A fashion influencer recommending beauty products to her followers Photo: TUCHONG

The gig economy has led to the emergence of numerous new forms of employment and occupational groups, while simultaneously transforming the power relations embedded in labor. The traditional employee–employer relationship has been replaced by a triangular structure involving demand-side clients, digital platforms, and self-employed workers. This trend has likewise triggered profound changes within the cultural and creative industries, reshaping the modes and practices of labor among traditional cultural producers and giving rise to a new occupational group—digital platform content creators. In a broader sense, this group can be categorized as digital platform cultural producers or creative laborers within the gig economy.

A classic concern in labor sociology has always been how capital extracts surplus value from labor. In current scholarship, two dominant theoretical approaches—structural and cultural—frame this question. However, digital platform cultural producers have transcended traditional forms of work organization and now operate within open commercial markets. Algorithms manage labor primarily through mechanisms such as task allocation, dynamic pricing, spatiotemporal organization, rule-setting, and performance evaluation. Yet this study finds that algorithms exert only limited direct intervention or control over the labor process. This raises an essential question: What kinds of managerial mechanisms does capital establish to regulate workers’ practices in this new context?

The theoretical contribution of this study lies in uncovering the novel forms of labor and the new modes of labor control distinctive to the digital economy. To explore these issues, the research focuses on a particular group of digital content creators who have drawn growing public and scholarly attention in recent years—fashion and lifestyle influencers on social media.

Regulative control mechanism: Algorithms and brands co-constructing profit distribution

Algorithms, through mechanisms of content recommendation and the construction of quantitative indicators, measure users’ attention to and recognition of influencers’ charisma. The quantification of charm produces several interrelated effects. First, algorithms generate comparability and differentiation among influencers, providing brands with data for quantitative evaluation. At the same time, algorithmic systems create conditions for influencers’ visibility and interaction while also imposing constraints. They also establish a competitive arena in which influencers vie for audience attention and affection. The combination of this competition for attraction and a “decentralized” mechanism of traffic control renders account data highly fluid. Within this classificatory and ranking system, individuals are datafied and quantified. For brands and intermediary agencies, each influencer is transformed into a continuously calculable, classifiable, rankable, and reconstructible “data unit.”

Brands rely on algorithmically generated data to conduct quantitative analyses of these data units, seeking to maximize marketing efficiency within a fixed budget—reaching the largest possible audience at the lowest possible cost. The logic of quantification rests on multiple assessment dimensions, encompassing follower numbers, recent performance data, follower demographics, qualitative traits of content such as lifestyle and aesthetics, and the influencer’s commercial production capacity, measured through indicators like the number and type of sponsored posts and their performance metrics. Based on these dimensions, brands determine pricing and selection by situating each influencer within a flexible, multidimensional hierarchy. The same influencer may occupy different positions across dimensions, creating a pattern of misaligned rankings.

In pricing, follower count remains the dominant factor, with commercial rates generally fluctuating between 5% and 20% of total followers. In selection, follower count again plays a central role, yet brands do not necessarily favor top-tier influencers. Instead, they often adopt a matrix-style investment strategy that combines a few high-profile influencers with a smaller number of mid-tier and a large number of lower-tier ones. In this way, brands—together with the algorithmic quantification systems on which they rely—construct a differentiated structure for distributing economic benefits.

An influencer’s position within this structure depends on the composite data unit that represents their account, integrating overall data, recent performance, follower demographics, and qualitative characteristics. Constantly datafied, quantified, and commodified, workers are situated within a hierarchical order of profit distribution centered around an object of desire—data capital. The process through which influencers strive to maintain or improve their position within this hierarchy becomes a process of accumulating data capital. As data capital represents the quantification of charisma, influencers’ labor ultimately takes the form of charisma labor driven by the pursuit of data capital.

‘Charisma labor:’ Driven by profits and data, with exploitation extending inward

“Charisma labor” refers to the process of managing and cultivating the self and personal image to generate attraction and influence over others. This study identifies three key dimensions of self-production through which charisma labor unfolds: comprehensive self-management, compulsory yet strategic self-exposure, and the immediate and continuous regulation and transformation of the self. Data-driven charisma labor functions as an ongoing process in which personal appeal is constantly refined and sustained. The “entrepreneurial self” persistently reconfigures its own subjectivity, while the growth of data and the development of the self reinforce one another. The influencer’s deliberate planning and management of their own labor process reveals how power operates at a deeper level: Exploitation extends not only outward toward others but also inward toward the self. For the self-entrepreneur who must bear risk individually, the object of primitive accumulation is, and can only be, the self.

The first dimension involves comprehensive management of the self. The charm influencers display on platform R, along with the narratives of self-optimization and self-realization embodied in their public personas, intertwine with the industry’s commercialized promise of upward mobility, making this profession appear highly desirable to many. Yet behind this glamorous surface lies a labor process that is complex, exhausting, unstable, and risky. The findings show that capital’s control and exploitation of labor have penetrated deeply into personal life and the self itself, occurring under the illusion of freedom—an exploitation of freedom itself. This control manifests in workers’ comprehensive management and development of the self, extracting content from their own lives to generate attention, data, and profit. What influencers manage and accomplish encompasses the body as the material carrier of the self, the relationships and emotional expressions built around the self, and the private life and trajectory of the self. One fashion influencer on platform R, pseudonymized as SJJ, who has nearly 100,000 followers, vividly illustrates this excessive self-development and all-encompassing self-management through her meticulous cultivation of body shape and aesthetic style, her sustained relational and emotional labor, and the instrumentalization and commodification of her private life.

The second dimension concerns the compulsory and strategic exposure of the self. The rapid growth of the digital influencer industry has produced a new market of algorithm specialists and social media consultants who provide training and guidance for newcomers. Due of the opacity of algorithms and the instability and unpredictability of data, influencers increasingly share their experiences and insights about account management with one another. Two forms of “knowledge” currently dominate the market. The first rests on what might be called “algorithmic imagination:” Since algorithms allocate traffic according to content tags, influencers at the initial stage of account growth tend to focus on a single niche or “track” to attract a specific audience before gradually expanding to other areas—a process of verticalization in content production. The second centers on the ideals of “authenticity” and “intimacy,” encouraging influencers to present themselves as complete and “real” individuals—a process of authenticity-driven production. Analysis of online posts indicates that most influencers combine these two knowledge systems in practice.

Another major mode of attracting audience attention lies in the performative narration of self-transformation, often through nostalgic disclosure of personal experiences. By showcasing the process of self-formation—through their own effort or with others’ help—influencers narrate the reshaping of their bodies, souls, thoughts, and behaviors, ultimately reaching a state of happiness, perfection, or wisdom. This narrative paradigm itself constitutes a crucial dimension of charisma performance.

The third dimension involves the continuous regulation and transformation of the self. The maintenance and growth of data are highly uncertain processes. On the one hand, the accumulation and circulation of data drive influencers to engage in constant self-optimization. Both interviews and online ethnographic observations suggest that the accumulation of data capital overlaps with the enrichment and diversification of the self. On the other hand, data are inherently unstable, and the primary strategy for managing this uncertainty is the continual production, testing, adjustment, and transformation of the self. Influencers must constantly assess their content according to data performance, extract lessons from success and failure, and further optimize their production practices.

Today, digital infrastructure forms the core of platform-driven labor organization and control, serving as the key means through which labor processes are quantified, supervised, and evaluated. Future research on digital platform labor should move beyond the framework centered primarily on control and resistance between algorithms and workers, situating digital labor within a broader network of sociocultural, technological, political, and economic relations, thereby enriching our understanding of labor–capital dynamics in the platform economy.

Shen Chao and Qin Yangguang are from the School of Philosophy and Social Development at Shandong University. This article has been edited and excerpted from Sociological Studies, Issue 4, 2025.