Datafication

Working Definition

Datafication is the transformation of aspects of the world and human life that previously had no quantitative record—bodily posture, casual conversation, mood, friendship ties, physical location, or driving style—into standardized, machine-readable data formats suitable for aggregation, indexing, and algorithmic prediction. Crucially, it is not synonymous with digitization—the conversion of analog media (such as a printed book or vinyl record) into binary zeros and ones: a digitized book is still a text to be read by a human, whereas a datafied reading experience is the real-time telemetry emitted by an e-reader recording which passages a reader highlighted and on which page they stopped. While popular commentary uses the word loosely to mean “everything is measured nowadays,” analytically it denotes a three-part socio-technical operation requiring a sensor or interface, a standardized classification schema, and an institutional actor extracting value from the database. Serious analysis begins by asking who designs the schema and who captures that value.

Genesis and Global Lineage

The term in its contemporary sense was introduced and popularized in 2013 by Viktor Mayer-Schönberger and Economist editor Kenneth Cukier in their bestseller Big Data: A Revolution That Will Transform How We Live, Work, and Think, which devoted its fifth chapter explicitly to “Datafication.” Drawing on examples from Matthew Fontaine Maury's nineteenth-century standardization of ship captains' logbooks to Japanese car seats that identify drivers by posture, Mayer-Schönberger and Cukier wrote in an affirmative register: datafication was presented as a cognitive leap extending early modern scientific measurement and state statistics (Statistik) to everyday life. Within a year, however, critical media scholars turned the concept inside out. In 2014, José van Dijck argued that the celebration of datafication rested on an unexamined ideology she termed dataism—a naive faith that digital traces are neutral footprints of human behavior rather than artifacts shaped by platform algorithms and commercial imperatives. By 2019, Nick Couldry and Ulises Mejias had reframed datafication as the core mechanism of data colonialism (the appropriation of social life as a factor of production), while Jathan Sadowski analyzed data as a distinct form of capital accumulation. Today the concept circulates across management literature and critical platform studies, anchored in the history of quantification pioneered by Theodore Porter and Alain Desrosières.

Central European Lineage

One of the two authors of the concept's canonical 2013 formulation was formed intellectually in the Austrian legal tradition of information and data-protection law. Born in Zell am See, Austria, in 1966, Viktor Mayer-Schönberger studied law at the University of Salzburg (earning his Mag. iur. in 1988 and Dr. iur. in 1991), founded the antivirus and data-security software firm Ikarus Software in Salzburg in 1986, and completed his Habilitation in information law, legal theory, and legal informatics at the University of Graz in 2001, before going on to faculty appointments at the Harvard Kennedy School, the National University of Singapore, and the Oxford Internet Institute. That Central European legal formation is more than a biographical footnote: Mayer-Schönberger came to the phenomenon of datafication not from Silicon Valley ad-tech engineering, but from decades of European scholarship on the legal regulation of information flows and institutional memory—most notably his influential 2009 book Delete: The Virtue of Forgetting in the Digital Age, which helped inspire the European Union's “right to be forgotten.” That background explains why Big Data frames datafication from the outset not merely as a database-engineering trick, but as a profound shift in how legal and social institutions remember, categorize, and judge human beings.

Why Now

Three interlocking developments between 2018 and 2026 turned datafication from a descriptive buzzword into a battleground of law and geopolitics. First, the European Union constructed an ambitious regulatory architecture around datafied life: beginning with the General Data Protection Regulation (GDPR, 2018) and expanding through the Data Governance Act (2022), the Digital Services Act and Digital Markets Act (2022–2024), the Data Act (2023, governing industrial and Internet-of-Things telemetry), and the AI Act (2024), European law turned the question “what may be datafied, by whom, and under what access rights?” into hard statutory rules. Second, privacy shocks—from the 2018 Cambridge Analytica revelations to post-2022 controversies over period-tracking apps, connected-car telematics sold to insurers, and workplace bossware—made clear that seemingly trivial lifestyle data can be recombined into intimate behavioral profiles. Third, the generative AI boom after late 2022 revealed the industrial prize of three decades of web datafication: large language and multimodal models devour the datafied archive of human text, speech, art, and code as training fuel, turning copyright lawsuits and licensing deals between publishers and AI laboratories (2023–2026) into a struggle over the retroactive price of datafying human culture.

Explanatory Power

Datafication separates three steps routinely conflated under “Big Data”: reductive measurement, classification standardization, and value extraction. First, it explains the business logic of seemingly unprofitable digital products: a free fitness tracker, smart thermostat, or navigation app is a sensor array converting unrecorded routines into proprietary datasets. Second, it illuminates the power asymmetry in platform work: a driver or warehouse picker sees only isolated dispatch prompts, whereas the platform datafies their braking patterns, route deviations, and acceptance thresholds to train the pricing and routing models that govern their wages. Third, it explains why machine-learning systems reproduce institutional bias: because datafication captures only what an institution previously chose to count (arrests rather than crime, billing codes rather than health, clicks rather than satisfaction), every dataset is an ossified archive of past administrative priorities masquerading as neutral nature.

Axes of Debate

Underappreciated Angle

Critiques of datafication almost always ask what is taken from the people being measured (privacy, autonomy, unpaid labor), while rarely examining the institutional blindness inflicted on the organization doing the measuring. Because datafication requires translating qualitative human realities into discrete, standardized tokens, whatever cannot be cheaply tokenized ceases to exist inside the organization's decision loop: the nurse's unrecorded conversation that prevents a patient's panic, the teacher's cultivation of curiosity that does not register on a standardized test, the engineer's tacit sense that a system feels brittle before any sensor trips. As organizations shift from human judgment to dashboard governance, they starve the unmeasured foundations that keep them alive—a pathology that historian Theodore Porter (Trust in Numbers, 1995) diagnosed long before the internet: rigid quantification is rarely the sign of a confident, high-trust institution; it is a defensive substitute adopted when authority no longer trusts its own professionals or is no longer trusted by the public.