Interoperability and Open Standards for Public Sector Websites

Interoperability is usually written about as a policy principle and experienced as a broken link. This article translates the principle into tests a team can actually run on its own website, on the assumption that a standard nobody verifies is a standard nobody follows.

Stable identifiers

Every published resource needs an address that is expected to outlive the software serving it. That means human-meaningful paths, no platform artefacts in URLs, no dependence on query parameters for identity, and a written policy that changing a URL requires a redirect rather than a decision.

The test: take a crawl from three years ago and request every URL. The percentage that resolves, directly or through a single redirect, is your identifier health. Anything below the high nineties indicates a policy gap rather than an accident.

Machine-readable meaning

Semantic interoperability means another system can tell what a page is without a human reading it. In practice: correct document structure, a single descriptive heading per page, language attributes, and schema.org types that match reality — Article for articles, Organization for the publisher, BreadcrumbList for hierarchy, FAQPage only where questions genuinely exist.

Structured data should be generated from content fields, not written into templates as fixed text, so that it remains true as pages change. Validate it as part of deployment rather than once at launch.

Exportable content and technological neutrality

Content belongs in the content system, not in the theme. The test is blunt: could the entire site's text, media and structure be exported today, in a documented format, and re-imported elsewhere? If the answer requires a consultant, the site is not neutral regardless of which licence its software carries.

Self-hostable software, documented data formats, standard media encodings and infrastructure described as code are what make that answer yes. They also make procurement honest, because an exit is a specification rather than a negotiation.

Accessibility as interoperability

An accessible page is one that works with software the publisher does not control — screen readers, translation tools, reader modes, crawlers. That is the same property interoperability describes, arriving from a different direction, which is why accessibility conformance and machine readability improve together.

Test with the keyboard, with a screen reader, at 200 percent zoom and with images disabled. Anything that fails those is also failing some machine consumer you have not thought about.

Technical references

The documentation below covers the procedures behind this article in far more operational detail than an editorial piece can.