Data is ‘crude’, digital refineries create value

For more than a decade, corporate strategy has been influenced by the proposition that “data is the new oil”. Businesses celebrate the scale of their information assets, while investors and executives increasingly equate ownership of information with competitive advantage. Data has become an essential economic resource. Yet it risks encouraging a misunderstanding of where value is ultimately created.

The more useful analogy is therefore not between data and petroleum, but between the digital and petroleum value chains. Crude oil acquires economic significance through refinement. Data follows an analogous journey. Its greatest value emerges through the institutional capability that systematically transforms raw information into intelligence worthy of commercial, regulatory, and societal trust.

The first phase of the digital economy rewarded accumulation. Data was scarce, computing expensive, and analytical capability limited. Every customer interaction, financial transaction, supply chain event, and digital footprint represented another potentially valuable resource.

That economic logic is changing. Smartphones, payment systems, industrial sensors, satellites, and connected devices now generate information continuously. Computing power has become progressively accessible. AI models are proliferating rapidly, while many algorithmic capabilities that once differentiated leading technology companies are steadily becoming available across markets. As industries mature, scarcity migrates.

That has profound consequences for corporate strategy. An enterprise may possess extraordinary volumes of information while lacking consistent definitions, coherent architecture, or sufficient confidence in the decisions emerging from its systems. Scale alone offers limited assurance of quality. Indeed, decades of acquisitions, incompatible platforms, changing regulations, and evolving operating models have often left organisations with information estates.

AI is steadily reducing the cost of information refinement. It is not reducing the value of institutional judgement. As the economics of processing improves, the economics of governance becomes correspondingly more important. It is increasingly capable of classifying documents, reconciling records, enriching data sets, identifying anomalies, and continuously improving information quality at a scale unimaginable only a few years ago.

As algorithms become progressively commoditised, competitive advantage migrates towards the environment within which those algorithms operate. AI can automate key aspects of refinement. It cannot determine which information should be regarded as authoritative, set up governance standards, reconcile competing legal obligations, or assume accountability for decisions carrying commercial, regulatory, or fiduciary consequences.

Petroleum refineries derive their economic significance from engineering discipline, operating standards, process integrity, and continuous quality assurance. Digital refineries demand comparable institutional attributes. Information must be governed throughout its lifecycle so that provenance remains traceable, definitions remain consistent, systems remain interoperable, and outputs remain reliable for consequential decisions.

The economics of refinement deserves greater prominence in boardroom deliberations. Cleaning information is expensive. Integrating fragmented systems requires sustained investment. Establishing common taxonomies across business units consumes management attention. Maintaining high-quality data sets for machine learning demands continuous effort because the underlying information itself never remains static. These expenditures often appear less visible than investments in AI applications, although they contribute more directly to the quality of decision-making, operational resilience, and long-term productivity.

Boards have traditionally regarded information quality as an operational concern residing within technology functions — this could be unsustainable. Capital allocation, regulatory compliance, operational resilience, cybersecurity, customer trust, and strategic decision-making are shaped by the quality of information entering organisational systems. As AI assumes a larger role in consequential decisions, information integrity becomes inseparable from fiduciary oversight.

The same reasoning extends to public policy. Data protection legislation has concentrated, understandably, on collection, consent, storage, and transfer. Emerging AI regulatory ideas are focusing on model behaviour, transparency, bias, and accountability. The intervening stage receives comparatively less attention — the institutional processes through which information is combined, interpreted, enriched, and transformed before intelligent systems act upon it. The long-term effectiveness of digital regulation may increasingly depend on recognising that governance of the refinery is as important as governance of the oilfield/finished product.

Investors may also need to reconsider where enduring competitive advantage will reside. Markets have historically attached significant premiums to ownership during the early stages of technological revolutions before recognising that durable returns accrue to capabilities. The digital economy appears to be approaching a comparable inflection point.

Companies capable of repeatedly transforming fragmented, incomplete, and continuously changing information into trusted intelligence occupy a fundamentally different economic position. Their competitive advantage rests not upon ownership alone but upon institutional capability that competitors cannot easily replicate. Trust, in this context, is the economic outcome of consistently producing intelligence that customers, regulators, investors, and boards are prepared to rely upon.

Disclaimer: The views expressed are the author’s own and do not reflect the official policy or position of Financial Express.