Skip to main content

Use Advanced Matching Options

Blast Audit can tune matching column by column. `Advanced Options` are useful when the default run is close but not quite right. They are not the first thing to reach for. Start simple, then add only the rule you actually need. You may stil

Written by William Karkegi
Updated over 2 weeks ago

Blast Audit can tune matching column by column. Advanced Options are useful when the default run is close but not quite right. They are not the first thing to reach for. Start simple, then add only the rule you actually need. You may still see Copilot Audit in a few older labels.

Who this is for

  • Power users refining a matching setup

  • Teams working with messy text, abbreviations, or tolerance-based numeric checks

  • Anyone trying to reduce false positives or false negatives after a baseline run

Before you begin

  • Run one clean setup without advanced tuning first if possible

  • Make sure the document scope is already correct

  • Open step 3 in Document Matching

How to do it

  1. In Finalize your document match, open the advanced settings for the input column you want to tune.

  2. Use Required Field when that column must match for the row to count as valid.

  3. Use Partial Matching when the source text is often truncated or only partly identical.

  4. Use Fuzzy Text Matching when abbreviations, spacing, or minor text differences are common.

  5. Use Threshold for numeric tolerance, for example when small value differences are acceptable.

  6. Use Force match on page / row within table only when the document structure is repetitive and you need to anchor the match more tightly.

  7. Run the job again and compare the output before changing more columns.

Expected result

  • Only a few columns have advanced options turned on

  • Each rule has a clear reason

  • Match quality improves without making the run brittle

  • Low-confidence results and obvious mismatches go down after the rerun

Avoid this

  • Marking too many columns as Required, which causes valid rows to be skipped

  • Turning on fuzzy or partial matching for fields that should be exact, like IDs or exact reference codes

  • Setting a Threshold so wide that unrelated values are accepted

  • Using force-match rules broadly and then getting fewer usable results

  • Changing multiple advanced settings at once, which makes debugging harder

If it still doesn't work

  • If results get worse, remove the newest advanced option and rerun

  • If many rows are skipped, check whether Required Field or force-match rules are too strict

  • If the output still looks noisy, narrow the document scope before tuning more columns

  • If you need help separating setup problems from engine problems, use Fix "No Matches" or Low Confidence

Practice with sample files

  • Search for Practice Matching 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

  • Map Columns Without Errors

  • Save and Reuse Matching Templates

  • Fix No Matches or Low Confidence

  • Fix Engine and Mapping Errors

  • Practice Matching with Sample Documents

Did this answer your question?