Seller Profit Guard

How to set refund vs replacement thresholds

Last updated: 2026-07-31

Written and reviewed by Seller Profit Guard Editorial Team.

Block invalid order, refund, credit, return, replacement, recovery, policy, currency, period, scope, or conflict evidence first. Then compare the lower modeled loss with a seller-owned maximum, review exact ties, expire stale evidence, and keep remedy authority outside the cost threshold.

resolution threshold charter from original order and case evidence to refund and replacement losses
This original diagram explains a non-compensating case-cost decision with synthetic values.

Set structural Block

Gate control 1: Invalid quantities, refund caps, costs, credits, currency, context, period, scope, or conflicts cannot be rescued by low modeled loss. This threshold role prevents favorable cost from compensating for invalid evidence or authority. Record case alias, source state, full-precision value, display value, currency, responsibility, reviewer, and review date.

Test the nearest valid boundary and counterexample. State how a mismatch changes a non-compensating case-cost decision, which evidence resolves it, who owns correction, and whether the packet must be blocked, reviewed, limited, superseded, or restored.

Set the loss maximum

Gate control 2: Define currency amount, owner, purpose, case population, approval, effective date, and expiry. This threshold role prevents favorable cost from compensating for invalid evidence or authority. Record case alias, source state, full-precision value, display value, currency, responsibility, reviewer, and review date.

Test the nearest valid boundary and counterexample. State how a mismatch changes a non-compensating case-cost decision, which evidence resolves it, who owns correction, and whether the packet must be blocked, reviewed, limited, superseded, or restored.

Set tie Review

Gate control 3: Require policy, agreement, timing, inventory, delivery, customer, risk, and authority evidence when costs tie. This threshold role prevents favorable cost from compensating for invalid evidence or authority. Record case alias, source state, full-precision value, display value, currency, responsibility, reviewer, and review date.

Test the nearest valid boundary and counterexample. State how a mismatch changes a non-compensating case-cost decision, which evidence resolves it, who owns correction, and whether the packet must be blocked, reviewed, limited, superseded, or restored.

Set recovery confidence

Gate control 4: Separate received, approved, submitted, eligible, estimated, disputed, denied, stale, and missing credits or claims. This threshold role prevents favorable cost from compensating for invalid evidence or authority. Record case alias, source state, full-precision value, display value, currency, responsibility, reviewer, and review date.

Test the nearest valid boundary and counterexample. State how a mismatch changes a non-compensating case-cost decision, which evidence resolves it, who owns correction, and whether the packet must be blocked, reviewed, limited, superseded, or restored.

resolution threshold charter: set recovery confidence
This original diagram makes a non-compensating case-cost decision reviewable.

Set timing thresholds

Gate control 5: Name refund deadline, return receipt, replacement dispatch, claim submission, credit posting, and case response clocks. This threshold role prevents favorable cost from compensating for invalid evidence or authority. Record case alias, source state, full-precision value, display value, currency, responsibility, reviewer, and review date.

Test the nearest valid boundary and counterexample. State how a mismatch changes a non-compensating case-cost decision, which evidence resolves it, who owns correction, and whether the packet must be blocked, reviewed, limited, superseded, or restored.

Set exposure bands

Gate control 6: Measure affected cases and loss privately without averaging away a critical policy or privacy failure. This threshold role prevents favorable cost from compensating for invalid evidence or authority. Record case alias, source state, full-precision value, display value, currency, responsibility, reviewer, and review date.

Test the nearest valid boundary and counterexample. State how a mismatch changes a non-compensating case-cost decision, which evidence resolves it, who owns correction, and whether the packet must be blocked, reviewed, limited, superseded, or restored.

Set approval roles

Gate control 7: Separate case owner, source owner, calculation reviewer, customer remedy authority, release operator, and rollback authority. This threshold role prevents favorable cost from compensating for invalid evidence or authority. Record case alias, source state, full-precision value, display value, currency, responsibility, reviewer, and review date.

Test the nearest valid boundary and counterexample. State how a mismatch changes a non-compensating case-cost decision, which evidence resolves it, who owns correction, and whether the packet must be blocked, reviewed, limited, superseded, or restored.

Set rollback triggers

Gate control 8: Name formula, privacy, source, schema, routing, accessibility, content, and health failures. This threshold role prevents favorable cost from compensating for invalid evidence or authority. Record case alias, source state, full-precision value, display value, currency, responsibility, reviewer, and review date.

Test the nearest valid boundary and counterexample. State how a mismatch changes a non-compensating case-cost decision, which evidence resolves it, who owns correction, and whether the packet must be blocked, reviewed, limited, superseded, or restored.

Set the lower-loss share maximum

Gate control 9: Divide lower modeled incremental loss by original gross buyer amount and accept a seller-entered boundary from above 0% through 100%. This threshold role prevents favorable cost from compensating for invalid evidence or authority. Record case alias, source state, full-precision value, display value, currency, responsibility, reviewer, and review date.

Test the nearest valid boundary and counterexample. State how a mismatch changes a non-compensating case-cost decision, which evidence resolves it, who owns correction, and whether the packet must be blocked, reviewed, limited, superseded, or restored.

resolution threshold charter: set the lower-loss share maximum
This original diagram makes a non-compensating case-cost decision reviewable.

Set the scenario-difference minimum

Gate control 10: Divide absolute loss difference by the larger absolute scenario loss and require a seller-entered minimum from 0% through 100%. This threshold role prevents favorable cost from compensating for invalid evidence or authority. Record case alias, source state, full-precision value, display value, currency, responsibility, reviewer, and review date.

Test the nearest valid boundary and counterexample. State how a mismatch changes a non-compensating case-cost decision, which evidence resolves it, who owns correction, and whether the packet must be blocked, reviewed, limited, superseded, or restored.

Define threshold equality

Gate control 11: A value exactly equal to the maximum share or minimum difference remains inside its gate; only a strict breach triggers threshold Review. This threshold role prevents favorable cost from compensating for invalid evidence or authority. Record case alias, source state, full-precision value, display value, currency, responsibility, reviewer, and review date.

Test the nearest valid boundary and counterexample. State how a mismatch changes a non-compensating case-cost decision, which evidence resolves it, who owns correction, and whether the packet must be blocked, reviewed, limited, superseded, or restored.

Prevent denominator substitution

Gate control 12: Do not divide loss by payout, contribution, or the smaller scenario, and do not divide the difference by original revenue; those ratios answer different questions. This threshold role prevents favorable cost from compensating for invalid evidence or authority. Record case alias, source state, full-precision value, display value, currency, responsibility, reviewer, and review date.

Test the nearest valid boundary and counterexample. State how a mismatch changes a non-compensating case-cost decision, which evidence resolves it, who owns correction, and whether the packet must be blocked, reviewed, limited, superseded, or restored.

Govern threshold changes

Gate control 13: Record old and new values, owner, rationale, affected case population, approval, expiry, expected status changes, and rollback before editing a boundary. This threshold role prevents favorable cost from compensating for invalid evidence or authority. Record case alias, source state, full-precision value, display value, currency, responsibility, reviewer, and review date.

Test the nearest valid boundary and counterexample. State how a mismatch changes a non-compensating case-cost decision, which evidence resolves it, who owns correction, and whether the packet must be blocked, reviewed, limited, superseded, or restored.

Challenge cost-only gaming

Gate control 14: Reject threshold changes made only to turn a known Review into Ready; require policy, agreement, customer, timing, inventory, coverage, and authority evidence. This threshold role prevents favorable cost from compensating for invalid evidence or authority. Record case alias, source state, full-precision value, display value, currency, responsibility, reviewer, and review date.

Test the nearest valid boundary and counterexample. State how a mismatch changes a non-compensating case-cost decision, which evidence resolves it, who owns correction, and whether the packet must be blocked, reviewed, limited, superseded, or restored.

resolution threshold charter: challenge cost-only gaming
This original diagram makes a non-compensating case-cost decision reviewable.

Use the resolution threshold charter with one protected case packet

Open the calculator only after the resolution threshold charter identifies order alias, original economics, refund cash, return cost, processing, credits, recovered inventory, replacement quantity and costs, recovery, buyer payment, absolute target, lower-loss-share maximum, scenario-difference minimum, source review date, nine evidence confirmations, currency, policy and case context, month, complete scope, and issues.

Save original contribution, both incremental losses, both remaining contributions, lower scenario, absolute difference, normalized difference rate, lower-loss share, threshold headroom, decision, and issues at full precision. Preserve every version when policy, posting, inspection, claim, or invoice evidence changes.

Separate cost review from remedy authority before a non-compensating case-cost decision

Use the model to compare declared seller costs only. Route buyer rights, shop obligations, platform eligibility, case instructions, agreement, consumer law, coverage, timing, and actual remedy to the authorized evidence and decision process.

A lower modeled loss never permits an inaccurate refund, unsupported refusal, delayed obligation, unwanted replacement, or customer-facing promise.

Version expected, approved, and received recovery

Keep fee credits, inventory recovery, carrier claims, insurance, platform coverage, supplier credits, and buyer payments in separate evidence states with amount, currency, source, date, owner, and dispute status.

Expected recovery is not received cash. Unknown, denied, expired, disputed, or missing recovery is not zero and cannot silently retain Ready.

Protect buyer, message, payment, and case records

Public examples are synthetic. Keep names, addresses, emails, messages, tracking, case records, payment data, claims, accounts, invoices, credentials, photos, and raw orders outside the resolution threshold charter.

Use aliases, aggregates, redacted pointers, access controls, and approved fixtures. An authorized reviewer should reproduce totals without public personal data.

Release, observe, correct, and restore the resolution asset

Before release, preserve narrow local and remote backups and a rollback identifier. Run typecheck, focused and full tests, integration, build, content and duplicate audits, SEO, routes, sources, images, links, mobile, keyboard, privacy, and restore checks.

After release, verify status, canonical, indexability, schema, answer blocks, images, hub discovery, strict 404, sitemap policy, events, and production scenarios. Record Day 0/7/14/28 evidence and restore on formula, privacy, source, accessibility, routing, content, or health regression.

Build a boundary grid

Test caps, zero, target equality, exact tie, recovery changes, currency, context, period, scope, and conflicts.

Store expected status, first issue, owner, and meaning.

Build severity bands

Separate a low-dollar estimate variance from wrong customer remedy, privacy exposure, formula error, or widespread policy defect.

Money cannot compensate for invalid authority.

Build expiry and reopening rules

Reopen when case, agreement, policy, fee credit, inspection, shipping, claim, or final invoice changes.

Preserve every prior version.

Noncompensation

Prove favorable cost cannot rescue invalid evidence. Deep check 1 stores case and order aliases, source versions, protected pointers, full-precision transformations, output, decision, authority, reviewer, expiry, and correction route. Another reviewer must not guess refund scope, recovery state, replacement grain, or formula sign.

Find the closest counterexample and state why it invalidates or narrows a non-compensating case-cost decision. Preserve competing policy, agreement, platform, payment, return, inventory, replacement, claim, fee, target, or scope evidence as named scenarios rather than blending uncertainty.

Equality

Define exact behavior at target and tie. Deep check 2 stores case and order aliases, source versions, protected pointers, full-precision transformations, output, decision, authority, reviewer, expiry, and correction route. Another reviewer must not guess refund scope, recovery state, replacement grain, or formula sign.

Find the closest counterexample and state why it invalidates or narrows a non-compensating case-cost decision. Preserve competing policy, agreement, platform, payment, return, inventory, replacement, claim, fee, target, or scope evidence as named scenarios rather than blending uncertainty.

Confidence

Separate expected from received recovery. Deep check 3 stores case and order aliases, source versions, protected pointers, full-precision transformations, output, decision, authority, reviewer, expiry, and correction route. Another reviewer must not guess refund scope, recovery state, replacement grain, or formula sign.

Find the closest counterexample and state why it invalidates or narrows a non-compensating case-cost decision. Preserve competing policy, agreement, platform, payment, return, inventory, replacement, claim, fee, target, or scope evidence as named scenarios rather than blending uncertainty.

Authority

Keep remedy decision outside the calculator. Deep check 4 stores case and order aliases, source versions, protected pointers, full-precision transformations, output, decision, authority, reviewer, expiry, and correction route. Another reviewer must not guess refund scope, recovery state, replacement grain, or formula sign.

Find the closest counterexample and state why it invalidates or narrows a non-compensating case-cost decision. Preserve competing policy, agreement, platform, payment, return, inventory, replacement, claim, fee, target, or scope evidence as named scenarios rather than blending uncertainty.

Rollback

Tie each critical failure to a tested restore. Deep check 5 stores case and order aliases, source versions, protected pointers, full-precision transformations, output, decision, authority, reviewer, expiry, and correction route. Another reviewer must not guess refund scope, recovery state, replacement grain, or formula sign.

Find the closest counterexample and state why it invalidates or narrows a non-compensating case-cost decision. Preserve competing policy, agreement, platform, payment, return, inventory, replacement, claim, fee, target, or scope evidence as named scenarios rather than blending uncertainty.

Sources and further reading

Related Seller Profit Guard tools

Next step: Open the Refund vs Replacement Calculator.

This is operational planning help, not tax, accounting, legal, financial, or platform-policy advice. Review the Terms and disclaimer, and verify current platform rules and fee assumptions before changing prices.