How to Prepare a PDF Bundle for Court or Tender Submission
Fourteen separate files have to arrive as one document, in a defined order, numbered continuously, by a deadline. A court bundle, a tender response, a grant application, a planning submission — the context differs but the mechanics are identical, and so are the mistakes.
The single most expensive one is doing the steps in the wrong order.
Before anything: read the requirement
Bundles are one of the few document tasks where somebody else has written down what “correct” means, and where getting it wrong can mean rejection regardless of the content.
Requirements vary by jurisdiction, by court, by procuring authority, and by funding body — so there is no universal specification and anyone who tells you otherwise is guessing. What is common is that a requirement exists, and that it usually covers some combination of:
- whether the submission is one file or several
- a maximum file size
- how pages must be numbered, and whether numbering runs continuously across the whole bundle
- whether an index or table of contents is required, and where it sits
- whether the document must be searchable
- naming conventions for the file itself
Read it before you assemble anything. Five minutes at the start saves rebuilding at the end, and rebuilding is what happens when you discover on the deadline that numbering had to be continuous.
The order of operations
This is the whole guide in one block. Every step depends on the ones before it.
1. Gather and prepare each source document
2. Merge into one file
3. Fix the page order
4. Clean up (rotate, crop, delete)
5. Make it searchable if required
6. Add page numbers ← must come after order is final
7. Protect if required ← must come last
8. VerifySteps 6 and 7 are the ones people get wrong, and both are unforgiving.
Step 1: prepare each document before combining
It is tempting to merge everything immediately and sort it out afterwards. Resist that. Problems are far cheaper to fix in a four-page source file than in a two-hundred-page bundle.
For each document, check it opens, check the pages are the right way up, and check whether it is a scan or a digital file. If a source arrived as a photograph or a fax, deal with it now — see why a PDF is not searchable for how to tell.
If any source is password-protected, remove the protection now with Unlock PDF using the password you hold. Encrypted files generally cannot be merged, and discovering that mid-assembly is a needless interruption.
Step 2: merge into one file
Merge PDF combines the sources. Add them in roughly the right order at this stage — approximately right is fine, since the next step exists to fix it, but starting close reduces the work.
A practical detail worth setting up before you begin: name your source files so they sort correctly. 01-claim.pdf, 02-defence.pdf, 03-exhibit-a.pdf will list in order in any file picker, while claim.pdf, defence.pdf, exhibit A.pdf will not. This costs nothing at the start and saves reordering later.
Everything runs in your browser, so a bundle containing privileged or commercially sensitive material is assembled on your own machine — see how to merge PDF files without uploading.
Step 3: fix the order
Organize PDF gives you the whole bundle as thumbnails. Drag anything that landed in the wrong place, delete blank separator sheets that came through from scanning, and remove duplicate copies of the same document — bundles assembled from several people’s folders almost always contain at least one.
Check the boundaries between documents rather than reading every page. If each source starts on a recognisable first page, verifying that those pages appear in the right sequence catches structural errors quickly. Reordering PDF pages covers the mechanics, including the interleaving problem if any source came from a two-pass duplex scan.
Finish this step completely. Everything after it assumes the order is final.
Step 4: clean up
Still in the same view: rotate any page that came through sideways, and crop scanner edges if the scans are messy. Neither affects order, so both are safe here.
Do rotation before the next step if the bundle needs to be searchable — recognition performs poorly on pages that are not upright, as covered in how to rotate PDF pages.
Step 5: make it searchable, if required
An increasing number of submission rules require a searchable bundle, and it is worth doing even where it is optional. A judge, an evaluator, or an assessor who can search your document finds your argument; one who cannot has to hunt for it.
If any part of the bundle is scanned, run OCR PDF to add a text layer behind the page images. The pages look identical afterwards — the same scans, unchanged — but the content becomes searchable and selectable.
Two caveats. Recognition is not perfect, and accuracy drops on poor scans, handwriting, and complex tables — how OCR makes scanned PDFs searchable covers what to expect. And OCR on a large bundle takes real time, because it runs on your own hardware rather than a server queue. Start it before you need it.
Step 6: number the pages — and only now
Here is the rule that decides whether you assemble once or twice.
Page numbers are stamped according to the order at the moment you apply them. Move a single page afterwards and its number moves with it, so the sequence breaks and the numbering actively misleads a reader who is looking for page 47.
Continuous pagination across the entire bundle — rather than each document numbered from 1 — is the most commonly specified requirement and the most commonly missed one. If your fourteen documents each restart at 1, nobody can be directed to a page.
Add Page Numbers handles position, format, and starting number. Two decisions matter:
Format. “Page 12 of 240” is stronger than a bare “12” for anything submitted, because it tells the reader whether the set they are holding is complete.
Position. Outer corners for anything printed and bound; the inner edge disappears into the spine. Keep clear of existing content — many source documents carry their own numbering, and a fresh number landing on top of an old one is worse than either alone. Adding page numbers to a PDF covers how to handle a bundle whose sources are already numbered.
The index problem
If the bundle requires a contents page listing each document and its page number, you have a circular dependency: the index needs the numbers, and the numbers need the final order including the index.
The way through is to build the index as a source document with the page references left blank, place it at the front, number the whole bundle, read off the actual numbers, fill them into the index, and re-merge that one page. Slightly tedious, and far quicker than the alternative of guessing and correcting.
Step 7: protect it, last of all
If the submission has to be encrypted, Protect PDF sets an open password — and this is genuinely the final step, because an encrypted file cannot be merged, reordered, or numbered afterwards.
Do not confuse this with a watermark. A password controls who can open the document; a watermark is a visible label on the page. If a copy needs marking as confidential or as a draft, that is Add Watermark, and it goes on before protection.
Check the requirement before encrypting anything. Some authorities specifically refuse password-protected submissions because their intake systems cannot process them, and an encrypted file arriving at a portal that cannot open it counts as nothing arriving at all.
Step 8: verify before you send
The list that catches nearly every error, in the order worth checking:
- Page count. Does the bundle contain what you expect? A missing exhibit is easier to spot as a wrong total than by reading.
- First and last page. Is the bundle bounded where you intended?
- Numbering start and end. Does the first numbered page show what it should, and does the last match the count?
- Each document boundary. Does each source begin where the index says it does?
- Search. If it should be searchable, search for a word you know is on a scanned page. If nothing is found, OCR did not cover that part.
- Open it fresh. Close everything and open the exported file as a recipient would — not the working copy in a tab you have had open for an hour.
That last one catches more than it should. A file that looks right in a tab you have been editing is not evidence about the file on disk.
Keep the unassembled originals
Keep the sources and the un-numbered, unprotected bundle. Not the finished submission alone.
Late changes happen — an exhibit is added, a document is withdrawn, an authority asks for a different format. Every one of those means renumbering, and renumbering from a clean master takes minutes while unpicking a finished bundle is not something a numbering tool can do at all. Once exported, page numbers are page content like anything else, described by the same ISO 32000-2 page model as the text around them.
A folder holding sources/, bundle-unnumbered.pdf, and bundle-final-v2.pdf is the difference between a twenty-minute revision and an afternoon.
Why this runs in your browser
Submission bundles are, almost by definition, the documents you least want to hand to a third party. Litigation material, commercially sensitive pricing, personal information gathered for an application.
Every step above happens on your own machine. Nothing is uploaded, nothing is queued behind other users, and there is no retention policy to evaluate because there is no copy anywhere to retain. The PDF Association publishes reference material on the format if you want to read further into how these operations work at file level.