Lessons learned in open standards development

Open standards for data are reusable agreements that make it easier for people and organisations to publish, access, share and use better quality data.

There are thousands of standards under development, in use, or retired worldwide. Here, we condense some of the important lessons emerging from the people and organisations using, sharing and developing open standards for data.

Use existing standards

Open standards for data are useful when you need an agreement from more than one party to support data to be produced. The problem needs a repeatable and reusable solution that maximises consistency and clarity. Consider open standards for data to strengthen data infrastructure.

If a standard exists that meets your needs, you are not yet clear on your needs, or cannot convene a community around your standard, creating a new open standard may not be the best option. Focusing on understanding the scope of the problem and potential solutions by commissioning a proof of concept to explore options, or working through a discovery phase to investigate the problem and solutions, can lead to a better result.

Read When not to create new standards for more about making a decision to create, adopt, or extend open standards.

Follow Open Stand principles

Adopt the five Open Stand principles, ensuring your open standard is developed fairly, transparently and cooperatively. An added advantage of open participation is engaging the community around your open standard before you launch to ensure there is sufficient interest.

Read Creating impact with open standards for more about creating impact through wide engagement.

Remember that language matters

People use language differently, even in the same organisation. Agreeing early on what concepts are included and being clear on what concepts mean will make your standard easier to understand and adopt. Include clear examples to make abstract concepts more concrete.

Read User experiences of open standards for data to learn more about what data users and data publishers revealed about open standards.

Consider common features and extensions

Standards are an excellent way of producing consistent data or models, but each stakeholder may hold or use data slightly differently. As part of stakeholder engagement, it is important to understand the common features as well as where vocabulary, language and data-use diverge.

Read experiences of mapping data to a standard, including US/UK legislative data mapping to Akoma Ntoso and the Supply side analysis report from the Open Contracting Partnership.

Understand how the standard will be used, and who will be using it

  • Will users of the standard be expected to publish data or use data?

  • Will users of the standard develop models to understand the flow of information?

  • Will they use the standard as a vocabulary to support sharing data?

  • Is the standard a guide to support developing a common language so others can publish data?

  • Is the standard a framework for collaboration around data?

  • What technology and skills do potential adopters have, and how will this shape their ability to adopt the standard?

  • What engagement, training or support will adopters need?

The outputs will affect the type of tools, guidance and support that users users of the standard need.

Choose the right development process

All open standards development processes fall somewhere on a spectrum between formal processes with clear governance and working groups and informal collaborations using loose guidelines.

The best development process is the one that works for the audience, sector and type of standard under development. Starting from the five Open Stand principles, consider what works for your open standard development and its community.

Be clear on purpose

Producing consistent data is a goal of all open standards for data. Depending on the purpose of the standard, this may be a secondary goal. The primary goal could be to change perceptions or convene stakeholders around a shared problem, make participation in a sector easier, or solve a problem affecting data infrastructure.

Clearly articulating the purpose of the open standard provides the basis for engaging the community consistently, brings focus to developing supporting tools and resources, and ensures the standard stays on track.

Read How to use the Open Standards for Data Canvas; this tool will help you document and share your understanding of the primary and secondary purpose of the standard.

Know your community

Open standards for data affect a wider community than data users and data publishers; they include the people and organisations that are affected by the standard, including the owner, sponsor and developer of the standard, and third parties.

It is important to understand the known roles in an open standard community and identify the key stakeholders. This will make engaging the community easier and will ensure you consider their technical, tool, communication and resource needs, including their levels of digital literacy, access to physical and digital spaces and anything else you might need to successfully engage with them.

Read Creating impact with open standards for more on understanding who open standards affect.

Consider who is missing

When deciding on technical language, file formats, location and format of guidance and resources, and where the community meets, be aware of who is and is not included. Using GitHub as a primary means of handling issues, for example, excludes people who do not use that platform. Convening meetings that require physical presence, especially in relatively expensive locations, excludes others.

Consulting the wide variety of people and organisations using or affected by the standard is important to develop a successful open standard. The needs of the community should influence these decisions rather than the personal preferences of the standard’s owner or developer.

Read the consult or fail section of our User experiences of open standards for data report to understand why poorly designed consultation and governance can cause problems with standards development.

Develop visible guidance

Clear and visible descriptions of the standard’s governance process and roadmap will increase the community’s trust and confidence. These do not have to be long documents: Google’s GTFS relies on a short set of guiding principles written in plain English.

Read Managing change in open standards to find out more about governance.

Agree ownership and copyright

Clearly stating who holds the intellectual property and copyright on the standard will help avoid issues such as misuse of branding. If the community is invited to contribute to the development of the standard, consider who retains the intellectual property and copyright on their contributions to avoid future legal problems.

Read Public transit data through an intellectual property lens: lessons about open data

Develop use cases

Open standards for data support the production of data, so it is important to have some idea of how the data will be used. This partly supports the argument for the standard to exist in the first place and, more importantly, helps the standard’s developer to understand what is essential and what is optional and could be excluded.

Differentiating between core and optional needs also helps focus resources and prevent bloated, over-specified standards that can be a barrier to adoption. Testing the standard with real-world data use will help refine the standard further and highlight issues early.

Read Use cases and lessons for the Data Cube Vocabulary from the W3C to learn more about the benefits and challenges in real-world use of this vocabulary. For a different perspective, review principals & use cases from the Open Ag Data Alliance on the challenges facing farmers and how open standards can help.

How to use this guide

There are a number of ways for you to learn more about the creation, development and adoption of open standards for data.

About this guide

This guidebook helps people and organisations create, develop and adopt open standards for data. It supports a variety of users, including policy leads, domain experts and technologists.

Something missing? Suggest a change or addition