The Next Frontier in Decision Intelligence

In analytics and operations research, we are trained to think in systems. We model flows, optimize decisions, and quantify trade-offs. We ask: What are the inputs? What are the constraints? What is the objective function?

But there is one element in today’s decision systems that we rarely question – yet it is the most critical one. 

Data. 

Where it comes from. Who owns it. Who controls it. And increasingly, who profits from it. 

We have built an entire decision economy on top of data without ever truly defining its ownership model. As a result, we are optimizing systems built 
on an unstable foundation. 

That is the problem data independence seeks to address. 

From Data Abundance to Data Imbalance

During the past two decades, organizations have become exceptionally good at collecting, storing, and exploiting data. The rise of machine learning and, more recently, generative AI has accelerated this trend, making data not just useful, but essential.

In many industries, data has become the primary driver of competitive advantage. It fuels forecasting models, personalization engines, pricing systems, and optimization frameworks. It powers everything from supply chains to financial services to healthcare decisions. 

But while organizations have mastered data utilization, the underlying structure remains fundamentally asymmetric. While individuals generate the data, organizations capture it, 
and systems optimize around it. As such, the individual – the original source of the data – is largely absent from the value equation. From an operations perspective, this is a structural inefficiency; from a societal perspective, it is a growing risk. 

The Missing Variable in Decision Systems

Traditional OR/MS frameworks assume that inputs are available, measurable, and appropriately governed. But what happens when the input itself 
is contested? 

Personal data today is not simply an input; it is a reflection of behavior, identity, and decision-making patterns. It is what I describe as “digital DNA”: a continuously evolving representation of how individuals interact with systems, markets, and environments.
 
When that data is aggregated across sources (e.g., transactions, devices, and platforms), it forms a 360-degree view that can be used to predict, influence, and optimize outcomes. This creates a new class of decision systems that are not just reactive, but anticipatory. Yet these systems often operate without transparent consent, clear ownership, or defined value exchange. 

From a decision science standpoint, this introduces hidden constraints, including: 

• Misaligned incentives between data producers and data users
• Lack of transparency in model inputs and outputs
• Increasing regulatory uncertainty across jurisdictions
• Growing erosion of trust in data-driven systems 

These are not theoretical concerns. They directly impact the quality, stability, and scalability of decision systems.

Introducing Data Independence

Data independence is a framework designed to rebalance this system. At its core are three principles:

• Consent. Individuals explicitly authorize how their data is used.
• Control. Individuals maintain ongoing authority over access and usage.
• Currency. Data is recognized as an economic asset, with value participation. 

Together, these principles form what I refer to as the “C3 model.” In addition, the framework incorporates broader governance principles – transparency, accountability, protection, equality, and respect (TAPER) – to ensure that systems are not only efficient, but also responsible. This is not about restricting data. It is about structuring it correctly.

Implications for Analytics and OR/MS

For the OR/MS and analytics community, this shift is significant. Historically, we have focused on improving models given available data. The next phase requires us to rethink how data 
enters the system in the first place. This has several implications:

• Higher-quality data inputs. When individuals participate knowingly in data exchange, data quality improves. Noise, bias, and incomplete signals are reduced, leading to more robust models.
• Transparent decision systems. With clearer data lineage, models become more explainable and auditable, which is critical for regulatory compliance and stakeholder trust.
• New objective functions. Optimization models may incorporate new variables, including fairness, consent constraints, and value-sharing mechanisms, expanding beyond traditional cost/revenue trade-offs.
• Trust as a performance driver. Trust is no longer an external factor; it becomes an internal component of system performance. Systems that violate trust degrade over time; systems that reinforce it improve engagement and data flow. 

From Extraction to Participation

The current data economy is largely extractive. Organizations collect data and optimize around it. Data independence introduces a participatory model in which individuals become active contributors to the system. This changes the dynamics of data-driven decision-making so that:

• Data becomes negotiated, not assumed.
• Access becomes conditional, not automatic.
• Value becomes shared, not concentrated.  

From a systems perspective, this is a more stable equilibrium.

Decision Intelligence in the Age of AI

As AI systems become more autonomous and more deeply embedded in decision processes, the importance of data governance will only increase. 
We are moving toward a world in which AI systems learn continuously from user-generated data, decisions are made in real time at scale, and models influence behavior rather than just predicting it.

In such a world, the question is no longer just how we optimize decisions, but how we define the inputs that drive them. If we get the data model wrong, no amount of optimization will fix the outcome. If we get it right, we unlock a new class of decision systems, ones that are not only efficient, but also sustainable.

Rethinking the Foundation

OR/MS has always been about making better decisions. Data independence extends that mission. It asks us to look beyond algorithms and models and examine the foundation on which they are built. It challenges us to treat data not only as a resource, but also as a governed asset with defined ownership and value. 

In doing so, it opens the door to a new generation of decision intelligence, one where optimization, AI, and human agency are aligned. Because ultimately, the quality of our decisions will depend on the integrity of our data and the systems we build around it.