This paper examines how platform delivery workers in Mexico experience, interpret, and contest algorithmic management in the course of their everyday labor. Unlike approaches that treat algorithmic control as a predominantly technical phenomenon, this study argues that its operation is mediated by structural inequality, pre-existing labor informality, and urban mobility—conditions that reconfigure both the modalities of control and the available margins of worker agency. The analytical framework reconceives algorithmic control not as a monolithic technological device, but as a labor management apparatus produced through opacity, incentives, ratings, sanctions, bonuses, territorial distribution, waiting times, and shifting interface signals.

The central argument holds that platform algorithms do not govern the labor process solely through order assignment and remuneration calculation, but also through a gamified environment that compels workers to decipher variable rules, anticipate algorithmic preferences, and continuously calibrate their labor strategies—route optimization, schedule management, and risk exposure—on the basis of partial, often ambiguous cues generated by the application and interpreted through collective sense-making. Far from endorsing technological determinism, the analysis recognizes that this environment also produces conditions for situated agency. Whether such agency constitutes resistance in an analytically robust sense, or rather functional adaptation absorbable by the system, remains an open empirical question.

Empirically, the study draws on 25 semi-structured interviews selected from the broader project Conceptions of Justice and Transformations of Labor Citizenship: The Case of Digital Delivery Workers in Mexico, funded by the Secretaría de Ciencia, Humanidades, Tecnología e Innovación (Mexico). These interviews were drawn from a corpus of 85 and selected for their analytical density relative to the categories generated through open coding, prioritizing cases that illuminate variation in workers' forms of agency and algorithmic subjection. The analysis proceeds inductively through open coding, identifying workers' daily evaluations of control, remuneration, sanctions, bonuses, and labor justice.

The findings demonstrate that delivery workers do not passively submit to algorithmic management; on the contrary, they develop practical repertoires oriented toward “playing” or “outwitting” the algorithm: simultaneous use of multiple applications, selective acceptance of orders, strategic disconnection, territorial specialization, active reading of demand patterns, and collective circulation of operational knowledge. These practices neither dismantle platform control nor constitute collective contestation in any strict organizational sense; yet neither are they reducible to mere functional adaptation. Rather, they disclose a terrain of everyday negotiation in which algorithmic discipline encounters practical, partial, and variable limits that workers recognize and exploit.

Taken together, these findings invite a critical reassessment of concepts such as resistance and discipline when transposed from Northern industrial contexts to labor environments defined by structural informality, high turnover, and the systematic absence of social protection. They suggest, moreover, that platform capitalism does not operate as a uniform global formation but assumes historically and institutionally specific forms according to the urban and regulatory conditions in which it is embedded.