From the Data Economy to the Context Economy
Data Economy
In 2006, British mathematician Clive Humby coined the phrase, “Data is the new oil.” The statement captured a defining shift in the digital economy: data was no longer merely a by-product of computing. It had become an economic asset.
Over the following two decades, the Data Economy transformed the way businesses operated. The ability to collect, store, analyse, and distribute data became a source of competitive advantage and a major driver of innovation, productivity, and economic growth.
At the same time, software itself underwent a fundamental transformation. It moved from being a product that users purchased, installed, and maintained to a service continuously delivered through the cloud. Software-as-a-Service (SaaS) became the dominant model, replacing perpetual licences and on-premise installations with cloud-based, subscription-driven software.
The Cloud fundamentally changed the economics of computing. Instead of owning and managing the infrastructure required to run software, users increasingly delegated responsibility for compute, memory, storage, security, maintenance, and much of the underlying software stack to service providers. The trade-off was simple:
Users gave services access to, and a degree of control over, their data in exchange for convenience, functionality, and better experiences.
It worked remarkably well. But it also created a structural problem.
As more aspects of life moved online, an individual’s data became increasingly fragmented across services and vendors. Identity might reside in one service, communications in another, financial information somewhere else, work in a productivity platform, documents in a cloud drive, purchases across multiple marketplaces, and preferences and behavioural history scattered across dozens of applications.
Each service became increasingly effective at understanding its own piece of the user’s world. Very few could understand the whole.
A calendar knows where a user needs to be.
An email service knows what the user has been asked to do.
A banking application knows what the user has spent.
A productivity application knows what the user is working on.
A search engine knows what the user is trying to learn.
A commerce platform knows what the user buys.
Each system possesses a fragment of understanding. But these fragments rarely come together to form a coherent picture of the individual. This created a paradox:
Users had accumulated more information about themselves than ever before, while having less ability to see, connect, and understand that information as a whole.
The burden of context remained with the user.
Users became responsible for integrating information across disparate systems and, ultimately, deciding what to do with it. They had to navigate between applications, search for information, reconcile calendars with emails, compare documents, reconstruct past decisions, connect conversations to tasks, and determine what mattered and what should happen next.
Individual applications could generate insights from the data they controlled. But connecting those insights into a coherent understanding of the user’s broader circumstances — and translating that understanding into action — remained largely the user’s responsibility.
Users became both the integration layer and the decision layer.
They had to stitch together fragmented data, applications, and signals to construct a holistic view of their own lives, and then apply their own intelligence to determine the appropriate course of action.
This reveals the fundamental limitation (or rather an advantage) of the Data Economy.
The Data Economy shifted control of data from users to software. Platforms became the custodians of increasingly large and valuable portions of an individual’s digital life. But control of the broader context — and therefore much of the intelligence that could be derived from it — remained with the individual.
Software could capture, store, process, and analyse the data it was given. But it generally could not see the full picture. It did not know enough about the individual’s circumstances, priorities, history, relationships, intentions, and constraints to understand what all that data meant together.
The user remained the only entity with access to the most complete view of their context. And therefore, the user remained the final intelligence layer.
Context Economy
The emergence of Agentic AI is beginning to fundamentally reshape the relationship between users and software.
Until now, most software has operated as a collection of discrete tools. Users issue commands within individual applications, retrieve information from separate systems, interpret the results, and manually connect the pieces themselves.
Agentic AI changes this model.
Rather than merely responding to isolated instructions, AI agents can potentially reason across systems, connect previously fragmented information, interpret context and intent, and act on the user’s behalf.
An agent can reconcile a calendar with emails, connect a task to the documents and conversations that inform it, interpret spending in the context of plans and commitments, identify relationships between seemingly unrelated pieces of information, and surface what matters — all without requiring the user to manually assemble the pieces.
This represents a fundamental shift in the role of software.
Software is moving beyond a collection of tools that users operate toward an intelligence layer that understands the user’s world, context, and intent.
It promises to behave like a user’s “second brain”: continuously building an understanding of the user’s information, relationships, preferences, objectives, commitments, and history — and using that understanding to assist them or act autonomously on the user’s behalf.
The economic opportunity that follows is therefore different from the opportunity created by the Data Economy.
The Data Economy solved the data problem: How do we capture, store, process, analyse, and distribute data?
The Context Economy addresses a different problem: How do we capture, store, connect, reason over, and distribute the context surrounding that data — and ultimately translate context into action?
Get Sushanta Pradhan’s stories in your inbox
Join Medium for free to get updates from this writer.
The distinction matters.
Data tells what happened. Context helps understand why it matters. Intelligence determines what should happen next.
A new economic layer is therefore emerging on top of the Data Economy — one in which value is created not merely by possessing information, but by understanding the relationships between information, the circumstances in which it exists, the user’s intent, and the actions that should follow.
In the Data Economy, competitive advantage came from collecting and controlling valuable data. In the Context Economy, competitive advantage may increasingly come from constructing and maintaining the most useful understanding of that data.
The Data Economy made data useful. The Context Economy shall make data purposeful.
The Questions That Follow
If context becomes the foundation upon which intelligence operates, its emergence as an economic asset raises questions that may prove as consequential as those created by data itself.
1. Who will own and control context?
In the Data Economy, users gradually entrusted platforms with enormous portions of their digital lives. Email providers accumulated communications. Social networks accumulated relationships. Search engines accumulated intent. Commerce platforms accumulated purchasing behaviour. Productivity platforms accumulated work.
Will the same pattern repeat with context?
Will users entrust a new generation of AI platforms with an even more valuable asset: a persistent understanding of their lives?
Or will Agentic AI invert the existing model, allowing individuals to retain ownership of their context while selectively granting agents and services permission to access and act upon it?
The answer could determine whether the Context Economy becomes another platform-controlled economy — or a more user-controlled one.
2. How will context be structured and governed?
Data is already fragmented across hundreds of applications, databases, services, and vendors. Context could inherit the same fragmentation.
Alternatively, a new category of systems may emerge whose primary purpose is to maintain a persistent, holistic model of an individual: their information, relationships, history, preferences, goals, commitments, constraints, and intentions.
If such systems become the foundation upon which agents operate, questions of privacy, portability, interoperability, permission, identity, and platform power become significantly more important.
The system that controls context may ultimately influence which intelligence can be built upon it.
3. How will context-driven software be delivered and monetised?
The Data Economy evolved alongside the Cloud and SaaS. Centralised infrastructure, subscription pricing, and vendor-controlled applications became the dominant architecture for delivering software.
But context introduces different requirements.
A persistent model of an individual’s life may contain far more sensitive information than any single application possesses today. It may need to remain continuously available across applications, agents, devices, and services. Users may demand stronger guarantees of ownership, portability, privacy, and control.
Will the Context Economy therefore continue the cloud-based, subscription-driven SaaS model?
Or will these requirements drive the resurgence of locally deployed software, user-owned infrastructure, new licensing models, decentralised architectures, or entirely new forms of software distribution?
The architecture of the Context Economy is still a WIP.
Conclusion
The shift from the Data Economy to the Context Economy is ultimately a shift in where intelligence resides.
The Data Economy gave software access to enormous quantities of information. But because that information remained fragmented across systems, the user still had to provide the missing layer: the intelligence.
The user had to remember history, understand relationships, recognise constraints, reconcile competing priorities, infer what mattered, and decide what should happen next.
Agentic AI creates the possibility of transferring part of that burden from the individual to software. And that changes what becomes economically valuable.
Data alone may no longer be the defining asset of the next generation of software. The context that gives data meaning — and enables intelligence to act on it — could become equally, if not more, valuable.
The Data Economy was built around the accumulation and utilisation of information. The Context Economy may be built around the accumulation and utilisation of understanding.
If that happens, the defining exchange of the digital economy may change as well.
In the Data Economy, users exchanged data for convenience. In the Context Economy, what will users exchange for intelligence?
And perhaps more importantly: how much of themselves will users need to reveal before software understands them well enough to act on their behalf?
These questions sit at the heart of the emerging Context Economy.
In the coming articles, I shall explore this economy in greater depth -
its architecture, ownership models, distribution mechanisms, competitive
dynamics, business models, and implications for the future of software.
