Every DAM Migration Inherits the Old Mess: Clean Up Metadata Before You Move
- Cleanup is usually out of scope before the move and deprioritised after it.
- Six months later, the old search problems are still there and the project team has moved on.
- Run metadata cleanup in parallel with the technical migration, reviewed by experts.
Why doesn’t a new DAM fix old metadata?
Moving to a new DAM platform can feel like a fresh start. It rarely is.
Migrations carry legacy metadata forward: inconsistencies, duplicate terms, gaps and all. Cleaning it up before the move is usually out of scope. Cleaning it up afterwards gets deprioritised once the new system is technically live.
What happens after a migration with dirty metadata?
The pattern is familiar. A new platform runs on the same tangled taxonomy as the old one, with the added confusion of assets that migrated with broken or partial fields.
Six months later, the search problems that were supposed to disappear are still there, and the project team has moved on.
For DAM vendorsFor SaaS DAM vendors, this is where churn starts: the customer blames the platform for a problem that came with them.
When should you clean up metadata in a DAM migration?
Treat metadata cleanup as its own workstream, running in parallel with the technical migration rather than after it.
| Stage | What to clean | Who should review |
|---|---|---|
| Before the move | Retire unused fields, merge duplicate terms, agree the target vocabulary | DAM admin and taxonomy owner |
| During the move | Map old values to new ones; flag assets with empty required fields | Subject experts by collection |
| Before go-live | Validate the most-searched collections and AI tags added in the new system | Experts in the loop, peer-checked |
| After go-live | Review zero-result searches and new uploads | Rotating team, weekly |
Who should do the cleanup?
Ideally a distributed team rather than one overworked administrator. Someone has to review, correct and standardise tags before or during the move, not after the ribbon-cutting.
The best reviewers are the people who know the content. They also catch the legacy assumptions baked into old tags, the ones nobody questioned because they had always been there.
Expert in the loop (EITL) is a review model in which the people who know the content best (brand managers, product specialists, archivists, rights owners) confirm, correct or reject metadata, including AI-generated tags, and their decisions are checked by peer consensus before they become the permanent record.
How expert-in-the-loop review works in TagQuest
How do you fit cleanup into a migration timeline?
Structure the review as short tasks: one asset, one decision. That makes it realistic to run alongside a migration plan without derailing it. Consensus scoring shows where experts agree, so administrators only handle the disputed cases.
The short version
A new platform is a good moment to fix old habits. It only happens if someone plans for it.
Frequently asked questions
Should you clean metadata before or after a DAM migration?
Before and during. Agree the target vocabulary before the move, map and review values during it, and validate the most-searched collections before go-live.
Why do DAM migrations fail to improve search?
Because the legacy metadata moves with the assets. Inconsistent and missing tags produce the same poor results on the new platform.
How long does a metadata cleanup take?
It depends on library size and how many experts take part. Spreading review across a team in short tasks lets it run in parallel with the migration instead of extending it.