Skills Theme (31.2% increase) The most prominent theme emergence centered on digital marketing skills compensation, characterized by keywords including “SEO optimization,” “data analytics,” “artificial intelligence,” “marketing automation,” “social media management,” “content marketing,” “digital advertising,” “conversion tracking,” “customer analytics,” and “performance metrics.” This theme reflects the fundamental shift toward technical competency recognition in marketing compensation structures. The discourse analysis reveals that marketing professionals increasingly frame their value proposition through technological proficiency rather than traditional creative or strategic capabilities. Comments within this theme frequently reference specific tools such as “Google Analytics,” “HubSpot,” “Salesforce,” and “Adobe Creative Suite,” indicating that technical tool mastery has become a primary determinant of compensation expectations. Remote Work Benefits Theme (20.2% increase) The remote work benefits theme demonstrated substantial growth throughout the study period, characterized by keywords including “work-life balance,” “flexible scheduling,” “remote collaboration,” “digital nomad,” “home office setup,” “virtual meetings,” “asynchronous communication,” “location independence,” “family time,” and “commute elimination.” This thematic evolution reflects the permanent integration of remote work capabilities into marketing professional expectations. The discourse analysis reveals that marketing professionals view remote work not merely as a temporary pandemic accommodation but as a fundamental component of contemporary total rewards packages. The keyword frequency analysis indicates that discussions emphasize productivity benefits and personal well-being outcomes rather than cost-saving considerations for employers. Traditional Marketing Theme (15.3% decrease) The traditional marketing theme experienced notable decline, characterized by keywords including “brand management,” “print advertising,” “trade shows,” “networking events,” “client presentations,” “creative campaigns,” “traditional media,” “billboard advertising,” “radio marketing,” and “television commercials.” This declining theme represents the diminishing relevance of conventional marketing approaches in professional compensation discussions. The discourse analysis reveals that marketing professionals increasingly perceive traditional marketing skills as baseline expectations rather than differentiating competencies worthy of premium compensation. The keyword analysis indicates that when traditional marketing concepts appear in total rewards discussions, they typically serve as contrast points to highlight the superior value of digital capabilities. Creative Benefits Theme (10.4% increase) The creative benefits theme showed moderate growth, characterized by keywords including “creative freedom,” “innovation time,” “design thinking,” “brainstorming sessions,” “artistic expression,” “creative collaboration,” “idea generation,” “concept development,” “creative recognition,” and “artistic autonomy.” This theme reflects the persistent importance of creative capabilities in marketing professional identity and compensation expectations. The discourse analysis reveals that marketing professionals value organizational support for creative processes as a distinct component of total rewards packages. The keyword analysis indicates that creative benefits discussions often emphasize psychological satisfaction and professional fulfillment rather than monetary compensation. Professional Development Theme (13.5% increase) The professional development theme demonstrated consistent growth, characterized by keywords including “skill enhancement,” “certification programs,” “training opportunities,” “career advancement,” “mentorship programs,” “leadership development,” “industry conferences,” “continuing education,” “professional networking,” and “competency building.” This theme reflects the growing importance of continuous learning in marketing professional career development. The discourse analysis reveals that marketing professionals view professional development opportunities as essential components of total rewards packages that directly impact long-term career trajectories. The keyword analysis indicates that discussions emphasize both employer-provided development programs and individual learning initiatives supported through organizational resources. Thematic relationship analysis The keyword co-occurrence analysis reveals significant relationships between themes that illuminate the complex nature of marketing professional compensation preferences. The digital marketing skills theme demonstrates strong positive correlations with both remote work benefits and professional development themes, suggesting that marketing professionals view technological competency, workplace flexibility, and continuous learning as interconnected components of contemporary total rewards packages. The declining traditional marketing theme shows negative correlations with all other themes, indicating that marketing professionals increasingly frame their compensation expectations in opposition to conventional marketing approaches. This finding suggests that traditional marketing capabilities may serve as baseline expectations rather than premium compensation drivers in contemporary marketing roles. The creative benefits theme demonstrates moderate positive correlations with professional development and remote work benefits themes, indicating that marketing professionals view creative autonomy, learning opportunities, and workplace flexibility as mutually reinforcing components of satisfying total rewards packages. Temporal pattern analysis The quarterly analysis reveals distinct temporal patterns in theme evolution that provide insights into the dynamic nature of marketing professional discourse. The digital marketing skills theme showed accelerating growth throughout the study period, with particularly pronounced increases during the second and fourth quarters when organizations typically conduct performance reviews and budget planning activities. The remote work benefits theme demonstrated steady growth with seasonal variations, showing peak discussion frequencies during summer months when marketing professionals may prioritize work-life balance considerations and during winter holiday periods when family time considerations become prominent. The professional development theme exhibited cyclical patterns aligned with traditional academic calendars and industry conference schedules, with peak discussion periods occurring during spring and fall quarters when educational and professional development opportunities are most readily available. Benefit types and sentiment correlations The analysis of specific benefit types revealed strong correlations with sentiment scores across various categories in marketing roles (see Table 5). Flexible work arrangements emerged as the benefit type most strongly correlated with positive sentiment (r = .48, p < .001), followed by digital tool allowances (r = .43, p < .001). Table 5Benefit Types and Sentiment Correlations in Marketing Roles.Benefit TypeFrequencySentiment Correlation (r)p-valueCI (95%)Flexible Work11,2450.48< 0.001[0.45, 0.51]Digital Tool Allowance9,8760.43< 0.001[0.40, 0.46]Skill Development8,9340.39< 0.001[0.36, 0.42]Creative Freedom7,8920.35< 0.001[0.32, 0.38] Benefit Types and Sentiment Correlations in Marketing Roles. Named entity recognition performance The implementation of our custom NER model demonstrated robust performance in identifying marketing-specific entities (see Table 6). The model showed particularly strong performance in identifying digital marketing tools (F1 = 0.88) and marketing-specific job roles (F1 = 0.86). Table 6NER Model Performance for Marketing-Specific Entities.Entity TypePrecisionRecallF1-ScoreSupportDigital Marketing Tools0.890.870.883,245Marketing Job Roles0.870.850.862,934Marketing Benefits0.840.820.832,567Industry Sectors0.820.800.812,123Overall0.860.840.8510,869 NER Model Performance for Marketing-Specific Entities. These results demonstrate significant patterns in how marketing professionals perceive and discuss total rewards on LinkedIn. The strong performance metrics of our analytical models (overall F1 = 0.85) provide confidence in the reliability of these findings. At the same time, the comprehensive coverage of our dataset ensures robust generalizability across different marketing contexts in the identified industry sectors. Operational validation of conceptual frameworks The systematic application of conceptual definitions and operational frameworks enabled precise identification and quantitative analysis of key discourse patterns in marketing professional total rewards discussions. Digital skills discourse analysis results The operational framework identified 12,847 comments (36.0% of total dataset) containing digital skills discourse based on the systematic recognition criteria. The most frequently mentioned digital skills categories were Data Analytics Skills (4,234 comments, 33.0% of digital skills discourse), Social Media Marketing Skills (3,567 comments, 27.8%), and Marketing Automation Skills (2,890 comments, 22.5%). The named entity recognition validation achieved 0.88 precision and 0.86 recall for digital skills terminology identification. Representative digital skills discourse examples include: “My Google Analytics certification led to a 15% salary increase,” “Companies value data-driven marketing skills over traditional advertising experience,” and “Marketing automation expertise opened new career opportunities.” The operational framework successfully distinguished between authentic digital competency discussions and superficial technology mentions. Skill-based compensation discourse analysis results The operational framework identified 8,934 comments (25.0% of total dataset) containing skill-based compensation discourse based on competency-reward linkage criteria. The most frequently mentioned skill-based compensation categories were Premium Salary Structures (3,245 comments, 36.3% of skill-based compensation discourse), Performance-Based Rewards (2,567 comments, 28.7%), and Career Advancement Opportunities (2,123 comments, 23.8%). The validation process achieved 0.85 precision and 0.83 recall for skill-based compensation terminology identification. Representative skill-based compensation discourse examples include: “Digital marketing specialists earn 30% more than traditional marketers,” “My SEO expertise qualified me for senior-level positions,” and “Company provides bonuses for marketing automation certifications.” The operational framework successfully differentiated between competency-linked compensation discussions and general salary conversations. Creative benefits discourse analysis results The operational framework identified 7,892 comments (22.1% of total dataset) containing creative benefits discourse based on creative process support criteria. The most frequently mentioned creative benefits categories were Creative Freedom Initiatives (2,890 comments, 36.6% of creative benefits discourse), Collaborative Creative Opportunities (2,234 comments, 28.3%), and Creative Resource Access (1,789 comments, 22.7%). The validation process achieved 0.82 precision and 0.80 recall for creative benefits terminology identification. Representative creative benefits discourse examples include: “20% time for creative projects improved job satisfaction,” “Cross-functional creative teams enhanced innovation capabilities,” and “Company-sponsored design software access supports creative work.” The operational framework successfully distinguished between authentic creative support discussions and general workplace satisfaction expressions. Validation-derived interpretive findings The qualitative expert validation process revealed important contextual nuances that computational analysis alone could not capture. Expert reviewers identified that the 31.2% increase in digital skills discourse reflected not merely frequency changes but qualitative shifts in how marketing professionals conceptualized their professional value proposition and articulated compensation expectations. The semantic depth analysis conducted during expert validation revealed that 87% of comments classified as digital skills discourse through computational analysis contained substantive discussions about skill-compensation relationships, while 13% represented peripheral mentions requiring interpretive nuance for appropriate categorization. Furthermore, contextual analysis confirmed that 92% of comments demonstrating positive correlations between flexible work arrangements and positive sentiment contained explicit causal attributions linking workplace flexibility to compensation satisfaction, validating that identified correlations represent meaningful relationships rather than statistical artifacts. The decline in traditional marketing discourse (-15.3%) was confirmed through expert review to reflect not merely reduced discussion frequency but a fundamental reconsideration of traditional marketing competencies within contemporary professional value frameworks. Marketing professionals were found to actively reposition traditional skills as baseline expectations rather than premium capabilities worthy of enhanced compensation, representing qualitative transformation in professional identity frameworks rather than simple topic preference shifts. Multivariate analysis results Hierarchical regression: predicting sentiment from industry, digital skill discourse, and benefit type The hierarchical regression analysis confirmed that the bivariate patterns reported in Sect. 4.1–4.3 remain significant when predictors are modeled simultaneously, while revealing meaningful interaction effects consistent with SBTRT predictions (see Table 7). Table 7Hierarchical Regression Predicting Sentiment Score.PredictorStep 1 βStep 2 βStep 3 βComment length (log)0.06***0.05***0.04***Quarter: Q20.03***0.02**0.02**Quarter: Q30.05***0.04***0.04***Quarter: Q40.08***0.07***0.07***Region: Europe− 0.02*− 0.01− 0.01Region: Australia0.010.010.01Sector: Technology—0.14***0.12***Sector: Retail—0.07**0.05*Digital skill discourse—0.18***0.15***Benefit: Flexible work—0.22***0.21***Benefit: Digital tools—0.17***0.16***Benefit: Skill development—0.11***0.10***Digital skill × Technology——0.09***Digital skill × Retail——0.05**R²0.020.100.14ΔR²0.02***0.08***0.04***F / ΔF146.02***531.84***828.91***Note. N = 35,691. Consumer goods sector and creative freedom benefit serve as reference categories. Standardized coefficients reported. *p < .05, **p < .01, ***p < .001. Hierarchical Regression Predicting Sentiment Score. Note. N = 35,691. Consumer goods sector and creative freedom benefit serve as reference categories. Standardized coefficients reported. *p < .05, **p < .01, ***p < .001. Step 1 (control variables) accounted for a modest but significant proportion of sentiment variance (R² = 0.02, F(5, 35685) = 146.02, p < .001), with temporal quarter emerging as a significant predictor reflecting increasing positive sentiment across the study period (β = 0.08, p < .001 for Q4 relative to Q1). Comment length demonstrated a small but significant positive association with sentiment (β = 0.06, p < .001), while geographic region effects were negligible, with European location showing a marginal negative association (β = − 0.02, p < .05) and Australian location showing no significant effect (β = 0.01, ns). Step 2 (main effects) contributed substantial incremental variance (ΔR² = 0.08, ΔF(6, 35679) = 531.84, p < .001). Technology sector membership predicted higher sentiment relative to the consumer goods reference category (β = 0.14, p < .001), confirming the ANOVA results while controlling for other predictors. Retail sector membership also predicted higher sentiment relative to consumer goods, though with a smaller effect (β = 0.07, p < .01). Digital skill discourse orientation positively predicted sentiment (β = 0.18, p < .001), with professionals classified as expressing digital competency discourse demonstrating significantly higher sentiment scores than those emphasizing traditional capabilities. Among benefit types, flexible work arrangement discourse emerged as the strongest positive predictor of sentiment (β = 0.22, p < .001), followed by digital tool allowance discourse (β = 0.17, p < .001) and skill development discourse (β = 0.11, p < .001), a pattern consistent with the bivariate correlation results. Step 3 (interaction terms) contributed significant incremental variance (ΔR² = 0.04, ΔF(2, 35677) = 828.91, p < .001), indicating that the relationship between digital skill discourse and sentiment varies meaningfully across industry sectors. The digital skill discourse × technology sector interaction was positive and significant (β = 0.09, p < .001), indicating that the sentiment premium associated with digital competency discourse is amplified in technology sector contexts relative to consumer goods. The digital skill discourse × retail sector interaction was also positive and significant, though smaller in magnitude (β = 0.05, p < .01), suggesting a gradient of moderating effects across sectors ordered by digital intensity. The inclusion of interaction terms attenuated the main effects of both digital skill discourse (from β = 0.18 to β = 0.15) and technology sector membership (from β = 0.14 to β = 0.12), consistent with partial mediation of the sector effect through the digital skill–sector interaction pathway. The full model (Step 3) accounted for 14% of the variance in sentiment scores (R² = 0.14). While this effect size is moderate, it is consistent with discourse-based prediction of attitudinal outcomes where substantial unexplained variance reflects the inherent heterogeneity of organic professional expression and the influence of individual, contextual, and situational factors not captured by the predictor set. Moderated regression: sector-specific benefit–sentiment relationships Separate moderated regression analyses confirmed that the relationship between digital tool allowance discourse and positive sentiment is conditioned by industry sector (interaction term: β = 0.09, p < .001). Simple slope analysis revealed that the digital tool allowance–sentiment relationship was strongest in technology contexts (b = 0.51, p < .001), moderate in retail contexts (b = 0.38, p < .001), and weakest though still significant in consumer goods contexts (b = 0.31, p < .001). The 95% confidence intervals for these simple slopes did not overlap between technology and consumer goods sectors, confirming that the sector-specific differences in correlation strength reported in Sect. 5.3 reflect statistically meaningful moderation rather than sampling variation. These multivariate findings provide stronger empirical support for SBTRT’s proposition that industry-specific digital intensity moderates skill-reward congruence effects, as the interaction patterns remain significant after simultaneously controlling for other predictors including temporal trends, geographic region, and comment characteristics. Discussion The analysis of marketing professionals’ discourse about employer-provided total rewards on LinkedIn provides significant insights into the evolving nature of organizational compensation preferences in the digital marketing era. The findings specifically address how marketing professionals evaluate and discuss the comprehensive benefit packages offered by their current employing organizations rather than external recognition systems or alternative employment opportunities. The organizational focus enables systematic examination of how different industry sectors structure their total rewards offerings and how marketing professionals respond to these employer-provided compensation packages. This approach provides actionable insights for organizations seeking to optimize their employee value propositions through strategic total rewards design and implementation. Understanding marketing professionals’ total rewards discourse (RQ1) Marketing professionals articulate sophisticated compensation frameworks through LinkedIn discourse, complicating assumptions that employees hold simple preference hierarchies ranking monetary rewards above non-financial benefits. The strong association between flexible work arrangements and positive sentiment (r = .48, p < .001), confirmed in multivariate analysis controlling for industry sector and digital skill orientation (β = 0.21, p < .001), challenges traditional compensation theory positioning workplace flexibility as supplementary benefits rather than core reward components. This finding engages ongoing debates about total rewards conceptualization. Armstrong and Brown’s (2006) influential framework treated flexibility as a discretionary benefit separate from core compensation, while Heneman and Ledford (1998) positioned non-monetary rewards as secondary considerations after establishing competitive base pay. The discourse patterns identified in this study suggest that marketing professionals, in their public professional communications, reject these hierarchical frameworks, instead articulating flexibility as an integral reward component deserving equal consideration with monetary compensation. A necessary interpretive caveat applies to this finding and to the broader discourse-based results presented throughout this discussion. LinkedIn constitutes a professional identity marketplace where discourse serves career positioning functions alongside and potentially in tension with authentic preference expression (Goffman, 1959). The emphasis on flexibility and digital skill recognition in professional discourse may reflect both genuine preference evolution and strategic professional signaling, as marketing professionals foreground reward categories that project technological sophistication and contemporary professional identity. Consequently, the patterns identified in this study are most accurately characterized as publicly signaled reward preferences rather than direct indicators of latent individual attitudes. This framing does not diminish the analytical value of the findings publicly articulated preferences shape organizational discourse, influence compensation benchmarking, and contribute to emergent professional norms that eventually influence formal reward structures but it does require interpretive caution when drawing inferences about underlying individual preference hierarchies. This differential aligns with broader methodological debates about authentic attitude measurement, as professionals have been shown to express substantially more negative workplace sentiment through social media compared to formal surveys3, while employer-administered surveys systematically overestimate compensation satisfaction relative to independent assessments2. These methodological variations raise fundamental questions about preference measurement validity. Do discourse-based findings reveal authentic preferences that surveys suppress through social desirability bias, or do different contexts elicit different strategic presentations rather than tapping into singular authentic preferences? We resist simple claims that either methodology accesses “true” preferences, instead proposing that compensation preferences may be context-dependent and strategically constructed rather than stable individual attributes that research methodologies accurately or inaccurately measure. LinkedIn discourse and formal surveys may capture complementary facets of professional reward orientations, with neither providing privileged access to a unified underlying preference structure. The sophisticated digital skill-based compensation frameworks evident in professional discourse extend recent theoretical work on skill-based pay evolution. Cantrell and colleagues34 argued that skill-based organizational models require fundamental compensation restructuring, while Lawler’s (2000) updated skill-based pay framework emphasized competency recognition over traditional job evaluation approaches. The discourse patterns identified in this study are consistent with these theoretical developments, suggesting that marketing professionals have independently developed comparable conceptual frameworks evident in how they discuss digital competency-compensation relationships through peer discourse. This convergence between theoretical developments and professional discourse patterns is suggestive of a dynamic in which compensation theory developments may reflect rather than drive professional community understanding. However, establishing the directionality of this relationship would require longitudinal designs beyond the scope of the current study. Industry-specific variations in total rewards perception (RQ2) The significant variation in sentiment across industries (F(2,35688) = 27.32, p < .001), confirmed in multivariate analysis controlling for digital skill discourse orientation and benefit type (see Sect. 4.6), provides empirical evidence supporting industry-specific approaches to marketing compensation. These patterns are consistent with SBTRT’s proposition that sectoral characteristics moderate skill-based reward preferences. Technology sector marketing professionals demonstrated higher positive sentiment (M = 0.45, SD = 0.28) compared to retail (M = 0.38, SD = 0.31) and consumer goods (M = 0.33, SD = 0.29), representing a meaningful sentiment differential that exceeds industry variation patterns reported in previous compensation research. These industry-specific sentiment patterns converge with and extend findings from Bradley and Gilmartin’s (2024) analysis of high-impact rewards across sectors, which identified technology sector organizations as more likely to implement comprehensive total rewards packages including flexible work arrangements and digital tool access. Notably, the magnitude of sentiment variation identified in this study (0.12 point difference between technology and consumer goods sectors on the standardized sentiment scale) exceeds the satisfaction differential Bradley and Gilmartin reported through structured organizational surveys (0.06 point difference on comparable scales). This discrepancy may reflect either stronger authentic industry effects captured through discourse analysis or differential performative dynamics across sectors, as technology sector professionals may face stronger professional norms favoring positive articulation of digital reward packages. The finding that technology sector marketing professionals express higher positive sentiment is consistent with the expectation that digitally intensive industries provide organizational environments more closely aligned with contemporary professional reward expectations. This pattern aligns with Văcărasu’s (2023) research on workplace trends among information technology professionals in Romania, which identified that technology sector employees prioritize flexibility, team quality, and fairness over traditional monetary rewards. The convergence across different geographic contexts and methodological approaches increases confidence that technology sector characteristics influence professional reward discourse, though the possibility of sector-specific performative norms cannot be excluded. The lower sentiment levels among consumer goods sector marketing professionals (M = 0.33) compared to technology (M = 0.45) and retail (M = 0.38) sectors diverge from expectations based on traditional compensation frameworks, which would predict that established consumer goods corporations with robust benefits packages would generate higher employee satisfaction. This pattern suggests that contemporary marketing professionals may prioritize characteristics associated with technology sector employment including innovation culture, rapid career advancement opportunities, and digital tool access over the employment stability and comprehensive traditional benefits typically associated with established consumer goods corporations. However, this interpretation must be qualified by acknowledging that consumer goods sector professionals may face different LinkedIn engagement incentives than technology sector professionals, potentially expressing satisfaction differently rather than experiencing it differently. This finding extends Errida and Lotfi’s (2021) research on organizational change management by suggesting that industry-specific change velocities may influence how marketing professionals articulate satisfaction through reward preference alignment. It is important to note that the cross-sectional nature of these industry comparisons precludes causal inference about whether industry context produces sentiment differences or whether professionals with pre-existing preference profiles self-select into specific industries. The observed patterns are consistent with both explanations, and disentangling these mechanisms would require longitudinal tracking of professionals transitioning across sectors. Potential sub-group variations within sectors Within each industry sector, further variation likely exists between marketing professionals operating in business-to-business (B2B) versus business-to-consumer (B2C) contexts. B2B marketing professionals typically engage in longer sales cycles, relationship-driven strategies, and technical content development, which may generate distinct reward preferences emphasizing professional development, industry conference access, and client relationship management support. In contrast, B2C marketing professionals often prioritize creative freedom, campaign performance bonuses, and access to consumer analytics platforms. The distinct skill profiles and performance measurement criteria characterizing these two marketing domains suggest that total rewards packages optimized for B2B contexts may differ substantively from those generating satisfaction in B2C environments. This sub-group distinction represents a theoretically motivated extension that the current dataset cannot empirically adjudicate. LinkedIn profile information does not consistently distinguish B2B and B2C orientations, and employer-level classification introduces substantial ambiguity for companies operating across both domains. Preliminary qualitative examination of discourse patterns offers suggestive but not systematic indications that B2B-oriented comments emphasize relationship continuity, thought leadership development, and account-based marketing tool access more frequently than B2C-oriented comments, which more commonly reference creative campaign recognition, consumer insight platforms, and social media engagement metrics. We flag this variation as a priority for future research employing supplementary classification criteria, such as employer product category analysis, self-reported role descriptions, or industry-specific keyword clustering, rather than presenting it as an empirically grounded finding of the current study. Industry-specific benefit valuations (RQ2b) The association between digital tool allowance discourse and positive sentiment (r = .43, p < .001), confirmed in multivariate analysis and particularly pronounced within the technology sector, indicates that industry-specific tools and resources represent important predictors of marketing professional sentiment beyond generic benefit categories. The moderated regression analysis (Sect. 4.6.2) provides formal confirmation that the relationship between digital tool discourse and positive sentiment is conditioned by industry sector, with the technology sector demonstrating significantly stronger associations than consumer goods contexts, and non-overlapping confidence intervals confirming that this moderation effect is statistically meaningful. This finding provides quantitative evidence consistent with Klompen and Walsh’s (2023) qualitative conclusions that digital talent prioritizes access to cutting-edge technology tools, while extending those conclusions through systematic measurement of the strength and sector-contingent nature of tool access-sentiment relationships. The technology sector demonstrated the strongest association between digital tool allowances and positive sentiment (r = .51, p < .001), compared to the retail sector correlation of r = .38 and the consumer goods sector correlation of r = .31. This variation in association strength across sectors is consistent with SBTRT’s proposition that industry characteristics moderate the relationship between digital skill recognition and reward satisfaction, with digitally intensive industries demonstrating stronger skill-reward congruence effects. These industry-specific patterns diverge from Terera and Ngirande’s (2014) cross-sectional survey findings examining 200 marketing professionals, which reported relatively uniform relationships between benefit categories and job satisfaction across industry contexts. The divergence may reflect either genuine industry-specific variation that traditional survey formats constrain respondents from expressing, or methodological differences between discourse-based association measurement and structured questionnaire approaches. Both explanations are plausible, and the current study cannot adjudicate between them. The finding that skill development benefits demonstrate positive sentiment associations in consumer goods sectors (r = .42) comparable to technology sectors (r = .44), despite lower overall sentiment levels, suggests that professional development opportunities represent broadly valued reward components across industry contexts. This pattern aligns with McIlveen and colleagues’ (2021) findings that career development opportunities significantly influence job satisfaction across diverse professional contexts. Evolution of marketing benefits preferences (RQ3) The substantial shift in discourse toward digital marketing skills compensation (+ 31.2%) and away from traditional marketing role discussions (-15.3%) represents a notable transformation in how marketing professionals publicly articulate career value and compensation expectations. This evolution, identified through topic modeling analysis demonstrating statistically significant thematic shifts (χ²(4) = 182.45, p < .001), is consistent with SBTRT’s Market Value of Skills Principle, which predicts that digital competency market premiums increase proportionally to technological change velocity. The 31.2% increase in digital skills discourse throughout 2023 exceeds the temporal evolution patterns Smet, Dowling, and Lim (2022) identified in their comprehensive survey of 1,200 marketing professionals across technology and retail sectors conducted in 2021. Their survey-based analysis identified a 12% increase in emphasis on digital competencies over a comparable twelve-month period. The magnitude difference between these studies may reflect a combination of accelerating digital transformation trends between 2021 and 2023, methodological differences in measuring preference evolution through organic discourse versus structured questionnaire responses, and the possibility that LinkedIn discourse amplifies emerging professional norms around digital skill emphasis through social reinforcement mechanisms not present in private survey contexts. These explanations are not mutually exclusive, and the current study cannot isolate the contribution of each factor. The declining emphasis on traditional marketing discourse (-15.3%) converges with Gillpatrick’s (2019) theoretical analysis of digital transformation’s impact on marketing practice, which anticipated that traditional advertising and brand management approaches would diminish in perceived professional value as data-driven digital strategies gained prominence. The discourse patterns identified in this study are consistent with this theoretical anticipation, though the causal mechanisms remain underspecified the decline in traditional marketing discourse may reflect genuine devaluation of these competencies in compensation markets, shifting professional signaling norms on LinkedIn, or a combination of both dynamics. The temporal evolution patterns identified in this study contrast with the relatively stable preference structures reported in WorldatWork’s (2022) Total Rewards Inventory, which found limited year-over-year variation in marketing professionals’ stated compensation priorities. This divergence is consistent with the possibility that formal survey contexts underestimate the dynamic nature of professional preference evolution, potentially because structured questionnaires measuring predetermined categories cannot capture emerging compensation preferences until they have already become established. However, it is equally possible that discourse-based measurement amplifies emerging trends through platform-specific dynamics, and the “true” rate of preference evolution falls between the estimates derived from these methodologically distinct approaches. Temporal changes in benefits valuation (RQ3b) The increase in discussions about skill development benefits (r = .39, p < .001) and creative freedom (r = .35, p < .001) indicates growing emphasis on growth opportunities and professional autonomy in marketing professionals’ public reward discourse. These patterns challenge traditional marketing compensation models emphasizing fixed compensation structures and standardized benefits packages, and are consistent with SBTRT’s integration of Self-Determination Theory principles, which predict that contemporary professionals increasingly prioritize rewards supporting competence development and autonomous work arrangements. The skill development benefits correlation identified in this study (r = .39) exceeds the professional development-satisfaction relationship Hoole and Hotz35 reported in their analysis of total reward system impacts on work engagement (r = .28). This difference may reflect strengthening of the development-satisfaction relationship as digital transformation accelerated skill obsolescence concerns between 2016 and 2023, though cross-study comparison must be approached cautiously given differences in measurement methodology, sample composition, and geographic context. The creative freedom association (r = .35) aligns with36 findings that social recognition and employee engagement demonstrate significant positive relationships in organizational contexts. The current study extends those findings by suggesting that creative autonomy may represent a distinct reward category valued alongside but separately from social recognition, with marketing professionals articulating specific preferences for organizational support of creative processes rather than merely desiring acknowledgment of creative outputs. The temporal evolution showing increased emphasis on both skill development and creative freedom suggests that marketing professionals’ public discourse increasingly emphasizes comprehensive professional growth environments supporting both technical competency advancement and creative expression opportunities. This dual emphasis diverges from traditional compensation frameworks treating technical training and creative autonomy as alternative benefit categories appealing to different employee segments, suggesting instead that digital-era marketing professionals may view technological competence and creative capability as complementary rather than competing professional attributes though this interpretation warrants empirical confirmation through individual-level preference elicitation studies. Theoretical contributions and preliminary empirical support for SBTRT This research offers three contributions advancing Skill-Based Total Rewards Theory: preliminary empirical support through discourse-based operationalization, industry-specific theoretical refinement, and methodological innovation establishing new frameworks for testing compensation theory through professional discourse analysis. Preliminary empirical support and quantitative operationalization The study provides the first systematic empirical examination of SBTRT propositions through analysis of authentic professional discourse patterns, transforming abstract theoretical propositions into quantifiable relationships enabling structured testing. The identification of specific association patterns between digital skill discourse and positive sentiment (r = .48 for flexible work arrangements, r = .43 for digital tool allowances) operationalizes the Digital Skill-Reward Congruence Principle through quantitative measurement, while the hierarchical regression analysis (Sect. 4.6) confirms that these associations persist when modeled simultaneously with industry sector and other predictors. The results are consistent with the Digital Skill-Reward Congruence Principle’s prediction that marketing professionals articulating digital competencies express more positive sentiment toward total rewards when organizations provide skill-specific recognition. The multivariate finding that digital skill discourse orientation interacts with industry sector to predict sentiment (ΔR² = 0.04, ΔF(2, 35677) = 828.91, p < .001) provides stronger evidence for this proposition than bivariate correlations alone, as it demonstrates that the digital skill–sentiment relationship varies systematically by sector after controlling for confounds. The 31.2% increase in digital skills discourse throughout 2023 is consistent with the Market Value of Skills Principle, suggesting temporal patterns aligned with theoretical predictions about skill-based compensation evolution. The systematic increase in digital competency discussions across all three industry sectors suggests that market-driven skill valuation changes operate consistently across organizational contexts, though the single-year observation window limits confidence in distinguishing sustained trends from transient fluctuations. The Skills-Based Flexibility Principle receives preliminary support through the identification of industry-specific patterns in flexible work arrangement discourse, with technology sector professionals demonstrating higher positive sentiment (M = 0.45) compared to retail (M = 0.38) and consumer goods (M = 0.33) sectors. The moderated regression analysis confirming sector-specific variation in the digital tool access–sentiment relationship provides formal evidence that SBTRT relationships are conditioned by industry context. We characterize these findings as preliminary empirical support rather than theoretical validation for three reasons. First, the correlational and cross-sectional design precludes causal inference about the mechanisms SBTRT proposes. The theory specifies directional relationships for instance, that digital skill mastery generates flexibility expectations but the observed associations are consistent with multiple causal orderings. Second, the single-year observation period limits assessment of the temporal dynamics central to the Market Value of Skills Principle, which predicts increasing digital skill premiums over time in proportion to technological change velocity. Third, the performative characteristics of LinkedIn discourse introduce uncertainty about whether identified patterns reflect underlying preference structures or publicly constructed professional identity positions. While these limitations do not invalidate the empirical contributions, they establish boundary conditions for theoretical inference and underscore the need for longitudinal, multi-method research to advance SBTRT toward more robust empirical standing. Industry-specific theoretical extensions and moderating factors The research contributes to theoretical refinement through identification of industry-specific moderating factors that condition SBTRT applications across sectoral contexts. The systematic variation in sentiment patterns across technology, retail, and consumer goods sectors, confirmed through both ANOVA and hierarchical regression analyses, reveals that SBTRT relationships differ depending on industry characteristics. Technology sector emphasis on digital tool access and flexible work arrangements, retail sector focus on performance-based compensation, and consumer goods sector emphasis on career development opportunities represent industry-specific patterns that can inform continued SBTRT development. These sectoral patterns suggest that digital intensity, customer interaction requirements, and organizational change velocity may serve as moderating variables influencing which specific reward components demonstrate strongest associations with expressed professional sentiment within skill-based compensation frameworks. The identification of distinct reward discourse patterns across industry sectors extends SBTRT beyond universal propositions to context-dependent applications. The technology sector’s stronger association between digital tool allowances and positive sentiment compared to consumer goods sectors demonstrates that industry-specific digital intensity conditions skill-reward congruence effects, suggesting that organizations may need to calibrate total rewards packages based on sectoral characteristics and workforce digital competency expectations. Methodological innovation for theoretical development The study contributes methodological approaches for theoretical development through integration of natural language processing techniques with professional discourse analysis, providing frameworks for examining theoretical propositions through systematic analysis of organic professional expression patterns. The combination of sentiment analysis, topic modeling, named entity recognition, and multivariate regression applied to authentic professional discourse creates analytical capabilities for examining theoretical propositions within specific professional contexts that complement traditional survey methodologies. The development of marketing-specific sentiment analysis models represents methodological innovation enabling domain-specific theoretical testing through custom lexicon development and model calibration. The integration of multiple analytical techniques within a comprehensive validation framework combining computational performance assessment, qualitative expert review, and multivariate statistical analysis provides methodological advancement applicable to future theoretical testing across diverse professional contexts. Ethical considerations, responsible implementation, and critical reflections on workplace discourse analytics The application of social media discourse analysis to human resource decision-making raises fundamental questions about surveillance, consent, privacy boundaries, and the potential for data misuse that extend beyond traditional research ethics frameworks into contemporary debates about algorithmic management and digital workplace monitoring. This research engages substantively with these critical concerns while acknowledging the inherent tensions between organizational knowledge generation and employee rights. Critical engagement with surveillance and algorithmic management debates The systematic analysis of professional discourse for organizational research purposes operates within contested terrain where legitimate market intelligence gathering intersects with potentially problematic employee monitoring practices. Recent scholarship on algorithmic management (Kellogg, Valentine, & Christin, 2020) demonstrates that computational analysis of employee-generated data enables increasingly sophisticated surveillance capabilities that fundamentally alter power dynamics within employment relationships, creating asymmetric information conditions where organizations possess comprehensive insights about employee attitudes while individuals lack comparable organizational transparency. The current research methodology focuses deliberately on aggregate pattern identification rather than individual employee assessment, establishing clear boundaries between appropriate strategic market intelligence and inappropriate individual surveillance. However, this methodological choice does not eliminate ethical concerns, as the identical analytical techniques employed for population-level pattern identification could be redirected toward individual employee monitoring with minimal technical modification. This dual-use potential creates ethical obligations extending beyond the immediate research application to consider how similar methodologies might be deployed in organizational contexts lacking comparable ethical constraints. Consent, contextual integrity, and data governance The ethical framework of this research must contend with the concept of contextual integrity (Nissenbaum, 2010), which holds that privacy expectations are governed by context-specific norms about information flows rather than by binary public-private distinctions. LinkedIn users share professional content within a context governed by norms of professional networking, career development, and peer engagement. The systematic aggregation and computational analysis of this content for compensation research purposes represents a different information flow than users likely anticipated when composing their posts even though the content is technically public. This contextual integrity concern has practical implications for organizational adoption of discourse analytics. When organizations implement social media monitoring systems to inform compensation strategy, they alter the informational norms governing professional social media participation. If professionals become aware that their LinkedIn discourse influences organizational compensation decisions, this awareness may modify their expression patterns in ways that undermine the methodological rationale for discourse analysis namely, that organic professional discourse captures preferences less constrained by social desirability bias than formal surveys. The recursive relationship between monitoring awareness and expression behavior creates a fundamental epistemological challenge for sustained organizational application of discourse analytics methodologies. Furthermore, existing data governance frameworks, including the General Data Protection Regulation (GDPR) in European contexts, establish legal requirements regarding the processing of personal data that intersect with discourse analytics in complex ways. While aggregated and anonymized analysis of public professional content may fall within permissible research use, organizational application of comparable techniques for human resource decision-making enters regulated territory where consent requirements, data minimization principles, and purpose limitation obligations may apply. Organizations contemplating implementation of discourse-based compensation analytics should obtain legal counsel regarding jurisdiction-specific data governance requirements and should establish explicit data governance protocols addressing collection scope, retention periods, access controls, and permissible use limitations. Blurred boundaries between professional self-presentation and authentic sentiment The methodological assumption that LinkedIn discourse reflects authentic professional sentiment requires critical examination considering the performative nature of professional social media communication. Goffman’s (1959) theoretical framework on impression management demonstrates that individuals strategically construct public presentations of professional identity designed to advance career objectives rather than expressing unfiltered authentic perspectives. Marketing professionals engaging in LinkedIn discourse may emphasize particular compensation preferences to signal professional sophistication or align with perceived industry norms rather than expressing genuine personal preferences. The performative dynamics of LinkedIn discourse extend beyond sentiment expression to influence which reward preferences gain visibility in professional discourse. Marketing professionals may strategically foreground preferences for digital skill recognition and flexible work arrangements topics signaling technological sophistication and alignment with contemporary professional norms while underrepresenting preferences for traditional monetary compensation, job security, or conventional benefits that may be perceived as professionally conservative or insufficiently forward-looking. This selective visibility creates potential asymmetry between publicly expressed and privately held reward preferences. In anonymous channels such as platforms like Glassdoor, Blind, or confidential organizational engagement surveys professionals may more freely articulate preferences for base salary increases, job stability assurances, or traditional benefits that they strategically minimize in LinkedIn’s professional identity marketplace. The implication is twofold. First, the strong emphasis on digital skills and flexible work arrangements identified in this study may partially reflect professional signaling behavior alongside authentic preference evolution. Second, the relatively muted discourse around traditional compensation structures may understate the continued importance of base salary competitiveness and financial security as foundational reward expectations. This tension between authentic expression and strategic self-presentation shapes how organizational decision-makers should weight discourse-derived insights. Organizations implementing discourse analysis insights should treat these findings as reflecting publicly constructed professional identity positions that complement rather than replace confidential assessment instruments capturing the full spectrum of reward valuations, including preferences professionals may strategically minimize in public professional contexts. Power asymmetries and employee voice An additional ethical dimension concerns the power asymmetries inherent in organizational discourse analytics. When organizations analyze employee social media discourse to inform compensation strategy, they exercise informational power that employees may not be aware of and cannot meaningfully consent to or opt out of, given the professional necessity of LinkedIn participation in many industries. This asymmetry is particularly pronounced for marketing professionals, whose occupational roles require active LinkedIn engagement, effectively limiting their ability to withdraw from potential monitoring without professional cost. The normative question of whether organizations should monitor professional social media discourse for compensation purposes even through aggregate anonymized analysis remains contested within the organizational ethics literature. Arguments favoring such monitoring emphasize improved organizational responsiveness to employee needs and evidence-based compensation strategy development. Arguments opposing it emphasize employee autonomy, the chilling effects of surveillance awareness on authentic professional expression, and the potential for discriminatory application if discourse patterns correlate with protected characteristics. This study does not resolve this normative debate but acknowledges that the analytical techniques demonstrated herein carry responsibilities extending beyond research ethics to encompass organizational governance, employee relations, and workplace democracy considerations. Responsible implementation framework and governance requirements Organizations seeking to apply social media discourse analysis insights require comprehensive ethical frameworks distinguishing between appropriate strategic applications and inappropriate individual monitoring practices. Appropriate strategic applications include analyzing industry-wide compensation trends across professional populations, understanding sector-specific professional preferences to guide total rewards strategy development, and identifying emerging benefit categories through market-level discourse patterns. These applications focus on market intelligence development rather than individual employee assessment. Inappropriate applications requiring explicit prohibition include individual employee social media monitoring for performance evaluation purposes, using social media discourse patterns to identify potentially dissatisfied employees for targeted retention interventions or preemptive termination decisions, and implementing discriminatory practices based on social media discourse patterns correlating with protected demographic characteristics. The implementation framework requires robust organizational governance structures including oversight committees with human resource professionals, legal counsel, employee representatives, and data ethics specialists monitoring appropriate use of discourse analysis findings. These governance structures should establish explicit policies regarding data collection scope, analytical boundaries, retention limitations, and prohibited applications, with regular audit procedures ensuring compliance. Advancing responsible practice through transparent communication Organizations implementing insights derived from social media discourse analysis should establish transparent communication policies informing employees about the general use of industry discourse analysis for compensation strategy development. This transparency should include clear distinctions between analyzing public professional discourse trends for strategic purposes and monitoring individual employee social media activity for assessment purposes. The communication framework should emphasize that organizational compensation strategy development incorporates insights from multiple data sources including industry research, competitive market analysis, and professional discourse trends, situating discourse analysis within broader compensation research practices. Conclusion Drawing from our analysis of 35,691 LinkedIn comments by marketing professionals, this research provides insights into how total rewards perceptions are publicly articulated in the contemporary marketing landscape. By addressing our three research questions through integrated computational analysis, we have identified patterns in marketing professionals’ discourse about compensation and benefits, with particular emphasis on the role of digital transformation in shaping publicly expressed reward preferences. Our sentiment analysis revealed substantial variations across industry sectors (F(2,35688) = 27.32, p < .001), with technology sector marketing professionals demonstrating notably higher positive sentiment (M = 0.45, SD = 0.28) toward their total rewards packages. Hierarchical regression analysis confirmed that these sector differences persist when modeled simultaneously with digital skill discourse orientation and benefit type, while revealing significant interaction effects (ΔR² = 0.04, p < .001) indicating that digital competency emphasis differentially predicts sentiment across industry contexts. These findings directly address RQ2, highlighting the role of industry context in shaping compensation discourse patterns. The strong association between flexible work arrangements and positive sentiment (r = .48, p < .001) further emphasizes the evolving nature of publicly expressed value perceptions among marketing professionals. In response to RQ3, our topic modeling analysis identified a significant shift toward digital marketing skills compensation discourse (+ 31.2%) and away from traditional marketing role discussions (-15.3%). This evolution in discourse themes (χ²(4) = 182.45, p < .001) indicates a transformation in how marketing professionals publicly conceptualize career value and articulate compensation expectations. These findings provide preliminary empirical support for the three propositions of Skill-Based Total Rewards Theory, while establishing boundary conditions for its application across industry sectors. We characterize this support as preliminary because the correlational design precludes causal inference, the single-year observation period limits temporal generalization, and the performative characteristics of LinkedIn discourse introduce interpretive complexity regarding the relationship between publicly articulated preferences and underlying attitude structures. Research limitations and boundary conditions This research demonstrates methodological innovation while acknowledging specific limitations affecting interpretation and generalizability. Sample representativeness and geographic constraints The study’s exclusive focus on English-language discourse from professionals in North America, Europe, and Australia creates geographic and linguistic limitations constraining generalizability to emerging economy contexts and non-English-speaking professional populations. The absence of emerging economy professionals from Asia, Latin America, and Africa represents a significant limitation, as digital transformation dynamics, labor market conditions, cultural workplace norms, and professional social media engagement patterns may differ substantially from those observed in developed English-speaking markets. The over-representation of professionals with advanced educational backgrounds (68% possessing bachelor’s degrees or higher) may not reflect global marketing professional demographic composition. Platform selection bias and self-selection effects The LinkedIn platform introduces multiple layers of selection bias requiring explicit acknowledgment. First, professionals who maintain active LinkedIn engagement may hold inherently more positive attitudes toward technology-mediated professional communication, potentially inflating observed emphasis on digital skills discourse and flexible work arrangement preferences. Second, professionals who comment on compensation-related content represent a self-selected subset of LinkedIn users who may hold stronger opinions about total rewards than the broader marketing professional population. Third, platform-specific norms governing professional discourse on LinkedIn including implicit expectations about professional tone, topical relevance, and career positioning shape expression patterns in ways that platform-neutral preference measurement would not. These self-selection and platform effects introduce systematic biases that limit the generalizability of findings to the broader population of marketing professionals, including those who prefer alternative professional networking platforms, who maintain passive rather than active LinkedIn presence, or who discuss compensation preferences primarily through private channels. Sectoral scope The exclusion of healthcare, education, and nonprofit sector marketing professionals represents a deliberate analytical choice enabling sectoral coherence while creating boundaries limiting theoretical generalization across all industry contexts. These sectors may exhibit different organizational structures, compensation constraints, and professional culture characteristics that produce distinct total rewards discourse patterns. Temporal context and dual transformation The study’s temporal scope encompasses a post-pandemic period where COVID-19 had already fundamentally altered workplace expectations regarding flexibility and remote work arrangements, creating interpretive complexity in distinguishing digital transformation effects from pandemic-induced workplace evolution. The findings reflect the combined influence of technological change and pandemic-accelerated workplace transformation rather than isolated digitalization effects. The single-year data collection period throughout 2023 provides robust cross-sectional analysis while constraining longitudinal interpretation. Correlational design and causal limitations The correlational and cross-sectional design constitutes a fundamental limitation for theoretical inference. SBTRT specifies directional causal mechanisms for instance, that digital skill mastery generates flexibility expectations but the observed associations are consistent with multiple causal orderings. Professionals who value flexibility may gravitate toward digital skill development because these competencies enable flexible work, rather than developing flexibility expectations as a consequence of digital skill acquisition. Similarly, industry-sector differences in sentiment may reflect self-selection of professionals with particular preference profiles into specific industries rather than industry context producing preference differences. Establishing causal directionality requires experimental or longitudinal designs beyond the scope of this study. Performative discourse and interpretive constraints As discussed extensively in Sect. 5.7, the performative nature of LinkedIn discourse means that publicly expressed reward preferences may not fully represent the complete spectrum of authentic professional compensation valuations. Professionals may strategically foreground certain preferences while minimizing others in professional networking contexts, creating potential asymmetry between publicly articulated and privately held reward orientations. The findings are most accurately interpreted as reflecting publicly constructed professional identity positions regarding total rewards rather than direct measures of latent individual attitudes. Construct validity implications of performative professional discourse The performative nature of professional social media discourse carries fundamental construct validity implications that warrant explicit and reflective examination. The central dependent construct of this study “total rewards perceptions” was operationalized through computational analysis of LinkedIn comments rather than through validated psychometric instruments designed to measure individual attitudinal constructs. This operationalization introduces a critical distinction between discourse about total rewards and perceptions of total rewards as traditionally conceptualized in organizational behavior research. Specifically, the dataset captures publicly articulated professional commentary on compensation-related topics, but it remains empirically indeterminate whether these comments reflect (a) authentic evaluations of one’s own organization’s rewards system, (b) general opinions about industry compensation trends, (c) professional signaling behavior intended to cultivate a particular professional identity, or (d) reactive commentary on content posted by other users rather than expressions of personal experience. This ambiguity fundamentally affects the construct validity of the study’s central dependent concept and necessitates interpretive caution. Drawing on Goffman’s (1959) dramaturgical framework, LinkedIn functions as a “front stage” environment in which professionals perform curated versions of their occupational identities for audiences comprising colleagues, recruiters, and industry peers. Within this performative context, expressions of reward preferences are not value-neutral reports of internal states but strategic communicative acts embedded in professional self-presentation logics. For marketing professionals in particular whose occupational competencies center on strategic communication and audience management the performative dimension of LinkedIn discourse may be especially pronounced, as these professionals possess heightened expertise in crafting messages calibrated to specific audiences and platforms. Consequently, the construct captured in this study is more precisely characterized as “publicly performed total rewards discourse” rather than “total rewards perceptions” in the psychometric sense, and the findings should be interpreted accordingly as patterns within professional discourse ecosystems rather than as direct windows into individual attitudinal structures. This construct validity consideration has cascading implications for several empirical claims advanced in this study. The strong positive association between flexible work arrangement discourse and positive sentiment (r = .48, p < .001), for instance, may reflect not only authentic preference strength but also the social desirability of expressing pro-flexibility positions within contemporary professional networking norms, where flexibility advocacy signals progressive workplace values and technological adaptability. Similarly, the observed decline in traditional marketing discourse (− 15.3%) may partially represent shifting professional signaling conventions whereby publicly emphasizing traditional competencies carries reputational risk in digitally oriented professional communities rather than a straightforward devaluation of these competencies in professionals’ private reward calculations. The industry-sector sentiment differences reported in Sect. 5.2 may likewise reflect sector-specific performative norms (e.g., technology professionals facing stronger expectations to publicly celebrate digital reward packages) alongside genuine sectoral variation in reward satisfaction. Furthermore, the unit of analysis introduces additional construct validity concerns. The study analyzes individual LinkedIn comments rather than identifiable individual professionals, meaning that the independence of observations cannot be guaranteed. Multiple comments from the same user may appear in the dataset, introducing potential clustering effects that inflate statistical significance and violate the independence assumption underlying the reported inferential statistics. Without user-level identifiers, it is not possible to model within-person consistency or account for individual-level confounds, which further constrains the degree to which observed discourse patterns can be attributed to population-level preference structures. Notwithstanding these construct validity limitations, it is important to note that the performative dimension of professional discourse does not render the findings analytically vacuous. Publicly performed reward preferences constitute a substantively meaningful phenomenon in their own right: they shape professional norms, influence organizational compensation discourse, and contribute to the social construction of reward expectations within occupational communities (Marwick & Boyd, 2011). The patterns identified in this study thus retain analytical significance as evidence of how total rewards are publicly negotiated and valorized within marketing professional communities, even if they cannot be straightforwardly equated with latent individual attitudes. Future research should pursue convergent validity strategies including triangulation with anonymous survey instruments, incentive-compatible preference elicitation, and cross-platform comparative analysis to establish the degree of correspondence between publicly performed reward discourse and privately held reward preferences, thereby advancing construct refinement for discourse-based organizational research. Future research directions Multimethod causal inference research Future research should prioritize designs capable of establishing causal directionality for the relationships identified in this study. Quasi-experimental research examining compensation preference changes following organizational total rewards policy modifications would provide stronger evidence for causal mechanisms. Experimental preference elicitation studies comparing stated preferences across different measurement contexts including LinkedIn-style public discourse tasks, anonymous survey instruments, and incentive-compatible choice experiments would help disentangle performative self-presentation effects from authentic preference expression, directly addressing a core limitation of the current study. Cross-platform comparative analysis Comparative studies examining reward preference discourse across platforms with different identity norms including anonymous professional forums such as Glassdoor and Blind alongside professional networks such as LinkedIn would clarify the extent to which platform-specific performative dynamics shape observed preference patterns. Such comparisons would enable researchers to identify which reward categories are consistently emphasized across contexts (suggesting robust preferences) versus those that appear platform-dependent (suggesting performative influence), advancing both methodological understanding and substantive compensation theory. Longitudinal SBTRT Development Extended longitudinal studies tracking professional discourse evolution over multi-year periods would distinguish temporary post-pandemic adjustment patterns from sustained transformations in total rewards preferences, enabling more confident assessment of SBTRT’s Market Value of Skills Principle. Time-series analysis examining discourse patterns across pre-pandemic, pandemic, and post-pandemic periods would isolate digital transformation effects from COVID-19-induced workplace changes. Practical implications for adaptive reward system design The findings offer three actionable pathways for managers seeking to design adaptive, skill-sensitive reward systems, with the caveat that discourse-derived insights should complement rather than replace confidential employee preference assessment through traditional engagement surveys. First, organizations can implement discourse-informed compensation benchmarking by systematically monitoring industry-specific professional discourse to identify emerging skill categories commanding premium compensation expectations before these trends appear in formal salary surveys. The 31.2% increase in digital skills discourse identified in this study signals compensation expectation shifts that proactive organizations could address through anticipatory reward adjustments. Practically, HR teams and compensation analysts can integrate periodic social media discourse monitoring alongside traditional compensation survey data to create early-warning indicators of shifting professional expectations, while recognizing that discourse trends may amplify emerging preferences relative to their actual prevalence in the broader workforce. Second, organizations should develop skill-sensitive reward calibration mechanisms that differentiate compensation structures based on verified digital competency portfolios rather than traditional job title hierarchies. The significant industry-specific variations identified in this study suggest that effective reward calibration requires sector-specific approaches. Managers in technology-intensive contexts should prioritize digital tool allowances, certification reimbursement programs, and flexible work arrangements as retention levers, while those in consumer goods sectors may achieve greater impact through career development investments and skill certification pathways. Acknowledgements No acknowledgements are applicable for this manuscript. Author contributions Author Contributions: S.T.- Conceptualization, Y.S.B.- Methodology, Formal Analysis, S.T.-C.C-A.B- Writing – Original Draft, Y.S.B.- A.K. Visualization. All authors have read and approved the final manuscript. Funding This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. Data availability The datasets generated and/or analysed during the current study are not publicly available due to LinkedIn’s data privacy policy, but are available from the corresponding author on reasonable request. Declarations Competing interests The authors declare no competing interests. Ethical Approval This article does not contain any studies with human participants performed by authors. Informed Consent This article contains no studies with human participants performed by authors. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The datasets generated and/or analysed during the current study are not publicly available due to LinkedIn’s data privacy policy, but are available from the corresponding author on reasonable request.