Quick answer
A download center is more than a folder of PDFs.
Each file belongs to a resource page, a category, a product, a department, a language or a custom field — and all of that is what makes the library usable. When a large set changes, the point is to keep every one of those relationships and change only the files underneath.
Upload the new documents, match them to the resources they belong to, check the ambiguous ones, keep replacements and new publications apart, and check the library and the public links afterwards.
Why a library is not the Media Library
The Media Library manages files.
A resource library adds meaning on top. Something like:
Safety Manual — Product A — English — Current
might actually be a custom post, with a file field, a category, an audience field and a link to a product.
So the thing you are replacing is not an upload path. It is a resource that happens to have a file in it.
How libraries are organised
By department, document type, product, topic, audience, language, date, region, or membership level — usually several of those at once.
A replacement has to leave all of it exactly as it was.
Replace, or publish new?
Replace when the resource is stable and only the file is new.
Current Supplier Handbook
keeps its title, category and description. Only the PDF changes, and everything that links to it, searches for it or filters to it carries on working.
Publish new when the old document should stay, the date or version means something, or the library is an archive.
2026 Annual Report
must not overwrite the 2025 one.
The files are usually in custom fields
Document libraries store files in ACF fields, in theme-specific data, or in a plugin’s own fields.
A tool that only looks at the Media Library will not find them at all — it will report a clean run and change nothing.
WP Bulk Replace Pro reads ACF file, image and gallery fields, including the ones inside repeater rows, and WooCommerce product images, galleries and customer downloads.
Keep the information around the file
A replacement must not quietly wipe the title, description, category, audience, related product, sort order or access rule.
That is the entire point of replacing rather than uploading: the resource stays itself, and only the document inside it changes.
Example: a 300-resource library refresh
An association maintains 300 resources. An annual review produces:
- 180 revised current PDFs
- 40 resources being retired
- 50 files that are unchanged
- 30 new publications.
A safe run: upload the 210 incoming files, match the 180 revisions to their existing resources, identify the 30 new ones, exclude the retired items from replacement, leave the 50 unchanged resources alone entirely, back up, replace only the approved revisions, and check the metadata, the access rules and the direct links.
That is a library job, not a bulk upload.
Search and categories
If your library search indexes the title, category, excerpt and filename, replacing the file keeps all of that intact.
Uploading a new document instead usually loses every one of those relationships, and the resource stops turning up where people expect it.
Files used by more than one resource
One file can be attached to several library entries.
If that is deliberate, one replacement updates all of them, which is efficient.
If it is not, the batch should stop that row until the library has been tidied up rather than changing entries you did not review.
Who can see what
A download center usually has public, member-only and staff-only resources.
WP Bulk Replace does not provide those permissions. Because the resource stays where it is, whatever does provide them carries on working untouched.
Retired is not the same as replaced
A resource can be current, superseded, retired or archived.
Replacing is only one of those. If readers need to know a document was superseded, overwriting it hides exactly the thing they needed to see.
Check the search afterwards
After a big replacement, search the library for a few of the resources.
If your search indexes the resource’s own information, keeping that resource intact is what preserves it. If it indexes file contents or filenames, you may need to reindex — which varies by site, so check rather than assume.
Do not reorganise and replace at once
If you are also restructuring categories, do that as a separate step.
Changing the taxonomy and the files in one action makes it much harder to work out what went wrong, and much harder to undo cleanly.
Summary
A resource library is a set of content objects that happen to have files attached.
The safe model: keep the resource, replace the file inside it, publish genuinely new resources separately, and check the metadata and access afterwards.
Related: Files in custom fields · Bulk replace files · Association document updates