The completeness and versioning of a Data Card should ideally be determined by the maturity of the dataset.
Changes may come from upstream sources of your dataset – or downstream context shifts in how your dataset is being used. Anticipate the types of updates or changes your dataset might go through over time, and have a plan for maintaining your Data Card.
Update your Data Card when you update your dataset.
Updates to your dataset should ideally be in lock-step with updates to your Data Card. This is an important part of ensuring that your accompanying documentation and dataset don’t become outdated over time.
Always have a plan for any time your upstream sources of information change. Be cautious of what would constitute a major or minor update to your dataset, and subsequently your Data Card. Set criteria for changes that will need to be reflected in your data card.
Regularly check in on your readers.
If your audience changes to include more job functions or prioritizes a niche audience, your Data Card can risk becoming less effective or alienating existing dataset users.
Strategize a way for existing dataset users to continue accessing the Data Card and any information that might have been deprioritized for new or updated readership. Even without this change, factor in a few minor, trailing adjustments after you release your Data Card.
Don’t assume that readers will be able to identify the changes made on an update to the Data Card. Continue to treat removals or erasures with care. Include a timestamp or changelog for your Data Card, and offer some outreach or communications about these changes.