A GDPR retention schedule should identify each personal-data category, why it is kept, when it is erased or anonymised, and how erasure requests are handled without inventing one-size-fits-all retention periods.
Use it to set the rule for each data category, align privacy notices and records of processing activities, handle erasure exceptions, and prove that systems and processors apply the rule.
Apply by recording, for each personal-data category, the purpose, legal basis, retention period or objective criterion, start event, expiry action, locations, owner, exceptions, and evidence. The GDPR does not supply one retention period for all personal data, so this artifact shows how to build and operate the schedule without inventing sector deadlines.
1
Section 1
Start with storage limitation, not a default number of years
Article 5(1)(e) requires personal data to be kept in identifiable form no longer than necessary for the processing purpose. Start with a specific purpose and data category, then state the period or objective criterion, the event that starts the period, and the action at expiry.
A criterion such as 'while the account remains active, then for the applicable claims period' is more usable than 'keep as needed', but the actual claims period must come from the law that applies to that record and jurisdiction. Where longer storage is for archiving in the public interest, scientific or historical research, or statistical purposes, separate it from ordinary operational retention and record the Article 89(1) safeguards.
Deletion and anonymisation are different outcomes. Data is anonymous only when the person is no longer identifiable by means reasonably likely to be used, considering factors such as cost, time, and available technology. Pseudonymised data remains personal data and still needs a retention rule.
List the processing purpose before the retention rule.
Define the personal-data category and the category of data subjects affected.
State whether the expiry action is deletion, effective anonymisation, aggregation of anonymous results, or restricted continued processing for a documented purpose and legal basis.
Record the system, backup, archive, and processor locations where the same data category exists.
Set a review trigger for changes to the purpose, system, processor, law, contract, or risk; a review date does not replace the expiry rule.
Do not add sector retention periods unless a cited legal or policy source in the evidence record supports them.
Use a complete schedule row and align it with Article 30
For controllers, Article 30 requires records of processing activities to include purposes, categories of data subjects and personal data, recipients, transfers where applicable, envisaged erasure time limits where possible, and a general description of security measures. A retention schedule can supply the erasure field and supporting detail, but it does not replace the rest of the RoPA.
Processor records under Article 30(2) have different required content and do not include the controller's envisaged erasure time limits. Controllers should put retention and deletion instructions in the Article 28 arrangement and verify that processors and sub-processors can apply them.
The exemption in Article 30(5) for an organisation employing fewer than 250 people is limited: it does not apply to processing that is likely to risk individuals' rights and freedoms, is not occasional, or includes Article 9 special-category data or Article 10 criminal-conviction and offence data. Even where the exemption applies, and other GDPR duties remain.
The Irish Data Protection Commission guidance warns against vague RoPA references such as pointing to an unavailable retention policy or saying data is kept under a policy without stating the rule. The schedule should be specific enough for a reviewer to understand the retention position.
Schedule ID, controller role, business owner, system owner, and approval status.
Processing activity and purpose.
Categories of data subjects and categories of personal data.
Legal basis and the legal, contractual, operational, or risk rationale for the chosen period.
Recipient categories, including processors and third-country recipients where applicable.
Retention period or objective criterion, start event, expiry action, and any review trigger for each data category.
Locations and copies, including live systems, logs, exports, archives, backups, processors, and sub-processors.
Exception or hold rule, approving role, restricted uses during the exception, and release trigger.
Deletion or anonymisation method, downstream and recipient action, test frequency, last test, and evidence location.
Source for the rule, distinguishing binding law and contract from internal policy.
Make notices, access answers, and the schedule agree
Articles 13 and 14 require the controller to tell people the storage period or, when that is not possible, the criteria used to determine it. Article 15 requires the same type of information in an access response. The schedule, privacy notice, RoPA, processor instructions, and access-response material should therefore describe the same rule at the appropriate level of detail.
A schedule change can require more than an internal edit. If the controller plans further processing for another purpose, Articles 13(3) and 14(4) require information about that other purpose and the relevant further information before the further processing. The separate lawfulness and compatibility analysis still has to be completed.
Compare each published period or criterion with the current schedule row and processing reality.
Use criteria a person can understand; 'as long as necessary' merely repeats the GDPR and does not explain the decision.
When a purpose ends, run the expiry action unless a documented purpose and legal basis supports continued processing.
When a period changes, record the source, decision date, affected systems and processors, notice impact, implementation date, and migration treatment for existing data.
Route Article 17 erasure requests through the same schedule
Article 17 requires erasure without undue delay when one of its grounds applies: the data is no longer necessary for its purpose; consent is withdrawn and no other legal ground remains; an objection succeeds; processing was unlawful; Union or Member State law requires erasure; or the data was collected in relation to an Article 8(1) offer of information society services.
Article 17(3) exceptions are case-specific, not blanket excuses. Continued processing may be necessary for freedom of expression and information, a legal obligation or public task, specified public-health grounds, protected archiving or research, or legal claims. Record the provision relied on, why continued processing is necessary, the retained categories, permitted uses, access restrictions, and the event that ends the exception.
Capture the request date, requester identity check, data categories in scope, and systems searched.
Map each data category to an Article 17 ground, an Article 17(3) exception, another documented outcome, or a no-data-found result.
Do not turn one valid exception into permission to keep every category or continue every use. Separate erasable data from data that must remain restricted for the supported purpose.
Where the controller made the data public and must erase it, record the reasonable steps taken, considering available technology and implementation cost, to inform controllers processing links, copies, or replications.
Keep the response rationale separate from the deleted personal data so evidence does not recreate the data that should be erased.
Coordinate routine expiry with active rights requests
Article 12 requires controllers to provide information on action taken on Articles 15 to 22 requests without undue delay and in any event within one month, with a possible two-month extension where necessary because of complexity and request numbers. EDPB access guidance says controllers should have measures that prevent short-lived data from being erased while an access request is being handled. That operational hold should be limited to the request and released when it no longer has a supported purpose.
Restriction is not erasure. When Article 18 applies, the controller generally stores the data but limits other processing to the grounds in Article 18(2). Article 19 requires the controller to communicate rectification, erasure, or restriction to each recipient unless this is impossible or involves disproportionate effort, and to identify those recipients to the data subject on request.
Does GDPR set a single retention period for all personal data?
No. Article 5(1)(e) requires identifiable personal data to be kept no longer than necessary for its purpose, and Articles 13 to 15 require the period or criteria to be communicated in relevant notices and access responses. The actual period depends on the purpose, data category, applicable law, contracts, and documented necessity; Article 30 asks controllers to record envisaged erasure time limits where possible.
Can a team erase all evidence after completing an Article 17 request?
Not automatically. Personal data covered by a valid erasure ground must be erased without undue delay unless an Article 17(3) exception applies. A controller may keep a limited request record only when it has a legal basis and defined retention rule for that record. It should keep only what is necessary to show the request date, scope, decision, response, and recipient action without recreating the erased data.
Does pseudonymising data end the GDPR retention duty?
No. Pseudonymised data remains personal data when the person can be identified using additional information. It still needs a purpose, legal basis, retention rule, security controls, and an expiry action. Data falls outside the GDPR only when it is anonymous under Recital 26's identifiability test.
Add a narrowly scoped rights-hold or restriction status for data that would otherwise expire while an active request is handled.
Link each data category to the systems and teams that must search for access or erasure requests.
Track recipient notification status after rectification, erasure, or restriction, including the supported reason when notification is impossible or disproportionate.
Record reasons for not taking action and the complaint or judicial-remedy information supplied to the requester when applicable.
Define a separate retention rule and legal basis for request records. Keep only what is necessary to show timing, scope, decision, response, and recipient action.
Evidence should show that the retention rule is tied to the purpose, has an accountable owner, and operates across systems and processors. It should not preserve unnecessary personal-data copies merely to prove the data once existed.
Use the processing activity as the control point. When it changes, review the retention schedule, Article 30 entry, privacy notice, processor instructions, rights workflow, and deletion-test evidence together.
Current retention schedule version, approver, owner, and change history.
RoPA entry showing purpose, data-subject categories, personal-data categories, recipients, transfers where applicable, erasure time limits where possible, and security measures.
Deletion or anonymisation control evidence for live systems, logs, exports, backups, archives, processors, and sub-processors, including what happens if a backup is restored.
Article 17 decision logs with data categories, grounds, exceptions, action taken, response date, and recipient-notification outcome.
Exception and hold register limited to the categories and uses supported by the recorded basis, with approval and release triggers.
Review evidence showing that expired data and obsolete processing are removed from active use and that any retained accountability record has its own rule.
This artifact helps connect RoPA fields, erasure requests, and deletion evidence
Sorena can help map processing activities to retention triggers, Article 17 decision records, recipient notifications, and evidence needed to show that deletion controls work.
DPC guidance supports maintaining the RoPA as a living, self-contained record with process owners, granular data categories, and retention information.