Data Has Currency: The Case for the Internal Data Brokerage

There's a conversation I keep having with boards and technology leaders that follows a predictable arc. We talk about data strategy. We talk about AI readiness. We talk about digital transformation. And somewhere in the middle of it, I ask a simple question: what is your data actually worth?

The silence that follows is not from lack of intelligence. It is from a structural blind spot that almost every organisation shares. Data is treated as an operational resource - a byproduct of doing business, a cost of running systems, an input to analytics. It is not treated as what it actually is: an asset with measurable, tradable, appreciating value.

That needs to change. And I think the mechanism that changes it is the internal data brokerage.

1. The Problem with How We Think About Data Today

Walk into any enterprise and you will find data siloed in division-specific systems, hoarded by teams who gathered it, duplicated across platforms that were never designed to talk to each other, and almost never valued in any formal sense. The sales team has customer behaviour data that would transform the product team's roadmap. The operations team has asset performance data that would reshape how finance models risk. The logistics function has movement data that would sharpen every demand forecast the business produces.

None of this data is on the balance sheet. None of it has a price. None of it is traded.

That is an extraordinary inefficiency hiding in plain sight and it is made worse every single time an AI model is trained on incomplete data, a strategy is built on a partial picture, or a decision is made without the context that already exists somewhere inside the same organisation.

The accounting treatment of data as an expense - the cost of storage, the cost of pipelines, the cost of governance - rather than as an asset is one of the most consequential misframings in modern enterprise. It determines how boards think about data investment, how CFOs model data-related spending, and ultimately how much organisational energy flows toward building data capability versus managing data cost.

We would never treat physical infrastructure this way. A building, a machine, a vehicle - all of these sit on the balance sheet and accumulate value or depreciate accordingly. We understand that owning them, maintaining them, and leveraging them intelligently creates competitive advantage. Data is no different. We have simply not built the frameworks to see it that way.

2. What a Data Brokerage Actually Is

The concept I have been developing and advocating for is not complicated in principle, even if the implementation requires discipline. An internal data brokerage is a governed platform through which data - structured data, operational data, behavioural data, sensor data - is assigned value, made discoverable, and traded between business units in exchange for a unit of internal currency.

Think of it as an internal marketplace, with the organisation as both the regulator and the market participant. The sales division does not simply share its customer data with marketing because someone requested it. It trades it. The operations team does not give away asset telemetry for free. It earns currency when that data is accessed and applied. The brokerage creates a mechanism through which the value of data becomes visible, traceable, and - critically - incentivised.

The brokerage operates on a combination of models: datasets carry a valuation score reflecting quality, recency, and demand. Access generates internal credits. Those credits can be used to acquire other data, fund data enrichment, or flow into divisional budgets as a recognised contribution. Supply and demand within the platform shapes value over time - just as it does in any functioning market.

The most important word in that description is governed. A data brokerage without governance is just a more complicated data lake. What distinguishes a brokerage is that it enforces rules about data quality, provenance, access rights, and valuation with the same rigour that a financial market enforces trading rules. The governance layer is not overhead - it is the entire point.

3. What Gives Data Its Value?

This is the question I am asked most often when I present this concept, and it is exactly the right question to be asking. Data value is not fixed - it is contextual, relational, and dynamic. But it is absolutely determinable. Several dimensions inform it:

Uniqueness. Data that only your organisation can generate - from your customers, your assets, your operations - has an inherently higher floor value than data that can be acquired from a third party. Proprietary signal is worth more than commodity input.

Quality and completeness. A well-governed dataset with documented lineage, validated accuracy, and consistent structure is worth measurably more than the same data in a degraded or incomplete state. Quality is not just a technical attribute - it is an economic one.

Recency and velocity. Real-time or near-real-time data commands a premium in most use cases. A demand signal from this morning is worth more than the same signal from last quarter. The brokerage needs to reflect this in its valuation mechanics.

Demand within the platform. If multiple business units are seeking access to the same dataset, that is a market signal. Price discovery through internal demand is one of the most powerful mechanisms for revealing what data is actually strategic versus what data simply accumulates.

Downstream impact. The most sophisticated valuation models track what happens after data is accessed. Data that demonstrably improves a decision, improves a model, or generates a measurable business outcome gains value through proven utility.

4. Data on the Balance Sheet

If data has value - and I am arguing it does, measurably and materially - then the question of how it is accounted for is not a philosophical one. It is a governance and financial reporting question that boards should be pushing their CFOs to address right now.

Current accounting standards do not make this easy. Internally generated intangible assets are treated inconsistently across most frameworks, and the specific treatment of data as an asset class is still evolving in practice, if not yet in regulation.^[1] But the direction is clear. Regulators, analysts, and investors are increasingly looking for data asset disclosure. The organisations that develop the internal capability to value their data now - through a brokerage model or otherwise - will be substantially better positioned when that disclosure becomes expected or required.

More immediately: a data brokerage gives the finance function something it has never had before. A mechanism through which the value generated by data across the organisation can be tracked, attributed, and reported. Not as a vague statement that "data is a strategic asset," but as a real ledger entry. The legal team created dataset X. Operations used it. The downstream decision improved margin by Y. That is a value chain that can be modelled, reported, and used to make investment decisions about where to focus data collection and enrichment effort.

The first organisations to put data on the balance sheet will not just be more financially literate - they will be more strategically capable. You cannot manage what you cannot measure. You cannot invest in what has no declared value.

5. The Internal Market First - Then the External One

My position on sequencing is deliberate. The internal data brokerage is the foundation. External data exchange is the future state - and it is a compelling one - but organisations that attempt to commercialise or share data externally before they have mastered internal data governance and valuation are building on sand.

The internal market does something critical that external engagement cannot: it forces the organisation to become serious about data quality, lineage, provenance, and access control under conditions where the stakes are controlled. Getting data governance wrong internally costs you efficiency. Getting it wrong externally costs you trust, reputation, and potentially compliance.

But the external horizon is real and worth naming. An organisation that has built a mature internal data brokerage - with clear valuation mechanics, well-governed datasets, and a culture that treats data as a tradable asset - is positioned for something significant. It can participate in industry data exchanges. It can negotiate data partnerships with suppliers and customers from a position of genuine understanding of what its data is worth. It can explore commercialisation pathways that are currently invisible to most enterprises precisely because they have never had to put a number on what they hold.^[2]

Most organisations will operate across multiple stages simultaneously. The framework is not a linear migration path - it is a map for building maturity deliberately, one stage at a time, without losing operational momentum.

6. Why This Is an AI Readiness Question

I want to make a connection that I think is underappreciated in most AI strategy conversations. The quality and accessibility of an organisation's internal data is not just an operational consideration - it is the single largest determinant of AI performance. Models are only as good as the data they reason over.^[3] And in most organisations, the data that AI systems most need is the proprietary, contextual, operationally specific data that lives in silos, is poorly documented, and has never been formally valued.

An internal data brokerage solves this problem directly. It forces data quality up because quality now has a price signal attached to it. It makes data discoverable because discoverability is a prerequisite for exchange. It creates accountability for data provenance because the brokerage cannot function without it. Every one of those outcomes is also an AI readiness outcome.

Organisations that are investing heavily in AI capability but have not addressed their internal data governance and valuation are essentially building a high-performance engine and running it on contaminated fuel. The brokerage cleans the fuel. It structures it. It ensures the right data reaches the right model at the right time and that the organisation understands the value of what it is contributing to that process.

When Physical AI systems - sensors, autonomous infrastructure, edge intelligence - are generating operational data at scale, the organisation that has a mature data brokerage will be able to capture, value, and leverage that data immediately. Those without one will simply accumulate more ungoverned signal and wonder why their AI investments are underperforming. If you haven't explored what Physical AI means in practice, I'd encourage you to read my earlier piece: From Code to Concrete: The Rise of Physical AI and the Power of Convergence. The data implications of that convergence are significant.

7. The Cultural Shift This Requires

I would be doing this concept a disservice if I presented it purely as a technical or financial architecture problem. It is not. The hardest part of building a data brokerage is not the platform. It is the culture change.

Data hoarding is deeply embedded in most organisations. Teams that gathered data view it as their data. Division leaders protect information asymmetry because it gives them political leverage.^[4] The idea of trading data - of making it accessible to other parts of the business in exchange for internal currency - runs directly against those incentive structures.

Reframing this requires leadership from the top. The board and the C-suite need to articulate clearly that hoarded data is a liability, not a strategic asset. Data that sits in a silo depreciates - through recency decay, through inaccessibility, through the decisions that were never improved because the right information never reached the right team. The brokerage needs to make the cost of hoarding visible and the benefit of sharing real and measurable.

Done well, it creates something genuinely transformative: an organisation where the incentive structure around data is aligned with the interest of the whole business rather than the individual division. Where contributing quality data to the platform is recognised, rewarded, and reported. Where the question "what is our data worth?" has a real answer.

Conclusion: The Balance Sheet Is Incomplete

Every organisation reading this article is sitting on data that is not on its balance sheet, is not being traded internally with any rigour, and is not being valued in any meaningful sense. That is not a technology problem. It is a strategic one and it is one of the most significant untapped opportunities in enterprise today.

The data brokerage is the mechanism through which that changes. It is not a product to be purchased or a platform to be deployed in isolation. It is a strategic posture - a decision to treat data as the fiscal asset it already is, to govern it accordingly, and to build the internal markets through which its value can be discovered, traded, and compounded.

The organisations that build this capability now will not just be better at data governance. They will be better at strategy, better at AI, better at investment decisions, and better positioned for a future in which data exchange extends beyond the organisation's walls.

The balance sheet is incomplete. It is time to finish it.