Skip to content

Scanning guide

How to name and file scanned documents automatically

A scanned document is only useful if someone can find it next year. Good names and a sensible folder layout do most of that work, and capture software can apply them to every page you scan. Here is how to design a naming convention, which folder layouts hold up, and how to handle duplicates, missing values and private numbers.

Practical guide · 9 min read · Updated October 2026

The short answer

Build each name from values on the document, in an order that sorts: date first in year-month-day form, then a unique number, with the type and the company or person as folders. Decide what to write when a value is missing, never overwrite on a clash, and keep tax IDs and account numbers out of names. CapturePoint 6 does all of this from the naming window in Job setup › Export.

Seven rules for names that work

  1. 01

    Build names from what is on the page

    Vendor, date, invoice number, employee, claim number. Never the scanner’s counter, like Scan_0042.

  2. 02

    Put dates in year-month-day order

    2026-09-14 sorts correctly in every folder view and means the same thing in every country.

  3. 03

    Lead with what people sort by

    Inside a vendor folder, people want date order, so the date comes first and the invoice number second.

  4. 04

    Include one value that makes the name unique

    An invoice number, claim number or policy number. A date and a name alone will clash.

  5. 05

    Use one separator, the same way everywhere

    An underscore between parts and hyphens inside dates is easy to read: 2026-09-14_INV-20931.pdf.

  6. 06

    Keep names short and plain

    Letters, numbers, spaces, hyphens and underscores. Windows refuses \ / : * ? " < > |, and long paths break in SharePoint and some backup tools.

  7. 07

    Write the convention down

    One line per document type, kept where the team can see it. It settles arguments and helps new staff.

Naming recipes by document type

These follow the rules above and match the kinds of documents in CapturePoint’s ready-made sample jobs. Adjust the fields to the ones your documents actually carry.

Example values are fictional, in the style of the sample documents.
DocumentFoldersFile nameExample
InvoiceInvoice › VendorInvoice date_Invoice numberInvoice/Northwind Supply/2026-09-14_INV-20931.pdf
Purchase orderPurchase order › VendorPO date_PO numberPurchase order/Northbridge Office Supply/2026-09-02_PO-1031.pdf
Vendor statementStatement › VendorStatement dateStatement/Northwind Supply/2026-09-30.pdf
W-9W-9Vendor_Signed dateW-9/Northwind Supply_2026-01-12.pdf
HR documentEmployee name › Document typeDocument date_Document typeLee, Jordan/Offer letter/2026-08-01_Offer letter.pdf
Bill of ladingBill of lading › Ship year-monthShip date_BOL numberBill of lading/2026-09/2026-09-12_BOL-55810.pdf
Insurance claimClaims › Claim numberDocument type_Date receivedClaims/CLM-104233/Adjuster report_2026-09-18.pdf

Folders: how deep, and by what

Folders should match how people browse, not every way they might search. Two levels is usually right, and three is as deep as most teams need.

LayoutWorks well whenWatch out for
Type › Vendor (or customer, employee)People look things up by who the document is from or aboutHundreds of small folders for occasional vendors
Type › Year-monthHigh volume, processed and audited by periodFinding one vendor across months needs search
Type › Vendor › YearLarge vendors with many documents a yearThree levels plus long names can hit path limits
Type only, rich file namesDestinations with columns or details that carry the values (SharePoint, Google Drive)Very large folders are slow to browse

When the destination can store values beside the file, let it. SharePoint columns and Google Drive details make a shallow tree just as easy to search. These guides show how:

Duplicates and missing values

When two documents want the same name

It happens more than you would think: a vendor reuses invoice numbers each year, a statement arrives twice, two people scan the same letter. The worst outcome is a silent overwrite. Good software keeps both and makes the clash visible. CapturePoint 6 never overwrites: the second file gets -2, the third -3, in your folder and in every cloud destination.

A -2 is also a signal. If you see many of them, add a value that tells the documents apart, usually the date, or check whether the same paper is being scanned twice.

When a value is missing

Some invoices have no purchase order; some forms leave a date blank. Without a rule, you get names like 2026-09-14__.pdf. Decide a word for each part, such as No PO or Undated, so the gap is obvious and still sorts sensibly. In CapturePoint 6 each field in a name has an If it’s missing, use setting for this, and the naming window shows how many of the shown examples used it.

Private values in names

File names travel further than file contents. They appear in folder listings, search results, sync logs, email notifications and sharing links, to anyone who can see the folder, even people who cannot open the file. Keep these out of names:

  • Social Security numbers and other tax IDs, including vendor tax IDs from W-9s
  • Bank account, card and loan numbers
  • Patient, member and policy numbers when a name alone would identify a person
  • Dates of birth

Use an internal reference instead: a vendor name, an employee ID that means nothing outside your company, a claim number. If you need the private value for search, store it inside the document’s data or in a column with restricted access, not in the name.

Setting it up in CapturePoint 6

CapturePoint 6 reads each document on your Windows PC, recognizes its type and pulls out the fields. Naming lives in Job setup › Export. The Export card shows the destination folder and a live preview of where a real document will land.

The destination, AP Scans on a file server, with the preview path AP Scans › Invoice › Northwind Supply › 2026-09-14_INV-20931.pdf. Show all types gives one example for each document type.
  1. 01

    Open the naming window

    Choose Change how files are named. On the left is every document type, with a search box; a small dot marks a type whose naming differs from the suggestion.

  2. 02

    Arrange the lanes

    A type’s naming is a set of lanes: Folder 1, Folder 2 and so on, then File name, which always ends in .pdf. Each lane is a row of parts. A part is the Document type, any captured field, or Your own words. Use + to add a part and Add a folder level for another folder.

  3. 03

    Pick separators and date formats

    Between parts choose an underscore, a hyphen, a space or nothing. For a date, Show as offers year, year-month, full date, month or day, so a folder can be 2026-09 while the file name carries the full date.

  4. 04

    Set the missing-value wording

    Click a field’s chip and fill in If it’s missing, use. Each chip shows its value on a real document as you work.

  5. 05

    Check the results on the right

    The right side draws the folder tree recent documents would produce and says how many current documents can be named, whether any two would share a name, and whether a private value is in the names.

  6. 06

    Reuse it across types

    Copy Invoice naming to… gives other types the same folders and file name. Suggest naming proposes naming for every type, and Back to the suggestion undoes your changes to one type. Suggestions replace your naming only after you accept.

  7. 07

    Use it and save

    Use this naming returns to Export with the change; Save in Job setup stores it.

A new job that has nowhere to send files starts with Documents › CapturePoint Exports and the job’s name, and you can change it to a network share. Each document arrives as a searchable PDF (optionally archive-grade PDF/A), a data file with every captured value and a text file.

Each document arrives as a searchable PDF, a data file and a text file, all named by the same naming.

The same naming carries to SharePoint or OneDrive, Google Drive, Dropbox and Nucleus One when you add them under Other destinations. If the job exports to Content Central instead, Content Central’s catalog decides folders and names; the help article on folder and file building explains how to set that up there.

Checklist before you go live

  • Every document type has a written naming line, and the team agrees on it.
  • Dates are in year-month-day form, wherever they appear.
  • Each file name includes a value that makes it unique.
  • Every field in a name has missing-value wording.
  • No tax IDs, account numbers or dates of birth in names.
  • Folder depth is two or three levels at most.
  • The preview shows no unexpected shared names.
  • A test batch has been scanned and the results opened by someone other than the person who set it up.

Naming only works on documents that are already separate. If you scan stacks, read batch scanning and document separation too.

For setup and troubleshooting articles, see the CapturePoint help library.

Questions about naming scanned files

What is the best date format for file names?

Year-month-day with hyphens, such as 2026-09-14. It sorts into date order in every folder view and cannot be misread as day-month or month-day.

Which characters should I avoid in file names?

Windows does not allow \ / : * ? " < > | in names, and SharePoint and OneDrive have a few rules of their own. Stick to letters, numbers, spaces, hyphens and underscores. CapturePoint 6 cleans names to each destination’s rules before it files them.

What happens when two documents get the same name?

In CapturePoint 6 the second file is saved with -2 at the end, the third with -3, and so on. Nothing is overwritten. The naming window warns you before you save when two current documents would share a name.

Can different document types be named differently?

Yes. Each document type has its own folders and file name in CapturePoint 6. Copy a type’s naming to other types to start them from the same layout; fields a type does not have are left out of its copy.

Does the naming apply to SharePoint and Google Drive too?

Yes. SharePoint and OneDrive, Google Drive, Dropbox and Nucleus One receive each document with the job’s own folders and file name under the folder you choose. Content Central is the exception: its catalog decides folders and file names.

Next step

Let every scan name and file itself.

Download CapturePoint 6, open a sample job and design your naming in the preview before you scan a page.

Windows 10 or 11, 64-bit. The free trial starts the first time you open it, with no form to fill in.