Data Retention, Deletion, and Minimisation

Article 5(1)(e) of GDPR requires that personal data be ‘kept in a form which permits identification of data subjects for no longer than is necessary for the purposes for which the personal data are processed’ — the storage limitation principle. This principle is violated by organisations that retain personal data indefinitely ‘just in case’, that keep data beyond the period for which it was originally collected, or that accumulate data in backups, archives, and legacy systems without a plan to delete it.

Storage limitation enforcement is increasing in DPA activity. Regulators have found that many organisations have detailed collection processes and security measures but no systematic approach to deletion. The result is expanding data lakes of personal data that were relevant at collection, retained beyond their useful life, and eventually lost to a breach or found in a regulatory audit. Data you do not hold cannot be breached and cannot generate a retention violation.

 

The Retention Schedule: Core Documentation

A retention schedule is the foundational document for storage limitation compliance. It maps each data category to a defined retention period and the criteria or legal basis for that period. The retention schedule should be specific: ‘customer contact data: retained for the duration of the contract plus 2 years for legal claims’ is a compliant entry; ‘customer data retained as long as necessary’ is not, because it provides no defined endpoint.

RETENTION SCHEDULE — REQUIRED FIELDS PER DATA CATEGORY

FieldContentExample
Data categoryThe specific category of personal data covered by the entryCustomer transaction records
Business system(s) containing dataWhich systems or databases hold this dataSalesforce CRM; order management database; email archive
Retention periodDefined period from a specific trigger point7 years from date of transaction
Trigger eventThe event that starts the retention clock runningDate of last transaction; date of contract termination; date of employee leaving
Basis for retention periodThe legal or regulatory obligation, business need, or limitation period that justifies the retention periodTax law requires 7-year retention of financial records (HMRC guidance)
Deletion methodHow the data is deleted at end of retention periodAutomated deletion via system workflow; manual deletion by data steward; secure erasure for archives
Exceptions or holdsAny circumstances that extend retention (litigation hold; regulatory investigation; consent to extended retention)Data subject consented to extended retention for future offers; regulatory investigation hold in place
Review dateWhen the retention period and basis should be reviewedAnnual review; or when relevant law changes; or when processing purpose changes

 

Defining Retention Periods by Data Category

Retention periods are not arbitrary choices. They should be determined by the purpose for which the data was collected, the legal and regulatory retention obligations that apply to the data, and the limitation periods for claims the data may need to support. The principle is that data should be retained for as long as it is needed for the defined purpose, and no longer. Where legal obligations specify a minimum retention period, that minimum is the floor, not the ceiling.

RETENTION PERIODS BY COMMON DATA CATEGORY

Data CategoryTypical Retention BasisSuggested Retention PeriodKey Consideration
Financial and tax recordsLegal obligation (tax law, accounting regulations)6–7 years from end of tax year in which transaction occurredSpecific period varies by jurisdiction; confirm local tax law requirement
Employment records (general HR)Legal obligation; legal claimsDuring employment plus 6 years (UK limitation period for contract claims)Longer retention may apply for pension, occupational health, or certain health and safety records
Recruitment records (unsuccessful candidates)Legal claims (discrimination)6 months to 1 year after notification of decision (sufficient for claims); longer if consent obtainedLonger retention (2 years) permissible with specific consent for consideration for future roles
Customer contract recordsLegal claims; contractual obligationDuration of contract plus 6 years (limitation period for contract breach claims)Limitation period varies by jurisdiction; confirm local law
Marketing consent recordsAccountability; legal claimsDuration of consent plus 3 years (sufficient to defend claims that marketing was sent without consent)Must retain evidence of the consent itself; not just the data subject’s details
Website analytics (user behaviour)No legal obligation; business analytics13 months maximum (CNIL guidance); review annuallyMany analytics tools default to longer retention; must be configured to comply
CCTV footageSecurity; legal claims30 days standard; longer if incident-related footage retained pending investigationLonger retention must be documented; CCTV policy should specify retention period
Access logs and audit trailsSecurity; legal claims; accountability1–2 years; longer if required for regulatory investigationMinimum period to support security incident investigation; review against regulatory requirements
Health data (patient records)Regulatory obligation (varies by healthcare regulation)As required by applicable healthcare regulation; varies significantly by jurisdiction and record typeNational healthcare law overrides GDPR minimum; DPA guidance and sector regulators should be consulted

 

Implementing Deletion: Automated and Procedural

Defining retention periods is only half of storage limitation compliance. The other half — and the more operationally challenging half — is ensuring that personal data is actually deleted when the retention period expires. This requires either automated deletion workflows built into systems, or procedural deletion processes that are reliably followed. Most organisations need both: automated deletion for high-volume, routine data and procedural deletion for archive, backup, and legacy systems.

DELETION IMPLEMENTATION — APPROACHES BY SYSTEM TYPE

System TypeRecommended Deletion ApproachImplementation Consideration
Primary operational systems (CRM, HR, e-commerce)Automated deletion triggered by retention schedule; deletion review queue for records approaching retention end; automated deletion at defined triggerMost modern SaaS platforms support retention rules and automated deletion; configure and test deletion workflows at implementation; validate deletion actually executes
Data warehouse / analytics platformScheduled deletion jobs; retention policies in data pipeline; ETL process filters out personal data beyond retentionAvoid persisting personal data in data warehouse longer than necessary; implement data lifecycle management in pipeline
Email archiveAutomated deletion of emails beyond retention period; litigation hold mechanism to preserve specific emails when requiredEmail archives frequently contain large volumes of personal data beyond retention; configure archive retention limits; test deletion
Backup and disaster recovery systemsFlag records for deletion in live system; record in deletion log; confirm deletion at next backup rotation cycle; document approachTrue deletion from backup is technically complex; acceptable approach is to flag deleted records so they are excluded from any restore; document the methodology
Legacy systems / data archivesManual deletion with data steward sign-off; project approach to decommission legacy systems holding personal data beyond retentionLegacy systems often have no deletion capability; manual export and deletion; decommission planning should include data deletion as a requirement
Paper recordsSecure disposal (cross-cut shredding or confidential waste service); disposal certificate for sensitive recordsPaper records subject to same retention schedule as digital records; often overlooked; include in retention schedule and disposal procedure

 

Data Minimisation at Collection

The storage limitation principle works upstream as well as downstream. Retaining data for shorter periods is easier if less data is collected in the first place. Data minimisation — collecting only the personal data that is necessary for the specific purpose — reduces the volume of data subject to retention management, reduces breach severity, and simplifies compliance.

DATA MINIMISATION — IMPLEMENTATION CHECKPOINTS

CheckpointQuestion to AskMinimisation Action
Form and registration designIs every field on this form collecting data that is actually needed for the stated purpose? Are any fields optional that could default to absent?Remove or make optional any fields not strictly necessary; periodic form review against current processing purpose
Third-party data purchasesIs the full purchased dataset necessary? Can the data purchase be scoped to only the fields and segments actually used?Scope data purchase narrowly; delete fields not used; do not retain purchased data beyond the campaign or purpose for which it was acquired
Analytics and trackingAre we tracking more user behaviour than we actually analyse and use? Are we collecting identifiable data where aggregate data would serve the same purpose?Configure analytics to collect only what is actually used; sample rather than full tracking where sufficient; anonymise where possible
Employee data collectionAre HR systems collecting data about employees that is not needed for employment management? Are monitoring systems collecting more data than the stated monitoring purpose requires?Review HR data fields against employment contract and legal requirements; remove fields not required; configure monitoring proportionately
BITLION INSIGHTThe organisations with the strongest storage limitation compliance are those that have made retention and deletion a system requirement rather than an afterthought. When a new system is procured or built, the data retention requirements should be specified as part of the technical requirements — including automated deletion capability, configurable retention periods, and deletion audit logging. Retrofitting deletion capability into systems that were built without it is expensive and technically complex. Requiring it at procurement costs nothing and ensures that the tool serves the compliance programme rather than undermining it.