How Do You Build a Long-Term Digital Preservation Strategy for Your DAM?

The gap between a DAM and a preservation system is most visible the moment something goes wrong. A storage migration completes, the project is signed off, and six months later someone opens a batch of files and finds them corrupted - checksums were never generated at ingest, so there is no way to know whether the corruption happened before or during the migration, or which files are affected. The investigation starts from zero.

This is not a hypothetical. It is a pattern that repeats across DAM programmes of every size, in every sector, when preservation is treated as a feature of the platform rather than a discipline that the platform must support. The platform rarely fails on its own terms. It was not designed to do what preservation requires.

The internationally recognised framework for understanding that gap is ISO 14721:2012, the Open Archival Information System (OAIS) Reference Model. Section 4.1 defines six functional entities a preservation system must implement: Ingest (with documented provenance), Archival Storage (with integrity verification), Data Management, Administration, Preservation Planning (monitoring format obsolescence and planning migrations), and Access. Most commercial DAM platforms handle Access and Data Management well. Ingest with documented provenance, Preservation Planning, and formal Archival Storage with integrity verification are rarely implemented out of the box.

The practical question is not whether your DAM is OAIS-compliant - most are not, and that is not necessarily a problem. The question is: which of those six functions does your DAM perform, which does it not, and what policies and processes do you need to fill the gaps? The following three sections address the gaps that cause the most operational damage: missing fixity checking, no format policy, and no retention schedule.

Ask a DAM manager whether their platform performs fixity checking and the most common answer is a pause, followed by 'I think so.' That pause is the problem. Fixity checking is binary: either you have a checksum for every file, generated at ingest and verified on a schedule, or you do not. 'I think so' means you do not have a record of it, which means you cannot rely on it.

Fixity checking works by generating a mathematical fingerprint - a checksum using an algorithm such as MD5 or SHA-256 - at the moment a file is ingested, and storing that value alongside the file. Periodically, the system recalculates the checksum and compares it to the stored value. A mismatch means the file has changed. That change could be malicious, accidental, or the result of storage hardware degrading gradually over time - what archivists call bit rot.

The Digital Preservation Coalition's Digital Preservation Handbook recommends fixity checks at three specific points: at ingest, after any storage migration, and on a regular schedule (at minimum annually for archival content). SHA-256 is preferred over MD5 for new implementations because MD5 has known collision vulnerabilities, but either algorithm is vastly better than no checking at all.

The storage migration point is the one most commonly skipped. Migrations are treated as completed when the file counts match and spot-checks look correct. But spot-checks do not find corruption in files that were not spot-checked, and file counts do not detect partial corruption within a file. A fixity check run against every file immediately after migration - comparing post-migration checksums against pre-migration checksums - is the only reliable way to confirm the migration succeeded.

Practical steps for DAM managers:

• Check whether your DAM platform generates and stores checksums at ingest. If it does not, raise this with your vendor in writing so the response is on record.
• If your platform supports scheduled fixity checking, enable it and document the schedule in your governance record.
• If your platform does not support fixity checking natively, determine whether your storage infrastructure provides integrity verification at the storage layer - for example, AWS S3 provides ETag checksums and optional object integrity checks via the x-amz-checksum header - and document exactly what that covers and what it does not.
• Record fixity check results and any failures. A fixity failure is a data integrity incident. Investigate the cause, attempt recovery from backup, and document the outcome. A log of fixity failures is also evidence of due diligence if a rights or regulatory dispute arises later.

Format obsolescence is not a theoretical risk. Organisations that stored video in early QuickTime formats, documents in early Microsoft Works formats, or images in formats tied to now-discontinued software have experienced it directly: files that open as blank screens, or that require specialist recovery tools that themselves require legacy operating systems to run. The assets exist. They are just no longer readable.

The Library of Congress Sustainability of Digital Formats resource is the most authoritative public reference for assessing format risk. It evaluates formats across five dimensions: disclosure (is the format specification publicly documented?), adoption (how widely is it used?), transparency (is the content human-readable without specialist software?), self-documentation (does the format carry its own metadata?), and external dependencies (does it require proprietary software to render?). Each format entry includes a sustainability factor rating. These ratings change over time as software ecosystems shift, which is why the resource is updated and why your format policy needs a review cadence, not just a launch date.

A practical format policy for a DAM programme addresses three things:

1. Preferred formats for ingest. Define which formats are accepted as preservation masters by asset type. For still images, TIFF and PNG have high Library of Congress sustainability ratings and no proprietary dependencies. For documents, PDF/A (standardised as ISO 19005) is explicitly designed for long-term preservation and is the format required by many regulated industries for archival records. For video, the choice is more complex; the Library of Congress moving image format guidance is the correct starting point, and the decision should involve whoever manages your storage infrastructure, not just the DAM team.
2. Formats requiring migration. Audit your existing library for formats that carry significant obsolescence risk. Build a migration plan - not necessarily immediate, but scheduled and resourced. A format risk register, even a simple spreadsheet listing format, estimated risk level, asset count, and planned migration date, is more useful than a policy document that does not name specific formats.
3. Review cadence. Review your format policy at least every two years, or when a major software platform in your ecosystem announces a significant change. Assign the review to a named person. A policy with no named reviewer is a policy that will not be reviewed.

The most common retention policy in practice is no policy at all. Assets accumulate. Storage grows. Nobody has the authority or the mandate to delete anything, because deletion feels risky and nobody wants to be the person who deleted the asset that turned out to be needed. The result is a library where current, approved assets are buried under years of superseded versions, expired campaign materials, and assets whose rights have long since lapsed.

That buried library is not just a findability problem. In many jurisdictions, retaining personal data - images of identifiable individuals, for example - beyond its legitimate purpose is a compliance violation under data protection law. Retaining assets whose rights have expired creates licensing exposure. And a library with no disposal process has no mechanism to control storage costs as the asset volume grows.

A retention and disposal schedule assigns a retention period to each asset category, defines what happens at the end of that period, and names who is responsible for applying the decision. The following questions need answers before you can write the schedule:

• Which asset types are subject to regulatory retention requirements, and what are those periods? Financial services records in the UK are subject to FCA retention rules that vary by record type. Pharmaceutical promotional materials in the US are subject to FDA requirements under 21 CFR Part 11. If you operate across jurisdictions, the requirements may conflict and legal input is needed.
• Which assets have ongoing brand or historical value that warrants indefinite retention, and who makes that determination? Not everything does. The criteria should be written down, not left to individual judgment at disposal time.
• Which assets are operational and can be disposed of after a defined period of inactivity - for example, campaign assets more than three years past their last use date?
• Who has the authority to approve disposal of an asset category, and how is that approval recorded? Deletion is not the end of the record; the fact of deletion, the date, and the authority for it should be logged.

Review the schedule annually and update it when business context changes. A new product line, a regulatory change, or a significant acquisition all affect what needs to be kept and for how long. The review owner should be named in the document.

This article covers the foundational practices of digital preservation as they apply to commercial DAM deployments: the OAIS functional model (ISO 14721:2012 section 4.1), fixity checking, format policy, and retention scheduling. It is written for DAM managers and digital operations leads who need to assess and improve their current programme, not for organisations building a formal archival repository from scratch.

It does not cover full OAIS implementation, which is relevant for national broadcasters, cultural heritage institutions, and regulated industries with mandatory records management requirements. Those organisations should engage a digital preservation specialist alongside their DAM programme. It does not address specific DAM vendor capabilities for preservation, which vary significantly and change with product releases. It does not cover the specific technical requirements of audiovisual preservation, which is a specialist field with its own guidance from IASA (the International Association of Sound and Audiovisual Archives).

The DAM Republic earns revenue through sponsorship and directory listings on this site. This article is produced independently by the TdR editorial team and is not influenced by commercial relationships with any DAM vendor or storage provider.

What is the difference between a DAM and a digital preservation system?

A DAM is designed for access, distribution, and workflow. A digital preservation system is designed for long-term integrity: ensuring assets remain readable, authentic, and findable over decades. ISO 14721:2012 section 4.1 defines six functional entities a preservation system must implement. Most commercial DAM platforms handle Access and Data Management well, but do not implement Preservation Planning or formal Ingest with documented provenance.

What is fixity checking and why does it matter for DAM?

Fixity checking verifies that a stored file has not changed since ingest by generating a checksum (MD5 or SHA-256) at ingest and recalculating it periodically. The Digital Preservation Coalition's Handbook recommends checks at ingest, after every storage migration, and at minimum annually for archival content. Silent file corruption is more common than most practitioners expect, particularly after storage migrations that are signed off on file count alone.

Which file formats are safest for long-term digital preservation?

Open, widely documented formats with no proprietary dependencies carry the lowest obsolescence risk. The Library of Congress Sustainability of Digital Formats resource evaluates formats across five dimensions including disclosure and external dependencies. For still images, TIFF and PNG are well-supported. For documents, PDF/A (ISO 19005) is the archival standard. For video, consult the Library of Congress moving image format guidance for your specific asset types.

What is the OAIS Reference Model and does my organisation need to follow it?

The Open Archival Information System (OAIS) Reference Model (ISO 14721:2012) is the internationally recognised conceptual framework for digital preservation. Section 4.1 defines six functional entities and the Submission, Archival, and Dissemination Information Packages that flow through them. Organisations with formal archival obligations should use OAIS section 4.1 as a design checklist when evaluating whether their DAM infrastructure is sufficient or whether a separate preservation repository is needed.

How long should digital assets be retained?

Retention periods depend on asset type, jurisdiction, and business context. Financial services records in the UK are subject to FCA retention rules; pharmaceutical promotional materials in the US are subject to FDA requirements under 21 CFR Part 11. Marketing assets may warrant three to seven years after last use. Retention decisions must be documented in a schedule, reviewed annually, and applied consistently. Keeping everything indefinitely creates storage costs, legal exposure, and findability problems.