Data infrastructure underpins business innovation, public services and civil society, but is often broken and neglected. These are the ODI‘s principles to guide how data infrastructure can be strengthened to benefit everyone
Data is infrastructure. Having a strong data infrastructure will only become more vital as our populations grow and our economies and societies become ever more reliant on getting value from data to meet people’s needs.Data infrastructure connects together different parts of our society and economy. Weather data is being used by everyone from farmers to the transport industry to individual citizens. Mapping data is created and shared by the public sector and then built on by diverse organisations, from Google to construction companies to the home insurance industry. People buying a home might use a service that combines data on house prices, schools, transport times and insurance premiums. Data is infrastructure for our cities, nations and globally across each and every sector.
Society has not learnt the lessons from the evolution of our road, railway and energy networks in the industrial revolution when we realised that while our infrastructure might be created by industrial giants, it had to be managed in a way to maximise benefits to society. To maximise value from data we need data infrastructure to be as open and accessible as possible while respecting privacy. Data could contribute more value to our economies and our lives than it currently does.
Design principles for data infrastructure
These principles for data infrastructure complement our principles for personal data. The personal data principles will help to build services and find insights in ways that people can understand and trust. Data infrastructure provides a foundation that allows those services and insights to flourish. These design principles should be used when that foundation is being designed, built and assessed
Data infrastructure includes datasets, the technology, training and processes that makes them useable, policies and regulation such as those for data sharing and protection, and the organisations and people that collect, maintain and use data.
1. Design for open
Open data, open culture, open standards, open source and collaborative models build trust, reduce cost and create more value than other approaches. Being open improves quality as more people can contribute to the outcome and increases the number of connections that can be made. Data benefits from network effects: it creates more value as more people use, contribute to and maintain it.
2. Build with the web
We need to learn how to publish, discover, use and link together data across the web. Data on and in the web is continuing to grow with more devices being connected and interconnected every day. The billions of people, sensors and services on the web produce and use data. Data infrastructure must support the web of data.
3. Respect privacy
In the most impactful and valuable data infrastructure, openness is maximised but what is private remains private. Different countries have their own data protection legislation and social contracts, which need to be adhered to. To build trust organisations using personal data should also be open with people about how they use and share that data.
4. Benefit everyone
Data infrastructure components should be designed and supported to benefit as many stakeholders as may use it. Everyone should benefit from the innovation, services and insights that the whole data infrastructure allows. Sometimes data infrastructure that is as open as possible will benefit the organisations that maintain the data, in other cases it will not. To benefit everyone it will be necessary for governments to provide support for some components.
5. Think big but start small
Don’t start big. Start with the problems that are making it hard for people to make decisions or build new services, be agile and learn from experiments. Concrete and tar don’t go out of date as quickly as data technologies do.
6. Design to adapt
Expect needs to change, and expect other needs to vary between different stakeholders and local contexts. Be prepared to experiment with new technologies and ideas, look for desire paths, measure impact, learn from what works and what doesn’t. Any part of data infrastructure might start as a small experiment but turn out to create significant impact and have high demand. If it is designed to adapt using approaches like human-centred design, by encouraging innovation and by using flexible modular approaches, this is most likely to happen.
7. Encourage open innovation
The best ideas can come from anywhere: individual citizens, large or small organisations and from the public, private or third sectors. Strong data infrastructure and open innovation will encourage and stimulate fair and equitable markets and innovative ways to both maintain data and use it to create new services.