Skip to main content

Handle Skipped Documents and OCR Risk

Blast Audit can only match and view documents reliably after OCR is complete. If OCR is interrupted or weak, you may see skipped documents, blocked previews, or low-quality matching output.

Written by William Karkegi
Updated over 2 weeks ago

Blast Audit can only match and view documents reliably after OCR is complete. If OCR is interrupted or weak, you may see skipped documents, blocked previews, or low-quality matching output.

Who this is for

  • Add-in users importing and matching PDFs

  • Teams working with scans, statements, invoices, or image-heavy files

  • Anyone seeing skipped documents or OCR-related warnings

Before you begin

  • You may still see Copilot Audit in a few older labels.

  • Work from the add-in pages Documents and Document Matching.

  • Keep Excel open while OCR runs.

  • Do not navigate away from Documents during first-time processing.

How to do it

  1. Check OCR Status in Documents.

Every file should move to Complete before you rely on it.

Files marked Not Started or Processing are not ready for matching or viewing.

  1. Respect the OCR guard.

During processing, Blast Audit warns Please stay on this page and keep Excel open. If you leave now, the OCR result may be lost and you'll need to run OCR again.

Treat that warning literally.

  1. Re-run OCR for weak or interrupted files.

Use Re-run OCR on documents that look incomplete, badly scanned, or were imported during a disrupted session.

  1. Confirm the viewer can open the file.

If the viewer says OCR must be completed before viewing, finish OCR first.

A file that cannot open cleanly in the viewer is not a good candidate for matching yet.

  1. Read skipped-document summaries after matching.

If a run ends with `Some documents were skipped: ...`, note the filenames and reasons exactly as shown.

That summary is your fastest clue about which files need cleanup or removal from the batch.

  1. Reduce the batch size.

Re-run the job with Choose specific documents or a clean folder only.

This tells you whether one bad file is poisoning a larger run.

  1. Replace low-quality source files when possible.

Cleaner PDFs beat post-processing every time.

Re-export from the source system if you can instead of reusing screenshots or low-resolution scans.

Expected result

  • Key files show OCR Status = Complete

  • Documents open normally in the viewer

  • Matching runs without a skipped-document summary

  • Re-runs produce more stable, higher-confidence output

Avoid this

  • Assuming import finished means OCR finished

  • Leaving Documents while OCR is still processing

  • Mixing clean exports and weak scans in the same batch

  • Ignoring skipped-document names after the run

If it still doesn't work

  • List the exact skipped filenames and messages

  • Note whether the issue affects one file or every file in the batch

  • Check whether the same documents fail in the viewer too

  • If you need support, include the OCR statuses and skipped summary in your support escalation packet

Practice with sample files

  • Search for Practice Troubleshooting with Sample Documents in this collection.

  • Use the attached PDFs and CSVs there to test this workflow before you use live client files.

Read next

  • Upload First Documents

  • Use Document Viewer and Create Snips

  • Folder Strategy for Clean Matching

  • Run First Document Matching

  • Fix "No Matches" or Low Confidence

  • Build a Support Escalation Packet

  • Practice Troubleshooting with Sample Documents

Did this answer your question?