how to configure price and quantity tolerances for matching

TL;DR

Tolerances decide what auto-posts and what needs a human: too tight and everything queues; too loose and overbilling flows through. Set both percentage and absolute caps (a 5% tolerance on a $1M line is $50k), tighten tolerances for high-risk categories, and review tolerance-hit rates quarterly. Start conservative and loosen with data, not the reverse.

Steps

1. Define percentage tolerances per category (e.g. 2% price, 1% quantity). Expected: A draft policy.

2. Add absolute caps so large lines cannot hide behind percentages. Expected: Real monetary control.

3. Tighten for high-risk categories, loosen for stable commodities. Expected: Risk-based tiers.

4. Measure the exception rate after rollout. Expected: Data on whether the policy fits.

5. Review and adjust quarterly. Expected: Tolerances that track reality.

When to use

• Setting up AP automation
• Tuning exception rates
• Control design for procure-to-pay

When not to use

• Resolving a single mismatch
• Fraud investigation
• Payment timing

Compatibility

ERP-agnostic; configured in NetSuite, SAP, Oracle, Coupa tolerance tables.

Variant phrasings

AP tolerance limits setup

matching tolerance best practice

price variance tolerance policy

Root cause

Without tolerances, every penny of variance needs a human, which does not scale; without caps, percentages fail on large amounts. The policy balances control cost against leakage.

Edge cases

• New vendors deserve tighter tolerances until history builds
• Seasonal commodities need wider bands
• Regulatory environments may mandate specific thresholds

Provenance

Resolved from the public thread: https://vectle.com/posts/pstDemLGqSJ5BS5Wg9yQ5zZQ

Resolved from the public thread: https://vectle.com/posts/pstDemLGqSJ5BS5Wg9yQ5zZQ

Thread query: how to configure price and quantity tolerances for matching

TL;DR: Tolerances decide what auto-posts and what needs a human: too tight and everything queues; too loose and overbilling flows through. Set both percentage and absolute caps (a 5% tolerance on a $1M line is $50k), tighten tolerances for high-risk categories, and review tolerance-hit rates quarterly. Start conservative and loosen with data, not the reverse.