Your company may already pay a vendor for information the European Union publishes free every day. Company registries, satellite imagery, transport schedules, and weather records sit on public portals under an open license, and most businesses never open the file. Competitors who did open it are already selling products built on it.
What Open Data Means in Practice
Open data is government or public sector information released under a license anyone can reuse, free or near free, in a format a computer can read without conversion. A PDF report from a ministry is public information. A structured dataset with an open license, updated on a schedule and available through an API, is open data. The European Union settled the difference in law with the Open Data Directive (2019/1024). The directive became law in 2019, and every member state had to write it into national law by July 2021. The directive replaced an older Public Sector Information Directive and pushed member states past basic access toward standardized, machine-readable, low-cost reuse.
A follow-up regulation gave the directive weight. The High-Value Datasets Implementing Regulation (EU 2023/138) took effect in February 2023 and became binding on member states in June 2024. The rule names six categories public bodies must publish free of charge, in machine-readable form, through an API: geospatial information, earth observation and environment data, meteorological data, statistics, company and company ownership records, and mobility data. Companies building products in property, agriculture, insurance, logistics, or finance now hold a legal claim to their sector’s raw material, backed by regulation rather than goodwill.
Where to Find It
The starting point is data.europa.eu, a merged portal combining the old EU Open Data Portal and the European Data Portal into one catalogue since April 2021. As of mid-2026, the portal lists more than 1.7 million datasets from 36 countries, spanning all 27 EU member states plus European Free Trade Association and candidate countries. The counter updates continuously, so treat the exact figure as a snapshot rather than a fixed number.
National portals sit underneath the EU-level catalogue. France runs data.gouv.fr, Germany runs GovData, and the United Kingdom, outside the EU framework since Brexit, still runs data.gov.uk. For satellite and environmental data, the Copernicus Data Space Ecosystem gives free access to imagery from the Sentinel satellite fleet, refreshed continuously and licensed for commercial reuse. For transport, every member state has run a National Access Point since December 2019, publishing schedules, fares, and increasingly real-time data from taxis, bike shares, and car-sharing fleets in the standardized NeTEx and SIRI formats mobility apps already read.
What the Data Is Worth
The European Data Portal’s 2020 study, “The Economic Impact of Open Data: Opportunities for Value Creation in Europe,” put the open data market at roughly €184 billion in 2019 and projected growth to somewhere between €199.5 billion and €334.2 billion by 2025. No newer EU-wide market sizing study has replaced it: data.europa.eu still lists the report as its core reference on open data value in current training material. Treat the 2025 figure as a six-year-old projection rather than a confirmed outcome, and watch for an updated study given how much EU data regulation has changed since 2020.
Copernicus offers a cleaner picture because the European Space Agency and the EU Agency for the Space Programme track it directly. A 2018 study projected the programme would generate between €67 billion and €131 billion in benefits to European society between 2017 and 2035, a return of ten to twenty times its cost. Earlier figures put cumulative economic value between 2008 and 2020 at roughly €13.5 billion against forecast investment of €7.4 billion over the same period. EUSPA’s market report puts global revenue from Earth observation data and value-added services at €3.4 billion in 2023, on a path to reach nearly €6 billion by 2033. More than a tenth of the income earned by European companies reselling or repackaging Earth observation services traces back to Copernicus data they never paid to acquire.
Government-side quality is improving too. The 2025 Open Data Maturity assessment, which scored 36 countries across policy, portal, data quality, and impact, found the average maturity score rose from 80% in 2024 to 81% in 2025. Cyprus, Estonia, Italy, Czechia, Lithuania, Spain, Ireland, Slovakia, Ukraine, Poland, and France scored in the top “trendsetter” band. A rising baseline means the data a company builds on today gets more reliable, not less, over time.
Companies Already Building on It
Spend Network, a UK-founded analytics company, pulled public procurement, tender, and contract data from roughly 330 government sources across the UK and EU and turned it into an API covering more than 30 million transactions worth over £878 billion. Consultancies, suppliers, and public bodies now pay to query the dataset rather than scrape 330 sources individually.
Doorda took a narrower slice of the same idea. The company aggregates open datasets from Ofsted, the Land Registry, the Food Standards Agency, the NHS, and the Home Office into a single map product, letting property buyers, retailers, and local businesses see school ratings, crime patterns, and hygiene scores for any UK address in one search.
Open Food Facts built a food product database from public labeling data and volunteer contributions, and the resulting dataset now powers nutrition apps and retailer tools across several European markets. In Amsterdam, the city’s Energy Atlas project publishes open energy consumption data at building level, and companies building efficiency and retrofit tools use it to target buildings without commissioning new surveys first.
None of the companies above invented the underlying data. Each one built a product around access, structure, or combination the public source lacked.
New Rules Reshaping What’s Available
Two additional laws expand the raw material available to companies, though neither one counts as classic open data.
The EU Data Act entered into force on 11 January 2024, with its main obligations phasing in from 12 September 2025 through full application by 12 September 2027. The law gives users of connected products and related digital services a right to access data the products generate during use, and it requires companies operating the services to support the access. A manufacturer selling a connected machine to a business customer will owe the customer, and potentially a competitor the customer names, a route to the usage data the machine produces. Companies building or buying connected products should treat the compliance deadline as a planning input now, not a scramble in 2027.
The Data Governance Act, in force since June 2022 and applicable since September 2023, works alongside the Data Act. It regulates a new category of data intermediaries and lays out rules for voluntary data sharing, forming the legal foundation for the sectoral Common European Data Spaces now being piloted in health, mobility, energy, and manufacturing.
Where the Friction Still Shows Up
Format quality varies sharply by dataset and by country. Beneficial ownership data is a clear example: the European Union names company ownership as one of the six high-value categories, yet only Denmark and Latvia currently publish the data in a structured, machine-readable format with API access. Most other member states still release ownership extracts as PDF documents, which technically satisfies a transparency requirement while doing little for a company trying to build automated checks on top of it.
Fragmentation compounds the format problem. Thirty-six national portals, hundreds of local catalogues, and a wide range of publishing maturity mean a company operating across several EU countries cannot assume one dataset schema works everywhere. The Open Data Maturity scores back this up: Albania, Malta, and Germany posted the largest year-over-year gains in 2025, which is encouraging, but a country climbing eleven or twenty percentage points in one year was starting from a lower base than the trendsetters above it.
Privacy law adds a separate layer of risk. Under the General Data Protection Regulation, only genuinely anonymized data falls outside the rules. De-identified data can still count as personal data if someone could re-identify individuals by linking it to another dataset, and combining two public sources creates exactly the kind of linkage regulators warn about. A company merging open government data with internal customer records should treat the combination as a privacy decision, not a technical one.
A Practical Path for Companies
An open data pipeline works better as a sequence than as a single research project.
- Map the six high-value dataset categories against your sector and identify which one supplies a core input: company ownership records for due diligence, mobility data for logistics, or geospatial data for site selection, among others.
- Start at data.europa.eu, then drop down to the matching national portal or National Access Point for local detail the EU-level catalogue will not carry.
- Check the license terms and update frequency before building anything on a dataset. A file updated once a year or released without a clear license can break a product faster than no data at all.
- Pair public data with internal records. Open data by itself rarely creates a lasting edge; combined with proprietary customer or operational data, it commonly does.
- Track the Data Act deadline if your company sells a connected product. New obligations to share usage data with customers and named third parties phase in through September 2027.
- Run a privacy review before publishing or combining any derived dataset touching identifiable individuals, even when every underlying source counts as anonymized in isolation.
The Window Is Still Open
Europe’s open data supply keeps expanding, and the Data Act adds a second stream by 2027, compelling companies building connected products to share the data their products generate. Companies already routing the material into internal pipelines will move first. Companies still waiting on a vendor invoice will keep paying for what a competitor down the street already gets for free.
