From Information to Enterprise

Executive Summary

European enterprises are not short of data. They are short of decisions. Industry estimates commonly place unstructured content — documents, emails, sensor feeds, media — at somewhere between 80% and 90% of enterprise data volume as of 2025, up from a lower share earlier in the decade, though exact figures vary materially by definition, industry, and measurement method (see Methodology and Limitations). Yet volume has not translated into use: only around a third of enterprise data is judged high quality, just 15% of organisations report mature data governance despite 85% having a formal governance framework on paper, and Gartner expects 60% of AI projects unsupported by AI-ready data to be abandoned through 2026.

This paper argues that the gap between information and enterprise value is not a technology gap; it is a missing layer. Raw information becomes useful only once it passes through governance, structuring, and context — becoming intelligence — and intelligence becomes valuable only once it is embedded in live operational workflows, where it becomes enterprise capability. What survives that process, reused and compounding over time, becomes an asset in its own right.

This four-stage progression — Information, Intelligence, Enterprise, Assets — is the structural logic behind BLACK&'s own positioning framework. Organisations with mature data governance report 24.1% higher revenue improvement and 25.4% greater cost savings from AI initiatives than those without, according to IDC research — direct evidence that the intermediate governance and structuring layer, not raw data volume, is what converts information into measurable enterprise value.

The European policy environment is moving in the same direction. The EU's data economy grew from an estimated €301 billion in 2018 to €829 billion by 2025, and the European Commission's Data Union Strategy, published in November 2025, is explicitly designed to convert Europe's data abundance into AI-usable, high-quality datasets through Common European Data Spaces, expanded high-value datasets, and simplified data-sharing rules under the Data Act. For BLACK&'s institutional and government audience, the practical implication is that the binding constraint on enterprise AI value is rarely data access — it is the governance, structuring, and operational integration layer in between.

1. Methodology

This paper synthesises data-management and data-economy research published between 2020 and 2026, drawing on: (i) unstructured-data volume and growth statistics from IDC and industry surveys, cross-checked across multiple independent publishers where possible; (ii) data governance and quality maturity surveys from DATAVERSITY's Trends in Data Management research and IDC; and (iii) European Commission data-economy valuations and policy documentation, principally the 2020 European Strategy for Data and the November 2025 Data Union Strategy, COM(2025) 835. A number of the unstructured-data volume statistics in Section 3.1 originate from data-management vendors and statistics-aggregation sites rather than peer-reviewed or governmental sources; these are flagged explicitly in Limitations and should be treated as commonly cited industry ranges, not as precisely measured figures.

Section 3.4 onward presents the Information-Intelligence-Enterprise-Assets progression as BLACK&'s own structural framework for interpreting this evidence; this portion is normative and proprietary rather than a synthesis of third-party findings, and is labelled accordingly.

Scope

The paper addresses enterprise and EU-level data volume, quality, and governance trends, and the policy response through the Data Union Strategy and related instruments. It does not provide a technical evaluation of specific data platforms, vendors, or governance tooling, nor does it include primary data-maturity assessments of BLACK& client organisations.

2. Definitions

• Information — Raw, unstructured or loosely structured data as generated by business systems, communications, and sensors, prior to governance, structuring, or contextualisation.
• Intelligence — Information that has been governed, structured, and contextualised such that it is decision-ready: attributable, current, and interpretable without further manual processing.
• Enterprise (as a stage) — Intelligence embedded directly into live operational workflows and decisions, as distinct from intelligence held in a static report or dashboard that requires a human to manually apply it.
• Assets — Enterprise capability that compounds and is reusable across time and use cases, e.g., a trained model, a decision workflow, or a structured dataset with ongoing value beyond its original use.
• Data Union Strategy — The European Commission's November 2025 policy communication, COM(2025) 835, aimed at increasing the availability of high-quality data for AI and completing the EU single market for data.
• Common European Data Spaces — Sector-specific data-sharing infrastructures (health, mobility, energy, media, and others) established under EU data policy to enable trusted, large-scale data reuse for AI applications.
• Data-ready / AI-ready data — Data that meets the quality, structure, and governance thresholds required for reliable use in an AI system, as distinct from data that merely exists in digital form.

3. Analysis

3.1 The Data Abundance Paradox

Enterprise data volumes have grown explosively, and the growth is concentrated almost entirely in unstructured formats. Unstructured content — documents, emails, images, sensor and IoT streams — grew from an estimated 64% of total enterprise data in 2020 to roughly 90% by 2025, and by some industry estimates now accounts for 80-90% of all enterprise information, including in data-intensive sectors such as financial services, where it comprises 85-90% of records including trade documentation. Forty percent of large enterprises now store more than 10 petabytes of data, per Komprise's 2026 State of Unstructured Data Management survey.

This abundance is not, by itself, a source of value. Traditional analytics and business-intelligence tooling was built for structured, tabular data; unstructured content requires fundamentally different extraction and interpretation methods before it can inform a decision. The practical consequence is that the fastest-growing share of enterprise information is also the share least accessible to existing decision-making infrastructure.

3.2 The Quality Bottleneck: Why More Data Does Not Mean More Decisions

Governance intent and governance reality diverge sharply. Eighty-five percent of organisations report having a formal data-governance framework, and 92% of business executives agree that governance is critical to digital transformation — yet only 15% report actually achieving mature data governance, and independent estimates put the share of enterprise data meeting basic quality standards at roughly a third, with some benchmarks placing it far lower. DATAVERSITY's 2025-26 survey found that 75% of leaders do not trust their own data for decision-making, and 61% still rank data quality as their top operational challenge.

The financial consequence is direct: poor data quality costs organisations an average of $12.9 million annually, per industry benchmarking, and Gartner expects 60% of AI projects unsupported by AI-ready data to be abandoned through 2026 — the same data-readiness failure mode identified as the dominant cause of enterprise AI project failure in BLACK& Research's companion paper, The Intelligence Company Builder. Conversely, organisations that do achieve governance maturity see a measurable return: IDC research finds mature data-governance programmes deliver 24.1% higher revenue improvement and 25.4% greater cost savings from AI initiatives than programmes without that maturity.

3.3 The EU Policy Response: The Data Union Strategy and the Single Market for Data

The European Commission has treated this gap as a strategic priority since its original 2020 European Strategy for Data, which valued the EU data economy at €301 billion and set out to build it toward a projected €829 billion by 2025 — a 175% increase — through the creation of a single market for data. The Data Act, in force since January 2024 and applicable from September 2025, and the Data Governance Act built the initial legal scaffolding; the Data Union Strategy, published 19 November 2025 as COM(2025) 835, is the Commission's explicit acknowledgement that legal access to data is not sufficient — the missing ingredient is high-quality, curated, AI-usable data.

The Strategy's practical instruments include the roll-out of Common European Data Spaces across priority sectors (health, mobility, energy, media, and research) backed by roughly €100 million in EU investment as of 2026; an expansion of high-value datasets under the Open Data Directive, including 30 million digitised cultural objects for AI training; and a push toward synthetic data generation funded through Horizon Europe. The accompanying Digital Omnibus package is intended to consolidate and simplify the resulting legal framework, which the Commission itself has acknowledged has become fragmented across overlapping instruments.

3.4 From Information to Intelligence: The Missing Layer

BLACK&'s own framework treats the movement from raw information to usable intelligence as a distinct, necessary stage, not an automatic consequence of data volume or even of legal data access. Intelligence, in this framework, requires three properties that raw information lacks by default: governance (clear ownership, provenance, and lineage), structure (extraction and normalisation, particularly for the unstructured majority described in Section 3.1), and context (linkage to the specific decision or workflow the information is meant to inform). This is consistent with the empirical pattern in Section 3.2: the organisations capturing measurable value are those investing in governance and structuring, not those simply accumulating more data.

3.5 From Intelligence to Enterprise: Operationalising Decisions

Intelligence that remains in a static report or dashboard has not yet reached the Enterprise stage of the framework; it still requires a human to notice it, interpret it, and manually act on it. The Enterprise stage is reached when intelligence is embedded directly into a live operational workflow — pricing that updates automatically, a fraud flag that triggers a review, a forecast that adjusts a procurement order — removing the manual translation step. Industry forecasting for 2026 anticipates this shift explicitly: Gartner projects that 75% of enterprise data will be processed at the edge, closer to the point of decision, and that non-technical business users will create 75% of new data-integration flows themselves via natural-language tools, both trends that compress the distance between intelligence and operational action.

3.6 From Enterprise to Assets: Compounding Value

The final stage of the framework distinguishes a one-off operational improvement from a durable asset. A decision workflow, a structured dataset, or a trained model that is reused across multiple business functions or ventures compounds in value with each additional use, in the same way a venture studio's shared infrastructure compounds value across its portfolio of companies (see BLACK& Research, The Intelligence Company Builder, Section 3.3). This is also the logic behind the EU's own high-value-dataset and data-space policy: a well-governed dataset made reusable across a Common European Data Space is explicitly designed to generate value repeatedly, across many downstream users, rather than once.

4. Data and Charts

The following exhibits quantify the growth of unstructured data as a share of the enterprise, the gap between governance intent and governance reality, the scale of the EU data economy, and the BLACK& value-chain framework presented in Section 3.4-3.6.

5. Strategic Implications

For institutional clients and government stakeholders evaluating data and AI investment priorities, three implications follow directly from the evidence above.

• Data-access initiatives should be paired with governance investment, not substituted for it. The EU's Data Union Strategy correctly targets data availability, but the evidence in Section 3.2 shows that organisations already possessing large data volumes are failing to capture value primarily on governance and quality grounds, not access grounds.
• The highest-leverage investment is the structuring layer, not additional data collection. Since 80-90% of enterprise information is unstructured and already exists, the marginal value of new data collection is lower than the marginal value of making existing unstructured information decision-ready.
• Reusability should be a design criterion from the outset. Treating a dataset, model, or workflow as a durable asset — reusable across functions — rather than a single-use deliverable is what allows the Enterprise stage to compound into the Assets stage.

6. Limitations

• Several unstructured-data volume and data-quality statistics in Section 3.1-3.2 are drawn from data-management vendors and statistics-aggregation publishers rather than peer-reviewed or governmental sources; figures vary across studies depending on definitions and survey populations, and this paper reports commonly cited ranges rather than claiming a single authoritative figure. Readers requiring audit-grade sourcing for any individual statistic should request the primary study referenced by the aggregator rather than rely on this paper's citation alone.
• The EU data-economy valuation (€301 billion in 2018, €829 billion by 2025) is a European Commission estimate and projection methodology current as of its publication; it has not been independently re-verified for this paper.
• The Information-Intelligence-Enterprise-Assets framework in Sections 3.4-3.6 is BLACK&'s own proprietary structural model; it is presented as an interpretive lens consistent with the cited third-party evidence, not as an independently validated academic framework.
• This paper does not include primary data-maturity assessments of BLACK& client organisations; no such assessments are cited or relied upon.
• The Data Union Strategy was a recently published policy communication (November 2025) as of the time of writing; specific implementation timelines and instruments remained pending and subject to change.

7. References

Citation format: Chicago author-date.

1. European Commission. 2020. "A European Strategy for Data." COM(2020) 66 final.
2. European Commission. 2025. "Data Union Strategy — Unlocking Data for AI." COM(2025) 835 final, 19 November 2025.
3. European Commission. 2026. "European Data Union Strategy." Shaping Europe's Digital Future.
4. IAPP. 2025-26. "Notes from the IAPP Europe: The Data Union Strategy" and "A View from Brussels." Europe Data Protection Digest.
5. Interoperable Europe Portal. 2026. "Data Interoperability (RP 2026)." HLF Workstream 14 report.
6. DATAVERSITY. 2026. "Data Management Trends in 2026: Moving Beyond Awareness to Action."
7. GitNux. 2026. "150+ Data Management Statistics: 2026 Verified Report." [Statistics-aggregation publisher; treat cited figures as directional — see Limitations.]
8. ZipDo. 2026. "100+ Unstructured Data | 2026 Report." [Statistics-aggregation publisher; treat cited figures as directional — see Limitations.]
9. Komprise. 2026. "State of Unstructured Data Management," fifth annual industry survey.
10. Congruity360. 2024. "The Future of Data: Unstructured Data Statistics You Should Know."
11. Hyperight. 2026. "Top 12 Data Management Predictions for 2026."
12. Typedef.ai. 2026. "Unstructured Data Management Statistics." [Secondary compilation; original sources cited where identifiable.]
13. IDC. 2026. Research on data governance maturity and AI initiative ROI, as cited in DATAVERSITY (2026).

8. Revision History

Next scheduled review: upon publication of further Data Union Strategy implementation measures, or Q4 2026, whichever is earlier.