
We’ve all found datasets we could not use. In research, we are no strangers to the thrill of a long search finally paying off, only to have it dampened by the realisation that we cannot make sense of the data we found. But where did it go wrong? Chances are, it’s not the data, but the metadata.
It’s easy to be confident about details and decisions while designing and conducting experiments. The longer time passes, the hazier everything gets, especially when we put too much faith in our memory and the random scribbles we call our notes. If even our own notes fail us after a few months, it’s no surprise that someone else’s data becomes unusable only years later.
Metadata provides the context surrounding your data. It enables future users, whether it’s machines, other people, or even yourself, to understand, manage, and reuse data. Metadata generally falls into three categories:
- Descriptive: What does the data represent?
- Structural: How are the different parts of the data organised and connected?
- Administrative: How was the data created, managed, and shared?

Metadata design and collection creates an inherent tension between individual researchers and the wider community. It’s undoubtedly extra work, but it’s a small investment with long-term returns.
- Metadata preserves context. It helps researchers return to their data and interpret it correctly. It also allows future researchers to judge whether the data is relevant, trustworthy, and suitable to build upon.
- Metadata connects knowledge. By using metadata standards and controlled vocabularies, researchers contribute to a shared understanding of what information should be collected, recorded, and communicated. In turn, this makes the associated data easier to discover, integrate, and reuse.
- Metadata empowers future research. As research becomes increasingly interconnected and data-intensive, there is a rising need for data to be both human and machine readable. Metadata helps ensure that data remains accessible not only to people today, but also to the automated tools that will drive tomorrow’s discoveries.
In POMATO, we plan to put these principles into practice. Guided by community standards, we designed our metadata from the outset so it could be captured alongside our experiments rather than reconstructed later. High-quality metadata makes data easier to discover, understand, and reuse, extending the impact of our research well beyond the lifetime of the project.
Questionnaire
Metadata standards evolve through the communities that use them. We’d love to hear about your experience. Your feedback will help guide future communication and resources within POMATO.