Many organizations that dominate our age heavily rely on complex data production processes, whereby data serve as the building blocks of organizational operations and the bases for developing novel goods or services.
3 Main Stages of the Social Life of Data
- The making of data
- The production of novel knowledge objects
- The systematic use of data in various contexts, including the packaging of data-based commodities exchanged in the market
1. Humans Make Data
Data are made, and digital data are no exception. Even when data already exist in other formats (e.g., analog pictures or movies), there is always more than one way of making them digital.
Yet most data used today by economic actors and organizations are data that did not exist in some other format and are native to the online or computational environments to which they belong. The conception, design and implementation of such data often entail a complex and distributed human and technical effort that tends to remain in the background or, worse, is taken as an accurate (and thus unproblematic) rendering of some events or facts “out there.”
Social data, the data that encode user engagement on social media platforms and other social online settings, are among the most conspicuous examples of digital native data that illustrate the institutional nature of the processes underlying their generation and use.
These social data, which are usually assumed to represent the opinions and whereabouts of users, are the by-products of user interactions that are designed mainly to accommodate the operations of social media platforms as business organizations, implemented under the constraints of technological infrastructures.
It is characteristic of digital data and their standardized formats to make data reuse and repurposing widely diffuse operations. Straightforward as they may initially seem, data reuse and repurposing are intricate.
On the one hand, several technical issues result from different formats, missing records, data inconsistencies, duplications, corrupted fields and the like. Rather than being trivial or limited, these are recurrent issues reinforced by the frequent updates and changes in software systems applications that are endemic to digital technologies.
As data travel across contexts and are reused, the predilections, assumptions and design choices on the basis of which they are initially produced become opaque and fade into the background. Reusing and repurposing data requires reimagining their roles and building the organizational capabilities to fuse such operations with established organizational and knowledge functions and practices.
In this regard, data reuse and repurposing should be better seen as increasingly widespread instances of data making.
2. Aggregated Data Become Data Objects
Data making and the practices with which it is associated are technology-driven, socially embedded processes that produce novel knowledge artifacts.
The modalities and different conditions that underpin the constant data manipulation, aggregation and processing give rise to a whole new breed of entities that weren’t there before, at least not in their current shape.
In their simplest form, these entities are only aggregations of multiple instances of the same data. In more complex forms, these objects are configured by putting together different data types under a given structure or shape. We call these entities “data objects.”
Data objects are technologically and structurally simpler than software objects and are brought into being by a structure or schema whereby data items are put together in a pattern or form. Such data-structuring schemas differ from software programs and the detailed instructions that the latter embody.
A typical example of a data object is a customer profile made of several attributes that are themselves clusters of data (i.e., transactions and login data) structured under a given format. Even the simplest arrangement of data requires some instructions as to how to identify, select and assign tokens to it.
Rather than capturing an underlying essence, [this] conception of data objects stems from the function that they fulfill in the process of knowledge making and use in which one or more organizations participate.
While technical entities, data objects remain semantic artifacts or cultural constructs, recurring arrangements of the events data encode, ordered according to certain logics and criteria that serve several aims. They are, in fact, the basic cognitive units, the elemental reality cuts of a complex infrastructure of knowing and acting, without which other more inclusive perceptions or knowledge management operations would be virtually impossible.
Credit scores, user profiles, ad impressions, click-through rates, viewability metrics, lookalike audiences and customer profiles are some conspicuous examples. They are all entities built by data aggregated under a schema or structure that makes the world legible and actionable in new ways, enabling new social interactions and new work practices within and across organizations.
3. Data Objects Become Data Commodities
Data objects work as coordinative devices that structure the operations of organizations and shape the relationships between economic actors and their environments.
There is, though, a distinct and growing class of data objects that enter the dealings of economic actors in the form of commodities. Advertising audiences, reputation scores, popularity indexes and credit records are some typical examples.
Unlike traditional production, distribution and consumption processes, a data life cycle is rarely planned and executed to end with a ready-to-market product. Data and data objects are never “finished” and data commodities present a stark contrast to traditional commodities in this regard.
Essentially a data commodity is a data object that acquires value because of specific market conventions. Unlike other commodities, however, even when data are stabilized into specific structures, data objects obtain only a temporarily bounded stability that is valid along specific dimensions (i.e., number and content of fields or metadata and functions that they execute in a system).
Data objects are constantly updated with new data while continuously executing themselves by performing specific functions.
Much like data, data objects are cognitive and communicative artifacts that are relational in nature; they are always linked to something else (e.g., a real event, data, other data objects, technologies or functions). These characteristics make data objects commodities of a particular kind and do not respond well to traditional market mechanisms or business models.
In general, commodification is implemented by repurposing data and data objects for a given commercial context (i.e., a market opportunity, a client need or a user behavior). In many settings and industries, the making of data commodities entails automatically recontextualizing and tailor-packaging data objects in real time.
The dynamics of data innovation and the role that these actors play in shaping the digital economy and society can be adequately framed only by unpacking the mechanics of data production and the journey that data undertake within and across organizations through the data life cycle.
Excerpted from Data Rules: Reinventing the Market Economy by Cristina Alaimo and Jannis Kallinikos. Reprinted with permission from The MIT Press. Copyright 2024.
