What really matters in government open data management?

Written by Mykhailo Kornieiev, Expert at eGA

In the AI era, open data is becoming a key enabler for the digital transformation of the government and society. Many governments in the world have therefore been keen to deal with open data. However, sometimes they start from the wrong angle by treating open data projects as publishing projects that increase the country’s position in the open data index, rather than as ongoing data-management processes. Publishing data once is relatively easy. Managing it as a reliable public resource is much harder. 

From a data management perspective, I see the central challenge in government open data as maintaining quality, consistency, and accountability across the whole data lifecycle. A dataset published once and then forgotten is not useful. For a government organisation, the more mature approach is to manage each important dataset almost like a product: assign an owner, define quality requirements, use common standards and metadata, automate updates, monitor quality, and collect user feedback. Ukraine’s rise in the European Open Data Maturity assessment was the result of applying this broader approach and treating open data as part of digital transformation reform, connecting policy, institutions, quality, reuse, and trust.

Four lessons of open data management

My role was to connect these different layers: policy design, coordination with ministries and municipalities, training civil servants, communication with data reusers, impact studies and alignment with European approaches. And along the way, we learned several important lessons that can be useful for other data managers as well.

First, government organisations need to understand that data – not only open data – is a crucial part of digital public infrastructure. They needed to move beyond simply publishing files. This means establishing and maintaining processes for appointing data owners and defining publication plans, introducing clear licensing and machine-readable formats, developing feedback mechanisms, and collecting evidence of the value data creates.

Secondly, manage open data as a portfolio rather than a collection of files. High-value datasets should be prioritised because they can support anti-corruption efforts, business services, research, journalism and evidence-based decision-making.

Thirdly, data quality needs to be monitored continuously. Good metadata, reliable APIs, regular updates and compliance with standards such as DCAT-AP determine whether data can actually be found, understood and reused.

Fourthly, the impact of open data needs to be demonstrated and documented. Impact also had to be documented to understand whether companies, civic-tech teams, journalists, and public bodies actually reuse the data.

Ukraine’s approach: risk-based data governance

After Russia began its full-scale invasion, security became one of the most difficult challenges. Ukraine had to restrict access to some datasets because information that is harmless in normal circumstances may create a genuine security risk during wartime. The answer was not to abandon openness, but to introduce risk-based data governance: assess datasets individually, protect potentially sensitive information that could be used against Ukraine, enable safe reuse of non-personal data and explain decisions transparently. Trust can be lost both when information is hidden without a clear reason and when it is published without appropriate safeguards.

This approach contributed to Ukraine moving from 17th place in the European Open Data Maturity assessment in 2020 to becoming one of Europe’s leaders, securing second place overall in 2022, maintaining it despite the full-scale war, and still remaining one of the European trendsetters in the sphere.

Data management challenges and lessons learned

At e-Governance Academy, I see similar data management challenges in other countries: Ecuador, Germany, the Netherlands, Croatia, and beyond. Ecuador’s lesson is that interoperability and data management need to support real public services rather than exist as purely technical exercises. Work on implementing the Data Governance Act in Germany, the Netherlands, Estonia and Croatia shows that Europe still needs practical procedures for the secure reuse of protected public-sector data.

Legislation is necessary, but legal texts alone are not enough; governments also need practical mechanisms for deployment and implementation. Our AI-readiness work with cities Riga and Rotterdam reinforces the same point from another direction: AI readiness starts with data governance: inventories, clear responsibilities, quality, ethics and skills.

In Ukraine, data governance is part of the country’s wider movement towards the EU digital space, including alignment with the Data Governance Act, Data Act, Interoperable Europe Act, eIDAS 2.0, and AI Act.

The most important lesson for me, as a data governance expert, is simple: an open data ranking is not the goal; it is a mirror. Moving up in the ranking has little value if datasets remain outdated, difficult to reuse, or disconnected from real public needs.

What matters is what the improvement represents beneath the surface: clearer responsibility, higher-quality data, better interoperability, evidence of reuse, and greater trust. When these elements work together, open data becomes part of the state’s digital infrastructure and supports better public services, economic activity, research, accountability and evidence-based decisions.