Market Accord aims to publish useful field learning while protecting participants, buyers, users and commercially sensitive information. Replacing a company name with “an SME” is not enough when the surrounding facts point to one organisation.
Assess the whole disclosure
Consider direct identifiers, combinations of facts, information already public, motivated users, likely search methods and the harm that re-identification could cause. The assessment must reflect the audience and context, not only the dataset in isolation.
Publication choices
- Named: specific, informed permission and a clear benefit.
- Organisation named, people removed: only where roles and events do not identify individuals.
- Generalised: broader sector, geography, dates or values.
- Aggregated: combine a cohort where a single record is too revealing.
- Delayed: publish after commercial or procurement sensitivity reduces.
- Composite: combine recurring patterns and label the result clearly as composite.
- Withheld: do not publish where useful detail cannot be separated safely.
Consent is not a universal answer
A participant may agree to a case study without understanding that a contract value, buyer type and date make them identifiable. Consent must be specific and does not remove obligations to handle other people's data or confidential information.
Corrections and review
Anonymisation can weaken over time as new public information appears. Market Accord keeps a review date and route for participants to challenge a disclosure. Material changes are logged rather than silently rewritten.