Rebuild a PDF page order visually before you send it

A scanned document arrives with the appendix in the middle, three blank separator sheets, and the signature page at the front. Fixing that by describing the problem to someone else takes longer than fixing it yourself, provided you can see the pages while you work.

Work visually, in a single pass

Page numbers are a poor interface for restructuring a document. Move page 14 to position 3 and every number you memorised is now wrong, which is how people end up exporting four times and still getting it wrong.

The page organiser shows thumbnails you can drag, select, rotate, and remove. That matters for one specific reason: every change should happen before a single export. Each export-reopen-fix cycle creates another file on your disk, and a week later nobody can tell which of the five similarly named PDFs is the final one.

A workable sequence:

1. Scan the thumbnails end to end before touching anything, and note what is wrong.
2. Delete the obvious junk first — blank separators, duplicate scans, cover sheets — so there is less to rearrange.
3. Fix orientation next, while the pages are still where you expect them.
4. Then reorder, moving large blocks before individual pages.
5. Export once, reopen the result, and read it in order.

Nothing degrades when you move pages

Reordering, deleting, and rotating are structural operations. The tool copies page objects and their resources into a new page tree; it does not re-render or re-encode anything.

So there is no quality cost, no matter how many times you rearrange. Text stays selectable, vector artwork stays sharp at any zoom, and the file size stays close to the original minus whatever you removed. This is the opposite of the image-based workaround — exporting pages, editing them as pictures, rebuilding a PDF — which loses the text layer and inflates the file to achieve the same visible outcome.

Deleting pages actually removes them

Worth stating plainly, because it is a common worry: removing a page from the page tree and exporting a new document means the page content is not carried into the output. This is not the same as hiding a page or covering it, and it is a genuine reduction in what the recipient receives.

That makes page deletion a legitimate minimisation step. If a report contains an internal annex the client should not see, removing those pages before sending is both simpler and safer than redacting their contents.

Two cautions:

• Keep the original. Deletion is only reversible if you still have the source.
• Deleting a page does not remove document-level metadata, attachments, or bookmark entries that referenced it. Those need their own pass.

What to re-check after the page set changes

Several PDF features are bound to page positions, and rearranging breaks the binding:

Feature	What to verify
Bookmarks and outline	Entries may now point at the wrong page or a deleted one
Internal links and cross-references	"See page 12" is now wrong in both the text and the link target
Printed page numbers	Numbers drawn on the page do not renumber themselves
Form fields	Duplicated pages produce duplicate field names that share one value
Digital signatures	Changing the page set invalidates the expectations a signature was made under

Printed page numbers are the one people miss most often. If the document had pagination stamped onto the pages, reordering leaves it scrambled, and the fix is to renumber after restructuring rather than before.

When a different tool is the better fit

• You want a subset, not a rearrangement. Extract the range with the split tool instead of deleting everything else.
• You are assembling several sources. Combine them with the merge tool first, then organise the result once, as described in the merging guide.
• Only orientation is wrong. Rotation alone is faster and needs no restructuring at all.

The general principle holds across all of these: decide the final shape of the document first, make every change in one editing session, and export a single file you can be confident about.