GDPR Deletion Concept: Guide & Retention Periods

Martin Bastius
16.11.2026
5
min.

Use AI to summarize this article

GDPR deletion concept: deletion periods and retention obligations at a glance

What is a GDPR deletion concept?

A deletion concept sets out which personal data your company deletes, when and how – once the purpose no longer applies or statutory retention periods expire. The GDPR does not use the term, but it requires storage limitation (Art. 5(1)(e) GDPR) and that you can demonstrate compliance with it (Art. 5(2) GDPR). A documented deletion concept is the usual way to provide this proof.

Many companies follow an unspoken rule: “We don’t delete data – storage is cheap.” This article shows you how to build a deletion concept based on DIN 66398, which retention periods apply in Germany and how to implement deletion in your systems.

Why do you need a deletion concept?

Personal data may only be stored for as long as the purpose requires. After that, you must delete or anonymize it. Data subjects also have a right to erasure (Art. 17 GDPR), which you can only fulfil if you know where their data is stored.

Keeping data without a legal basis breaches the principles in Art. 5 GDPR. This can lead to fines of up to €20 million or 4% of worldwide annual turnover (Art. 83(5) GDPR). Deletion is particularly difficult for AI systems trained on personal data – read more in our article on the right to be forgotten for AI models.

What does DIN 66398 cover?

DIN 66398 (“Guideline for the development of a deletion concept with derivation of deletion periods for personal data”) is a guideline published by the German Institute for Standardization (DIN). It is not mandatory, but it provides a common terminology that data protection and IT teams both understand.

The core terms of DIN 66398

  1. Data type: data with the same purpose, e.g. application data, payroll data or customer contact data.
  2. Starting point: the event from which the period runs, e.g. end of contract, receipt of a rejection or end of the calendar year of invoicing.
  3. Deletion period: the time between the starting point and deletion, e.g. 6 months, 6 years or 8 years.
  4. Deletion class: several data types with the same starting point and period, grouped together to simplify implementation in IT.

Which retention periods apply in Germany?

The biggest challenge is the conflict between the GDPR’s deletion obligation and statutory retention obligations under commercial and tax law. As long as a legal obligation applies, you must keep the data; there is no right to erasure in that case (Art. 17(3)(b) GDPR). You do, however, restrict processing (blocking) and delete the data once the period expires.

Since 1 January 2025, accounting records such as invoices only have to be kept for 8 instead of 10 years (Germany’s Fourth Bureaucracy Relief Act). Books, inventories and annual financial statements remain at 10 years. Different rules apply to banks, insurers and investment firms.

RecordsStarting pointPeriodBasis
Application documents after rejectionReceipt of the rejection6 months in practiceDeadlines under Section 15(4) AGG and Section 61b ArbGG plus a buffer
Candidate data in a talent poolConsent givenas stated in the consent, typically 1–2 yearsArt. 6(1)(a) GDPR
Commercial and business letters, business emailsEnd of the calendar year6 yearsSection 257 HGB, Section 147 AO
Accounting records, e.g. invoicesEnd of the calendar year8 years (since 2025)Section 257 HGB, Section 147 AO
Books, inventories, annual financial statementsEnd of the calendar year10 yearsSection 257 HGB, Section 147 AO
Contract data after the contract endsEnd of the year the contract ended3 years (standard limitation period)Sections 195, 199 BGB
Server logsCreation of the entry7–30 days in practice, depending on the purposeArt. 6(1)(f) GDPR

Sources: Section 257 HGB and Section 147 AO as amended since 1 January 2025.

How do you create a deletion concept in 5 steps?

An audit-proof deletion concept comes together in five steps.

Step 1: Take stock

Use your record of processing activities (Art. 30 GDPR) as the basis and clarify:

  • which data types you process, e.g. about customers, employees and applicants,
  • in which IT systems, databases, cloud storage and paper archives they are stored.

Step 2: Determine the periods

Assign a purpose and a period to each data type:

  • What is the purpose, and when does it end?
  • Which retention obligations under the HGB or AO still apply afterwards?

Step 3: Create deletion classes

Group data types with the same starting point and period into deletion classes to keep the number of rules manageable.

  • Deletion class A: application data without hiring – starting point: rejection – period: 6 months.
  • Deletion class B: accounting records – starting point: end of the year of origin – period: 8 years.

Step 4: Implement deletion rules technically

Deleting is easy on paper; in complex IT environments it needs clear rules:

  • Automated deletion routines: set up routines in CRM, HR and ERP systems that delete or anonymize data once the period expires.
  • Blocking concept: data you cannot delete yet because of a retention obligation is blocked for normal operations.
  • Backups: backups do not have to be overwritten immediately. It is sufficient if deleted data is automatically deleted again after a restore and backups expire on fixed cycles.
  • Service providers: set out deletion and return of data in the data processing agreement (Art. 28(3)(g) GDPR).

Step 5: Document, train, review

Document the deletion concept in writing, have management approve it and train the relevant departments. Check at least once a year whether the deletion rules actually take effect in your systems.

How does the deletion concept relate to IT security?

A deletion concept does not end with clicking “delete”: data must be removed so that it can no longer be restored. For destroying storage media and files, DIN 66399 sets out protection classes and security levels. Record in your technical and organizational measures how you delete and destroy data, and bind disposal service providers by contract.

Storing less data also means less damage in the event of an attack. Our guide to reporting a data breach shows what to do if it happens.

Is your deletion concept audit-proof? The checklist

  • Record of processing activities used: are all processing activities from the record covered?
  • Starting points defined: does every data type have a clear trigger?
  • Laws considered: are the current HGB and AO periods incorporated?
  • Deletion classes created: are there clear deletion classes and deletion rules?
  • IT implementation done: are deletion and blocking routines running in CRM, HR and ERP systems?
  • Backups covered: is there a documented rule for backups?
  • Service providers included: are cloud and disposal providers covered by data processing agreements?

For particularly high-risk processing, also check whether a data protection impact assessment is required (Art. 35 GDPR).

Conclusion

A deletion concept is not a bureaucratic exercise but the foundation of good data hygiene: it cuts storage costs, limits the damage from cyberattacks and shows supervisory authorities that you take storage limitation seriously.

With clear deletion classes based on DIN 66398, up-to-date HGB and AO periods and automated deletion routines, you make your deletion concept audit-proof and keep it current.

Want to set clear deletion periods and rules for your company? With the external data protection officer from heyData, you create your deletion concept, implement deletion rules and document everything in one place.

Take the first step: Book a free demo now

FAQ

Is a data deletion concept mandatory for every company?

In effect, yes. The GDPR does not use the term “deletion concept”, but it requires storage limitation (Art. 5(1)(e) GDPR) and that you can demonstrate compliance with it (Art. 5(2) GDPR). A documented deletion concept is the most practical way to provide this proof, and supervisory authorities regularly ask for it during audits.

How should you handle legacy data from before you had a deletion concept?

With a one-off cleanup. Assign existing data to your new deletion classes, check whether a retention obligation still applies, and delete or anonymize everything else promptly. Document the cleanup so you can prove it during an audit.

Is it enough to move files to the operating system’s recycle bin?

No. Files in the recycle bin can be restored with little effort and therefore do not count as deleted. Deletion within the meaning of the GDPR requires that the data can no longer be restored with reasonable effort, for example through secure overwriting or destruction of the storage medium.

How long do you have to keep invoices in Germany?

Eight years since 1 January 2025. Germany’s Fourth Bureaucracy Relief Act shortened the retention period for accounting records such as invoices in Section 147 AO and Section 257 HGB from ten to eight years. The period starts at the end of the calendar year in which the record was created. Books, inventories and annual financial statements remain at ten years.

What is the difference between deleting and blocking data?

Deleted data is gone for good, while blocked data remains stored but may no longer be used. Blocking is the right approach as long as a statutory retention obligation applies: you keep the data only for that purpose and delete it once the period expires. The GDPR calls this restriction of processing (Art. 4(3) GDPR).

Published
16.11.2026
Last updated
16.11.2026
Martin Bastius
Co-Founder & CLO

More articles

Porträt eines lächelnden Mannes mit kurzem dunklem Haar und Bart vor grauem Hintergrund.
View all articles
Compliance in Practice
2/2/23

What Are Security Controls in Companies?

What Are Security Controls in Companies?
Compliance in Practice
9/25/24

Is Your DNA Safe? Genetic Testing Risks and How to Protect Your Data

Is Your DNA Safe? Genetic Testing Risks and How to Protect Your Data
Compliance in Practice
8/20/26

Managing Multi-Framework Compliance: GDPR, ISO 27001 & NIS2 efficiently and without duplicate work

Managing Multi-Framework Compliance: GDPR, ISO 27001 & NIS2 efficiently and without duplicate work
Porträt eines lächelnden Mannes mit kurzem dunklem Haar und Bart vor grauem Hintergrund.
Discover all stories