DAM migration plan

What a DAM migration plan contains

A DAM migration plan is a programme document that sets what moves, in what order, under which rules, and how each batch is proven correct. The plan sits inside phase 2 of the DAM implementation and depends on a frozen metadata schema. Six parts make the plan complete, each with an owner and a sign-off.

The DAM migration checklist turns these six parts into gate items, and the metadata mapping guide covers the field rules in detail.

Inventory and cleanup

The inventory lists every place assets live today, and the cleanup rules decide per source what moves, what archives and what goes. A DAM content audit from the selection phase gives a head start. The migration inventory adds the file-level detail the audit did not need.

Use one inventory row per source:

Cleanup rules work best as written rules applied by script, with exceptions logged by name. Typical rules:

• Delete exact duplicates found by file hash, keeping the copy with the richest metadata or the clearest folder path.
• Keep only the final version of artwork unless the version history has legal or regulatory value.
• Archive outside the DAM any file with no recorded use for a period agreed with each content owner.
• Exclude work files such as temporary exports, thumbnails and system files.
• Keep layered source files such as PSD or INDD packages only where teams will reopen them.

Duplicate detection by hash finds identical files. Near-identical files, such as the same photo in two crops, need a visual similarity check or a manual decision by the content owner.

Assets with unknown rights

Assets with unknown usage rights migrate with a restricted status, or stay out of the DAM until an owner confirms the licence. DAMSelection assessment: an unrestricted asset with unknown rights becomes riskier after migration, because the DAM makes it easier to find and share. The restricted status keeps the asset visible to the rights team and blocks download and publication for everyone else.

Three rules handle the group:

1. Mark every asset without a licence record as "Rights unknown" through the mapping default.
2. Give the legal and rights team a review queue sorted by expected use, starting with assets in active campaigns.
3. Set a review deadline, and move assets still unconfirmed at the deadline to the archive.

The Rights Exposure Estimator sizes the exposure before migration, so the review queue gets realistic staffing.

Migration approaches

A DAM migration follows one of three approaches, and by DAMSelection assessment phased migration by priority suits most mid-sized libraries. The choice depends on volume, the number of sources and how long old and new systems can run side by side.

Coexistence needs one firm rule. From the day the first batch is live, new assets go into the new DAM only. Old systems accept no new uploads from that day, so the delta at cutover stays small.

Batch sequence

The batch sequence migrates high-value, well-described assets first, so users find useful content on their first visit. DAMSelection assessment: the first batch shapes user opinion more than any training session. A first batch of old archive material teaches users that search returns nothing useful.

A typical sequence, highest value first:

1. Brand core: logos, brand guidelines, templates and approved photography.
2. Product images and documents for active products, linked to the PIM by SKU.
3. Current and upcoming campaign assets.
4. Assets of the last two to three years with recorded use.
5. Older material selected for reuse or for legal retention.

Batch size balances speed and control. By DAMSelection rule of thumb, a batch should be small enough to validate in one or two working days, and the first production batch should be smaller than later ones. Increase batch size once two consecutive batches pass validation without manual fixes.

The batch sequence also follows the release waves in Driving DAM user adoption. Each team's most-used assets should be live before that team's training session.

Validation

Validation proves after each batch that every asset arrived intact with the right metadata, and a batch counts as done only when all checks pass. Run automated checks on every asset and manual checks on a sample.

The sample size and the three-filter rule are DAMSelection rules of thumb. The checksum check rests on the Secure Hash Standard: a matching SHA-256 digest shows the file content is unchanged after transfer. Log every failure in an issue log with an owner and a fix date, and re-validate a failed batch before the next batch starts.

Cutover and decommissioning

Cutover switches the organisation to the new DAM as its single source, and it ends when old sources are read-only and old links redirect. Plan cutover as a short, rehearsed sequence:

1. Announce the freeze window and the read-only date to every team, two weeks ahead by DAMSelection rule of thumb.
2. Freeze uploads to the old sources at the start of the window.
3. Run the delta migration for files added or changed since the last batch.
4. Validate the delta batch with the same checks as every other batch.
5. Switch integrations to the new DAM and confirm the first successful syncs.
6. Redirect old asset links with permanent HTTP 301 redirects where the old system serves public URLs.
7. Set old sources to read-only and publish the decommission date.
8. Export and archive what the organisation must retain from the old system before it closes.

A 301 status tells browsers and other clients that the resource has a new permanent address. Websites, emails and partner portals that embed old asset URLs keep working while their owners update links. Keep a redirect map from old URL to new asset ID for as long as the redirects run.

Plan the exit while planning the entry. The export format and ID map from this migration also inform the exit strategy for the new platform.

Worked example: a 250,000-file migration plan

A DAM migration plan for a 250,000-file library shows how cleanup sets scope and batch sizes. Worked example: a mid-sized consumer brand migrates from two shared drives and an old DAM. Shared drive A holds 120,000 files, shared drive B holds 60,000 and the old DAM holds 70,000.

Cleanup runs before any batch:

• Hash-based duplicate detection removes 45,000 identical copies.
• The archive rule, agreed with content owners, moves 75,000 files with no recorded use out of scope.
• The exclusion rule removes 10,000 work files and system files.
• The remaining 120,000 assets migrate, and 9,000 of them carry a restricted status for rights review.

Batches 1 to 5 add up to the 120,000 assets in scope. The plan migrates under half of the original files, and the library users meet on day one contains current, described and rights-checked content.