Data & Datafication
Turning what happens in the world into records a computer can read, including things nobody used to count, like footsteps or soil moisture. You cannot steer what you cannot see, and no amount of clever analysis replaces a missing measurement. In Europe, more than half of firms with ten or more staff pay for cloud computing, but only four in ten analyse their data at all. Nobody measures how much stored data is good enough to decide on.
Maturity
Where it feeds in
Where you have already met this
In 2025, 53% of EU firms with ten or more staff paid for cloud computing. 40% did any data analytics at all. The data is being kept. Nobody counts how much of it is good enough to decide on. No statistics office publishes that number.
If this moved
If institutions measured what their decisions need rather than what is easy to capture, the diagnostics gap would lose one support - and stay at 2, held by epistemic trust. Precision everything, government as a service and cognitive-work automation would also stay at 2, held by the other enablers in their ties, managerial capability among them.
What to watch for
The first national accounts that record data as an asset under the 2025 System of National Accounts. A statistics office adding a survey question on the quality of firms' own records.
Not to be confused with
Digitisation. Digitisation scans the paper form. Datafication turns the behaviour behind it into records that can be counted and queried.
What it is
The capture of activity as machine-readable records - and the extension of that capture to domains that were not previously measured at all.
How it works
- 01Something happens and is recordedA sensor reading, a purchase, a click or a form entry becomes an event with a time stamp.
- 02The record is structuredThe event is written in a fixed format so software can read it, and sent into a pipeline.
- 03It is cleaned and joinedPipelines drop malformed records, enforce the expected fields and link records that refer to the same person, asset or transaction.
- 04It is stored for queryingRecords land in databases or data warehouses built to be searched and summed.
- 05Someone asks a question of itDashboards, models and rules run over the stored data. Whether the answer is right depends on whether the data was captured for that question - which is rarely checked.
State of play
- In 2025, 52.7% of EU enterprises with at least ten employees bought cloud computing services and 39.9% performed data analytics (Eurostat, digital economy and society statistics – enterprises, data extracted Jan 2026).
- 46.5% of EU enterprises used enterprise resource planning (ERP) software in 2025 (Eurostat, data extracted Jan 2026).
- The UN Statistical Commission unanimously recommended the 2025 System of National Accounts in March 2025; it records data as a new category of produced fixed asset (UN Statistics Division, 2025 SNA, Mar 2025).
- The OECD's Going Digital Measurement Roadmap calls for new data infrastructures able to monitor digital activity and data flows on a timely basis (OECD, 12 Mar 2026).
- No national statistical agency publishes an audited share of enterprise records that meet a decision-grade quality standard (documented absence across OECD, Eurostat and UNCTAD sources, re-checked Sep 2026).
Why it matters
It sets the gradient on anything that requires knowing the current state of a system. Where the data does not exist, the constraint is not analysis but measurement, and no amount of modelling substitutes.
- It sets the reading on anything that needs to know the current state of a system. Where the data does not exist, the constraint is measurement, and no model substitutes.
- It feeds the diagnostics gap, precision everything, cognitive-work automation and government as a service - all at 2, and it shares the constraint on each.
- The most quoted number in the field, that 80% of enterprise data is dark, has no census behind it. The honest figure is that nobody has measured it.
What holds it back
Measurement design
Data is captured for billing, compliance or marketing, and then asked to answer questions it was never built for.
Quality and lineage
Few organisations can say where a record came from and what was done to it. Without that, a figure cannot be trusted for a decision.
Privacy law
Rules on personal data limit reuse, which is intended; it also means much stored data cannot legally be joined.
No common yardstick
There is no accepted standard for how good a dataset is, so nobody reports it.
Terms
- Datafication
- Turning activities, behaviour and physical states into structured, machine-readable records.
- Data lineage
- The record of where data came from and every step that changed it.
- Schema
- The blueprint that fixes which fields a record has and what each may contain.
- Decision-grade data
- Data checked for accuracy, origin and timeliness well enough to act on without re-verifying it by hand.
- Data as an asset (2025 SNA)
- The new national-accounts treatment that counts the information content of data used in production as fixed capital.
Steepens
| Megatrend | Weight | Role |
|---|---|---|
| The Diagnostics Gap | large | shared weakest link (with 1 other) |
| Precision Everything | large | shared weakest link (with 4 others) |
| Automation of Cognitive Work | moderate | shared weakest link (with 5 others) |
| Government as a Service | moderate | shared weakest link (with 6 others) |
| Climate Resilience Imperative | moderate |
The gradient behaves like a sticky rubber band: it is pulled hardest toward the weakest enabler. Marginal links are shown but do not pull.
If this moved one step
Computed from the map as it stands: this score raised from 2 to 3, every other score unchanged. It is arithmetic, not a forecast.
Raised one step on its own, it lifts no megatrend's weakest link; each gradient it feeds still rises a little.
Tied as the weakest link
Here it shares the weakest spot with other enablers. Raising it alone leaves the weakest link where it is and lifts the gradient a little; the enablers it is tied with must move too.
- Automation of Cognitive Work2.7/5 , 2.8/5tied with Digital Transformation, Managerial Capability, Neuromorphic & Alternative Compute, Regulatory Clarity, Workforce Trust
- Government as a Service2.5/5 , 2.6/5tied with Digital Transformation, Generational Turnover, Managerial Capability, Public Procurement, Regulatory Clarity, Smart Cities
- Precision Everything2.7/5 , 2.8/5tied with Digital Twins, Managerial Capability, Smart Cities, Workforce Trust
- The Diagnostics Gap2.6/5 , 2.8/5tied with Epistemic Trust
The gradient also rises in
- Climate Resilience Imperative2.3/5 , 2.4/5
What would change the gradient
Conditional, not predictive. No dates, no probabilities.
- 01Measuring what decisions actually require, rather than what is easy to capture
- 02Data quality and lineage treated as infrastructure rather than as a project
- 03Access rights that let the measured see their own data
Common misconception
That more data means better decisions. Data volume has grown enormously while institutional ability to answer basic questions about their own state has not - which is the diagnostics gap, and it is a measurement-design problem rather than a volume problem.
Talin, B. (2026). Data & Datafication. The Future Explorer, dataset v0.4.0 (2026-11-16). https://explorer.talin.digital/e/data-and-datafication/