Migrate away from Beaver Builder
Beaver Builder stores its pages as flat node map in post meta. That single fact decides what a migration can carry across and what it cannot, so it is where this page starts.
What Beaver Builder stores
Beaver Builder keeps a flat map of nodes in post meta, each node carrying its own parent reference. The structure is recovered by following those references rather than by nesting, which makes it one of the cleaner data models to read programmatically.
The part that actually costs time
The flat map is easy to read and easy to get subtly wrong: a node whose parent reference is stale still renders in Beaver but has no defensible position in a nested tree, so the conversion has to decide where it belongs.
Carries across
- Row, column and module structure
- Text and images
- Most module settings
- Saved rows, as reusable parts
Needs a decision, or a rebuild
- Beaver Themer layouts, which sit outside the page
- Modules from add-on packs
- Per-node responsive overrides
Where you can go from Beaver Builder
How the migration actually runs
The agent works on a duplicate of the page, not the live one. The original keeps serving while the converted copy waits for review, and every write captures a before-state so a bad conversion is one call to undo. Nothing publishes without a human approving it.