There is a comfortable belief that keeping every document indefinitely is the safe option. Under modern data protection law the opposite holds: personal data kept beyond the purpose that justified collecting it is unnecessary exposure - and every extra file is one more item to locate when a data subject request arrives.
Start with the retention schedule
A retention schedule defines, per document type, the active retention period, the intermediate period and the final disposition - disposal or permanent preservation. It emerges from three inputs: legal obligation, operational need and historical value.
This document belongs to the business, not IT. Technology's role is to apply the schedule, not invent it - and skipping that step is precisely where governance projects stall.
Retention labels applied automatically
With the schedule ready, each row becomes a retention label. The critical point is automatic application: a label that depends on users choosing it manually is not a control, it is a suggestion.
- By content type: the most predictable route - every Contract is born with the right label.
- By location: a default label applied to an entire library.
- By sensitive content: pattern detection to classify documents containing personal data automatically.
The retention trigger: the most common error
The clock can start at creation, last modification or an event. For contracts the right trigger is rarely the creation date - it is the end of the term. Event-based retention requires someone to raise the event, and that must be part of the contract closing flow.
Disposition review before disposal
For sensitive archives, configure a disposition review: at the end of the period a reviewer receives the list of eligible documents and chooses to dispose, extend or reclassify. Each decision is recorded - that is what makes disposal defensible if questioned later.
Data subject rights in practice
When a request arrives, the operational question is blunt: where are the documents containing this person's data? Answering in hours rather than weeks depends on two things done beforehand - consistent confidentiality classification and metadata that supports targeted search.
Well-configured retention is what turns a document archive from legal liability into an organised asset.
Recurring mistakes
- Applying one period to everything, ignoring the difference between an invoice and an internal policy.
- Configuring labels without publishing the schedule that justifies them.
- Leaving label application to the end user.
- Never testing disposal - and discovering on first execution that the trigger was wrong.