Seller Profit Guard

Listing Cost Library audit checklist and change log

Last updated: 2026-07-28

Written and reviewed by Seller Profit Guard Editorial Team.

A defensible SKU cost-library audit records identity, units, source evidence, formulas, required layers, effective dates, imports, downstream joins, privacy, approvals, and rollback. Every checked item links to evidence. Completion requires material unknowns and collisions to be resolved or formally held; clean boxes alone are not proof.

Listing Cost Library audit checklist and change log evidence workflow from source and units through SKU version and seller decision
A reusable cost record keeps identity, units, source, version, and decision connected.

What belongs in the identity and unit checklist?

Verify every active SKU is nonblank, unique after approved normalization, attached to the intended product or variation, and protected from historical reuse. Record listing and variation mapping, old-to-new aliases, active dates, currency, purchase unit, usable quantity, consumption unit, conversion, and quantity basis.

Hand-check representative pack, weight, length, batch, hour, minute, item, order, and bundle conversions. Confirm imported decimals and delimiters, block conflicting duplicate rows, and preserve the pre-import backup.

SKU cost audit trail from source and unit through record version join approval and rollback
Every control must point to reproducible evidence.
ControlRequired evidenceFailure response
IdentityUnique SKU and mappingResolve collision
UnitsConversion hand-checkCorrect denominator
LayersIncluded/excluded scopeMeasure or hold
SourceLineage and dateRefresh evidence
VersionEffective intervalRepair history
JoinDummy and representative matchFix mapping

What belongs in the source and formula checklist?

For materials, labor, packaging, fulfillment, allocation, and risk, record source, currency, unit, formula, effective date, owner, and review trigger. Recalculate each subtotal and the final total. Verify waste, yield, quantity, and labor are not counted twice.

Classify every layer as observed, measured, allocated, estimated, stale, or unknown. Confirm zeros have evidence and negatives use explicit credit or recovery treatment. Check that marketplace and order costs are not hidden inside a production layer and then subtracted again downstream.

Audit matrix connecting cost layers to source unit formula version owner and decision
A total passes only when every material layer has a visible control.

How should changes and approvals be logged?

Record timestamp, SKU set, old and new values, source change, unit change, formula change, reason, affected periods, fixtures, expected results, operator, reviewer, approval, rollback, and follow-up. Preserve the prior library and report fingerprints.

Review a bounded sample of unchanged and changed records, plus missing, duplicate, stale, and variation-sensitive cases. Approval states the exact scope and does not certify profit, tax, accounting, employment, legal, or marketplace compliance.

How do you design an audit sample that can find defects?

Define the population before choosing rows: active SKU versions, versions used by the review period, imported source records, or sold-item joins. Record population size, selection rule, exclusions, and review date. Include a deterministic random sample for broad coverage and a risk sample for newly added, high-value, high-volume, multi-currency, renamed, bundled, stale, manually overridden, or variation-sensitive records. Do not choose only familiar bestsellers or records that already passed a dashboard.

Trace each selected item in both directions. Starting from the SKU record, reproduce source quantity, conversion, layer subtotal, effective interval, and total, then confirm a dummy or private controlled transaction selects that version. Starting from a transaction reference, find the one valid SKU version and reconstruct its layers. Record population and sample error counts separately; a passing sample provides bounded assurance for that procedure and date, not proof that every untested row is correct.

Add deliberate negative fixtures outside private data: a blank identifier, a normalization collision, a pack-versus-each error, overlapping version dates, missing labor, stale shipping evidence, and an unmatched variation. The validator must reject or classify each fixture as expected. If a previously detected defect no longer triggers, investigate the control before celebrating a cleaner report.

Sample stratumWhy inspect itReperformance
Random active recordsBroad library coverageSource through total
Changed versionsUpdate and interval riskOld/new selection
Variation extremesAverages can conceal lossOption-specific join
Negative fixturesValidator sensitivityExpected rejection
Historical ordersCurrent-cost rewrite riskDate-effective match

What should the final audit evidence package contain?

Keep a compact manifest containing audit scope, population counts, sampling method, reviewed control version, library fingerprints before and after review, source-register summary, exception register, fixture results, recalculation worksheet, version-overlap test, join-test summary, reviewer questions, dispositions, approval boundary, and rollback location. Reference private evidence by controlled identifier or path; do not copy invoices, buyer details, supplier terms, or raw order exports into a public SEO report.

Separate findings from management actions. A finding states the observed condition, affected population, evidence, and control objective. The response names an owner, correction, validation method, due date, and residual uncertainty. The reviewer then marks the finding open, remediated, accepted within a stated operational boundary, or outside scope. A blank status, deleted row, or reduced warning count is not a disposition.

Finish with a reconciliation of counts and material changes. Explain why active records, unique SKUs, versions, unknown layers, exceptions, and fingerprints changed from the opening snapshot. Retain unanswered reviewer questions independently from the operator queue. This makes the audit reproducible even when the person approving it did not perform the import or cost measurement.

SKU cost audit questions

Can the checklist replace an accountant? No. It is an operational data control.

Should historical costs be changed after an audit? Correct errors through a logged version or correction; preserve prior evidence.

How are inactive SKUs handled? Archive with dates and mappings needed for historical orders.

What blocks approval? Ambiguous identity, invalid units, material unknowns, broken versions, failed joins, or missing rollback.

What proves completion? Reproduced totals, passing controls, resolved or formally held exceptions, and preserved evidence.

What negative evidence should an auditor inspect?

Challenge sudden zero warnings, missing cost categories, reduced row counts, identical costs across unlike variations, unchanged labor after workflow changes, and all records sharing one source date. A clean result can reflect a weakened control.

Compare the current package with the prior approved state and explain material changes. Keep reviewer questions separate from operator exceptions so approval pressure cannot silently close either list.

Define audit criteria before inspecting results: approved identity rules, unit dictionary, required layers, evidence classes, refresh intervals, effective-date policy, import tolerances, fixture expectations, join logic, and approval roles. Freeze the criteria version in the evidence package. Changing a tolerance after seeing failures is a control change that needs its own reason, test, and approval.

Use reviewer independence proportionate to risk. The operator can self-check ordinary additions, while bulk imports, mapping changes, formula changes, corrections to historical intervals, and overrides of material unknowns deserve a second reviewer. Independence does not mean copying private data into a new system; provide controlled access to the source reference and retain only the minimum audit result.

Listing Cost Library audit checklist and change log page-specific cost comparison and control boundary
The comparison isolates the page's decision without presenting a planning estimate as observed cost.

Which evidence supports this SKU cost audit?

A reusable cost record needs seller-owned evidence. Material and purchased-component costs come from invoices, bills of materials, measured usage, scrap, and yield. Labor comes from timed tasks and an explicit loaded hourly rate. Packaging comes from component quantities and purchase cost. Fulfillment comes from label receipts, carrier adjustments, outsourced fulfillment bills, handling, insurance, and documented subsidies. Etsy exports can carry seller-added SKUs and transaction context, but they do not supply the seller's private production economics.

Record source type, supplier or internal process, currency, quantity basis, unit conversion, effective date, operator, and evidence location without copying private invoices into public reports. A checksum or file identifier can protect lineage, but it does not prove the unit basis is correct. Recalculate pack, sheet, roll, batch, hour, minute, weight, and quantity conversions with a public dummy example before importing a large cost library.

Keep observed, allocated, and estimated values distinguishable. A current invoice is observed evidence; a monthly subscription divided by forecast orders is an allocation; expected scrap, return loss, or future shipping is an estimate. Store the formula and review trigger with every non-observed amount. Unknown remains unknown until measured; zero requires evidence that the cost is genuinely absent.

Privacy and commercial sensitivity for SKU cost audit

A cost library does not need buyer names, delivery addresses, private messages, payment credentials, marketplace passwords, or raw order history. It may contain supplier pricing, labor rates, production time, margins, sourcing notes, and fulfillment terms that are commercially sensitive. Keep the working library in controlled seller storage and use redacted identifiers or public dummy values for documentation and support.

Seller Profit Guard stores the free browser workflow locally by design. A downloaded cost library or profit report is still a private business record. Do not paste real supplier invoices, buyer data, order rows, or SKU economics into public communities, email drafts, or AI prompts. Share the column map, formula, units, and a made-up reproduction case when troubleshooting.

How to apply this SKU cost audit in the SKU Cost Library

Open the SKU Cost Library and add one representative SKU manually before importing a file. Confirm material, packaging, labor minutes, labor rate, shipping, transaction extra, other cost, target margin, notes, currency, source date, and unit basis. Save locally, reload, and export the record. Then import the public template or a controlled working copy and review missing, duplicate, invalid, and suspicious values before accepting the batch.

Use the saved library in Profit Guard with a representative order export or the public dummy sample. Inspect missing-cost warnings and variation joins before interpreting contribution. Run a base case and a page-specific stress case, then record the cost version, result, seller decision, owner, and next review date. A reusable record is valuable only when its units, source, effective date, and SKU join remain traceable.

  1. Preserve the source evidence and current library backup.
  2. Validate the SKU, currency, units, quantity basis, and effective date.
  3. Calculate each cost layer independently before summing.
  4. Import and resolve duplicate, missing, or invalid rows.
  5. Test one handmade and one variation-sensitive scenario.
  6. Record the version, decision, rollback, and next review.

Related resources

Sources and further reading

Related Seller Profit Guard tools

Next step: Open the SKU Cost Library.

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.