Standardize a field with an approved list

Upload a list of approved values, such as your vendor names, and Paige replaces what it reads on each document with the matching value from that list. Choose how strict matching is, what happens when nothing matches, how to break ties, and how to keep the list current.

Catalog administrators15 minutesVerified 2026-10-07

Before you begin

A company admin login in Paige (only company admins see the Settings gear)
A CSV file of your approved values with a header row, for example a vendor list exported from your accounting system

At the end of this guide, a field like Vendor Name comes out of Paige spelled exactly the way your own records spell it, because Paige swaps what it read on the document for the matching value from a list you provide.

Invoices from one vendor rarely agree with each other. One shows the brand from the logo, another the legal name on the remit-to block, a third has a scanning slip like Acrne Office Supply. Without a list, Paige exports what each page printed, so your accounting system or file share sees three vendors. Standardize the field with your vendor list (with the brand listed as another spelling) and all three export as Acme Office Supply, so the documents carry one name and file into the same folder downstream.

How it works

Paige reads the field first, the same way it always does, following the field's AI Extraction Guidance.
Paige then compares what it read with your list. An exact value or a listed alternate spelling matches directly. For names, Paige can also correct a close variant, but only when one approved value is clearly the best match.
When Paige identifies one clear approved value, it stores that value. When it can't, the field follows the no-match choice you set for it.

Note

Fixing a vendor name by hand during review corrects that one document; Paige does not learn from those corrections. The list is how you make every future document consistent.

Before you start: prepare the file

Save it as a CSV. The first row must be column headers, for example Vendor Name.
Put one approved value per row, spelled exactly the way you want it exported.
Optional: other spellings. When a vendor goes by more than one name (a brand in the logo and a different company on the remit-to block), add a column such as Also Known As and put the other name on that vendor's row. Use more than one column if a vendor has several.
Optional: more columns. Columns like Vendor Number or GL Code can fill other fields from the same row later. See Fill other fields from the same list row.

Important

In a file with exactly two columns, Paige treats the second column as other spellings of the first. If your second column is something else, such as a vendor number, add a third column or clear it from the other spellings afterward (see Manage your lists below).

Standardize the field

Open Settings (the gear at the bottom of the left menu). Under Scan Jobs, select the scan job.

Choose Document Types, then click the pencil (Edit Document Type) on the document type.

In Fields, select the field, for example Vendor Name. Its details panel opens on the right, titled Vendor Name Details.

Click Standardize this field with a list…, the dashed box under the field's format settings. The line under it reads "Replace what AI reads with an approved value from your data."

For a column of a line-items field, use Standardize this column with a list… under that column in Line Item Columns.

5
Choose where the values come from:
- Use a list I already have: pick the list. Lists are shared across your company, so the vendor list you set up for one scan job works in all of them.
- Upload a file: drop your CSV on the box or click it to choose the file.

6
Tell Paige which column holds the approved values:
- One column: Paige uses it.
- Two columns: Paige shows Approved value: with the first column and Also accept spellings from: with the second. Click Change if they're the wrong way round.
- Three or more columns: under Which column holds the approved Vendor Name?, click the column. To add alternate spelling columns, click Add other spellings we should accept and pick them. Leave columns like Vendor Number alone.
What you should see: a preview of the first rows of your file, so you can confirm Paige read the columns correctly.

Click Use this.

What you should see: the line Standardized against followed by the list name and its count of approved values, with Try a value, Edit, and Remove beside it. In the Fields list, the field now carries a Standardizes tag.

Click OK to save the document type.

Important

The field isn't standardized until you click OK. Cancel discards the change to the field (a list you uploaded stays in your company's lists for later).

Paige names a new list after the column that holds the approved values. You can rename it in Lists & Standardization.

Check it with Try a value

On the Standardized against line, click Try a value.
Type a spelling the way it appears on a document, then click Run test.

What you should see: Match found: with the approved value Paige would store, and The matched row showing every column of that row. If Paige can't pick one value, you see No clear approved value, followed by what the field will do instead.

Choose how strict matching is

Under the Standardized against line, Paige states each setting as a plain sentence. Click Change on a sentence to see its choices.

Setting	Choices	Applies to
What kind of values are in this list?	Names and text: corrects likely typos only when one approved value is clearly better than the rest. Best for vendor names, product names, and descriptions. Codes and identifiers: uses an exact value or a listed alternate spelling only, and never guesses when a character differs. Best for vendor numbers, account numbers, and SKUs.	Every field that uses this list
How close must a match be? (names and text only)	Strict (85% similar), Balanced (recommended) (75%), Lenient (60%), or Always use the closest value (no minimum)	Every field that uses this list
If Paige cannot identify one clear approved value	Keep the value found by AI (best when new vendors can still appear), or Leave Vendor Name blank (best when only approved values may be stored)	This field only

Paige suggests the kind from your values when you upload the file. Check it: a list of codes treated as names could correct a code to a different, similar code.

Warning

Always use the closest value replaces every value with something from your list, even a vendor who isn't on it yet. A new vendor's invoices would then export under an existing vendor's name. Use it only when the field must always hold a value from the list.

Break ties when several rows match

When more than one row could be the match, for example two vendors with the same name in different cities, tell Paige which document field settles it. Your file needs a column to compare against, such as City.

Find the line that starts No tie-breakers. and click Add.
Under Tie-break qualifiers, click Add a tie-breaker.
Pick the Document field, for example Vendor Address, and the List column to compare it with, for example City. Add more pairs if you need them.
Click OK to save the document type.

What you should see: the line now reads "When several rows match, Paige compares Vendor Address with City and keeps the row that agrees best." The values don't need to match exactly. When you use Try a value on this field, it also asks for the tie-break values the document would show.

What changes, and what doesn't

Guidance still decides what Paige reads. If the logo and the remit-to block name different companies, the field's AI Extraction Guidance decides which one Paige reads; the list then corrects its spelling. If Paige keeps reading the wrong text, fix the guidance: Write extraction guidance the AI can follow.
New documents only. The list applies to documents Paige processes after you save. Documents already processed or exported keep the values they had.
One way per field. A field is either standardized or filled from another field, not both. Setting up a fill on a standardized field replaces its standardization.
It replaces Force Selection from List. Standardizing replaces the older Force Selection from List option under Advanced Settings, which is hidden while the field is standardized. If the field used that older option, you may see a Legacy list line: the old file is kept on the field but no longer used.

Turn it off or switch lists

Switch to a different list: click Edit on the Standardized against line, choose the other list, and click Use this.
Stop standardizing: click Remove and confirm with Stop. Paige goes back to keeping exactly what AI reads.

Either way, click OK to save the document type.

Manage your lists

Open Settings and choose Lists & Standardization, just below Company Settings. Every list in your company appears under Lists; select one to see its details.

To	Do this
Update the list when your vendor master changes	Click Replace file…, choose the new CSV, and click Replace file. The new file must still contain every column the list uses.
Get a copy of the current file	Click Download.
See what's in it	Click Sample rows.
Change which column holds approved values or other spellings	On the line that starts Approved values come from, click Change, then adjust Approved values come from and Other spellings we should accept.
See every field that uses it	Check Where this list is used: one row per scan job, document type, and field. Click a row to open that field.
Use it on another field	Click Add to a field…, choose Standardize a field, pick the Scan job, Document type, and Field to standardize, then click Attach.
Rename it	Edit the name at the top and click Rename.
Start a new list	Click New list and upload a CSV.
Delete it	Click the trash icon. It's available only when no field uses the list; remove each use under Where this list is used first. Deleting is permanent.

Settings for the kind of values and match closeness live on the list, so changing them here changes them for every field that uses it.

Note

Right after you upload or replace a file, Paige needs a short time to prepare the new rows. The file line shows preparing… until it's ready.

When it doesn't work

Symptom	Where to look
The field is blank	Its no-match choice is Leave Vendor Name blank and Paige couldn't identify one clear value. Run Try a value with the spelling from the document, then add that spelling to your list, or switch the field to Keep the value found by AI.
A value kept the document's spelling	Paige couldn't identify one clear value, so it kept the AI value. Add the spelling as an approved value or other spelling. For a names list, a looser closeness setting may also catch it.
The wrong approved value was used	Matching is too loose (Lenient or Always use the closest value), or two rows look alike. Move to Balanced (recommended) or Strict, or add a tie-breaker.
Codes are being "corrected" to similar codes	The list's kind is Names and text. Change it to Codes and identifiers.
Older documents still show the old spelling	Expected. The list applies to documents processed after you save.
The trash icon is greyed out	The list is still in use. Remove each use under Where this list is used.

Still stuck?

Contact Ademero support and include:

The scan job, document type, and field name
The list name
The value as it appears on the document
What Try a value shows for that value
The field's no-match choice and the list's kind of values