# OCR Exception Queue Triage Worksheet

> Source: https://netray.co/tools/ocr-exception-queue-triage-worksheet
> Publisher: Netray (netray.co), AI and integration engineering for ERP and manufacturing systems.
> Section: Free Tools

## Short answer

Log each document that falls out of OCR into an exception queue, tag the root cause, and see aging and hours lost per category.

## Category

ai-automation

## Industry

Discrete Manufacturing

## Audience

- AP supervisor
- Quality engineer
- Finance operations

## Tool

**Kind:** worksheet

**Mode:** editable

**Columns:**

#### Document ID

**Key:** doc_id

**Type:** text

#### Document type

**Key:** doc_type

**Type:** select

**Options:**

- Supplier invoice
- Credit note
- Certificate of conformance
- Packing slip
- Mill test report
- Other

#### Source

**Key:** source

**Type:** select

**Options:**

- Email inbox
- Scanner
- Supplier portal
- EDI image
- Mobile photo

#### Root cause

**Key:** cause

**Type:** select

**Options:**

- Low image quality
- New layout
- Missing PO number
- Handwriting
- Multi-page split
- Wrong document type
- Vendor not found
- Other

#### Received

**Key:** received

**Type:** date

#### Age (days)

**Key:** age_days

**Type:** number

**Unit:** days

#### Handling minutes

**Key:** touch_min

**Type:** number

**Unit:** min

#### Loaded rate per hour

**Key:** rate

**Type:** number

**Format:** currency

#### Handling cost

**Key:** cost

**Type:** number

**Expr:** touch_min / 60 * rate

**Format:** currency

#### Owner

**Key:** owner

**Type:** text

#### Status

**Key:** status

**Type:** select

**Options:**

- Open
- Waiting on vendor
- Fixed
- Rejected

**Rows:**

#### Missing PO number

**Doc id:** INV-20931

**Doc type:** Supplier invoice

**Source:** Email inbox

**Received:** 2026-10-05

**Age days:** 14

**Touch min:** 12

**Rate:** 28

**Owner:** R. Meyer

**Status:** Waiting on vendor

#### Low image quality

**Doc id:** INV-20944

**Doc type:** Supplier invoice

**Source:** Scanner

**Received:** 2026-10-08

**Age days:** 11

**Touch min:** 9

**Rate:** 28

**Owner:** L. Ortiz

**Status:** Fixed

#### New layout

**Doc id:** COC-5512

**Doc type:** Certificate of conformance

**Source:** Supplier portal

**Received:** 2026-10-12

**Age days:** 7

**Touch min:** 15

**Rate:** 34

**Owner:** K. Patel

**Status:** Open

#### Handwriting

**Doc id:** MTR-7781

**Doc type:** Mill test report

**Source:** Mobile photo

**Received:** 2026-10-14

**Age days:** 5

**Touch min:** 20

**Rate:** 34

**Owner:** K. Patel

**Status:** Open

#### Wrong document type

**Doc id:** CN-1180

**Doc type:** Credit note

**Source:** Email inbox

**Received:** 2026-10-15

**Age days:** 4

**Touch min:** 6

**Rate:** 28

**Owner:** R. Meyer

**Status:** Fixed

**Totals:**

#### Total handling cost

**Key:** cost

**Agg:** sum

**Format:** currency

#### Total handling minutes

**Key:** touch_min

**Agg:** sum

**Format:** number

#### Oldest item (days)

**Key:** age_days

**Agg:** max

**Format:** integer

**Summary:**

#### Average minutes per exception

**Key:** avg_touch

**Expr:** sum_touch_min / count

**Format:** decimal

**Primary:** true

#### Average age in days

**Key:** avg_age

**Expr:** avg_age_days

**Format:** decimal

**Notes:** Sort by root cause when you export to Excel and pivot count and minutes. The top two causes usually account for most of the lost time.

## Content

**Intro:** Every OCR or invoice capture project ends up with a queue of documents the engine could not post. This worksheet turns that queue into data: what failed, why, who owns it, and what it costs in handling time.

**Sections:**

#### How to use the worksheet

**Body:** Add one row per document that leaves straight-through processing. Record the source channel, pick the root cause from the list, and note handling minutes when the item is closed. Age in days is entered at review time so you can sort the oldest items first. Handling cost is computed from minutes and your loaded hourly rate, so you can compare a cause such as missing PO numbers against one such as handwriting. Export to Excel weekly and keep each export as a snapshot so trends are visible.

#### Choosing root causes that lead to fixes

**Body:** A cause list is only useful if each value points to an owner. Image quality belongs to the scanning or email intake process, new layout belongs to the capture configuration, missing PO belongs to procurement and the supplier, and vendor not found belongs to master data. Keep the list short, under ten values, and add a new value only when a cause repeats at least three times. Avoid a catch-all Other bucket that grows beyond ten percent of rows, because it hides the pattern you are looking for.

**Bullets:**

- Intake problems: fix at the scanner, mailbox or portal
- Model problems: add a template or retrain on the failing layout
- Master data problems: fix the vendor or part record once
- Process problems: change the supplier requirement or PO rule

#### Reading the summary

**Body:** Average handling minutes tells you the cost of a typical exception, and the maximum age shows whether items are being forgotten. If the average is above ten minutes, exceptions are probably being fixed by retyping everything instead of correcting the one failed field. If the oldest item is past your payment terms, you are paying late fees or losing early payment discounts because of the queue, not the supplier.

#### Common mistakes

**Body:** Teams often log only invoices and ignore certificates of conformance and test reports, which are slower to fix because a quality engineer must read them. Another mistake is recording the symptom, such as wrong total, instead of the cause, such as a skewed scan. Finally, do not delete fixed rows. Closed items are the evidence you need to show that a template change or vendor request reduced exceptions.

**Cta:** If your exception queue sits outside the ERP, Netray can build a private agent that reads the queue, proposes the fix and posts after human approval.

**Faq:**

#### What is an OCR exception queue?

**Answer:** It is the list of documents that the capture engine could not read with enough confidence or could not match to ERP data, such as a vendor or purchase order. A person must review and correct each item before it can post.

#### How many exceptions are normal?

**Answer:** It varies widely by document mix. Clean electronic invoices can run with few exceptions, while scanned certificates and photos produce many more. Track your own rate weekly and focus on the trend and the top causes rather than a generic benchmark.

#### Should quality documents share the queue with invoices?

**Answer:** You can log them together for analysis, but give them separate owners. A certificate of conformance needs a quality reviewer, while an invoice needs an AP clerk, and mixing owners slows both.

#### Can I import this into my ERP?

**Answer:** The worksheet exports to Excel and CSV for analysis. It is not an ERP import file. Use it to decide what to fix, then change capture rules or master data in your system.

## Search terms

ocr exception queue, invoice capture exception handling, ocr exception log excel, ap exception queue template, document capture exceptions, invoice exception triage, ocr error root cause log

---
Talk to an engineer: https://netray.co/contact?intent=erp_ai | Machine catalog: https://netray.co/agent/catalog.json
