Seller Profit Guard · How it works · CSV privacy

Missing SKU cost checker

Check synthetic sold-item counts for exact, missing, ambiguous, and stale cost-library joins. Enter two aggregate packets, their SKU or variation mapping context, an unresolved-join threshold, maximum cost age, currency, period, and evidence scope. The browser-local checker returns coverage plus Block, Review, or Ready without uploading orders or cost files.

Maintained by Seller Profit Guard Editorial Team. Last reviewed: 2026-07-31.

Sold item keys pass through normalization alias variation and dated cost records to exact missing ambiguous or stale join decisions
Classify every synthetic sold item exactly once before using costs in downstream margin work.

Start with one sold-item grain

Count one row per sold item or declared line-item event before checking costs. Order totals, payment rows, shipment rows, and listing rows are different grains.

A balanced classification requires every sold line to appear once as exact, missing, ambiguous, or stale.

Keep this intent distinct

Use the CSV column mapper to normalize headers, the duplicate checker to review repeated rows, and this checker to measure cost-library coverage after those controls.

The result does not calculate contribution, accounting profit, inventory value, or tax.

Define an exact usable join

An exact join links one normalized sold-item key to one cost record that is valid on the sale date, in the declared currency, at the required variation grain.

A syntactic SKU match is not enough when the record is expired or applies to another variation.

Define a missing join

A missing join has no cost record after approved SKU normalization, alias lookup, listing fallback, and variation checks are exhausted.

Never substitute a zero cost merely to make the join complete.

Define an ambiguous join

An ambiguous join finds two or more plausible cost records with no approved precedence rule.

Do not average competing records or select the first spreadsheet row.

Define a stale join

A stale join reaches a record whose effective window or approved age has expired before the sold date.

Keep stale separate from missing so the repair owner can update rather than recreate evidence.

Normalize keys conservatively

Trim declared whitespace and apply a versioned case or punctuation rule only when that transformation is approved for the source.

Do not remove characters that distinguish genuine variants.

Preserve renamed-SKU aliases

Map legacy SKU to canonical SKU with effective dates, reason, owner, reviewer, and prior value.

An alias should not rewrite the historical sold export.

Use listing and variation keys

When a SKU is blank or reused, require a documented listing-plus-variation mapping at the same sold-item grain.

A listing-level cost can be wrong when materials, size, or personalization change unit cost.

Version the cost library

Store canonical SKU, variation scope, unit cost, currency, effective-from, effective-to, source, and reviewer version.

A current cost is not automatically valid for an older sale.

Calculate exact coverage

Exact cost coverage equals exact usable joins divided by sold item lines.

Coverage is a data-readiness measure, not a margin metric.

Calculate unresolved joins

Unresolved joins equal missing plus ambiguous plus stale joins.

Report both the count and the reason mix so repair work is actionable.

Balance the classification

Exact, missing, ambiguous, and stale counts must sum exactly to sold item lines for each packet.

An imbalance Blocks interpretation because rows are double-counted or omitted.

Version the cost-coverage contract

Record a nonfuture source-review date covering the evidence month, a policy effective date no later than that review, each packet's oldest usable cost effective date, synthetic sold date, and oldest usable cost age, plus minimum closed evidence duration and a maximum sold-line input bound. Confirm the nine grain, mapping, cost, quarantine, counterexample, restoration, privacy, and authority controls before interpreting coverage.

Invalid governance controls mask every derived join result.

Read Ready

Ready means the synthetic counts balance and unresolved joins are within the declared threshold, with complete evidence and no known conflict.

It clears only the entered coverage packet.

Read Review

Review means the packet is structurally valid but unresolved joins exceed the approved threshold.

Assign each reason class to a named evidence owner before any margin analysis.

Read Block

Block means counts, grain, context, currency, period, source scope, privacy, or restoration evidence is invalid or incomplete.

Repair the join contract before interpreting coverage.

Work a renamed SKU

Preserve the sold SKU, look it up in the dated alias table, resolve one canonical SKU, then join to the cost record valid on the sold date.

If one legacy SKU maps to multiple canonical records for the same date, classify it as ambiguous.

Work an unmapped variation

Use listing ID, variation values, canonical option order, and dated SKU assignment to identify the exact cost record.

Do not fall back to a parent listing cost without a reviewed rule.

Separate blank and unknown values

Blank SKU, unrecognized SKU, deleted listing, changed variation, missing cost row, and expired cost row need different reason codes.

A single 'N/A' bucket hides the repair path.

Control currency

The library currency, converted cost, conversion source, rate date, and rounding convention must be explicit.

This tool reports join coverage and does not convert money.

Control dates

Compare sold date with alias and cost effective windows using one documented timezone and boundary convention.

Month-only records need an explicit first-day or period rule.

Preserve source evidence

Fingerprint protected sold-line, listing, alias, variation, and cost-library sources with versions and filters.

Public examples use invented counts and labels only.

Minimize private data

The checker needs aggregate counts, not buyer names, emails, addresses, order numbers, payment data, credentials, or raw CSV rows.

Keep operational identifiers in authorized systems.

Assign human authority

Name the mapping owner, cost owner, independent reviewer, correction approver, monitoring owner, stop authority, and restoration owner.

Ready cannot authorize overwriting the source library.

Back up before repair

Preserve the sold export, normalized-key version, alias table, variation map, cost library, prior classification, exception log, and tested restoration instructions.

The browser-local public tool creates no backup.

Retest counterexamples

Test full coverage, renamed SKU, missing alias, one-to-many alias, blank SKU, unmapped variation, reused SKU, stale cost, overlapping dates, malformed counts, and open conflicts.

Every public fixture remains synthetic and bounded.

Monitor join drift

Track schema changes, new aliases, variation edits, cost-source updates, recurring reason codes, coverage shifts, and reviewer overrides.

Stop downstream profit automation when coverage degrades unexpectedly.

Do not infer zero cost

A missing join is an evidence gap, not evidence that cost equals zero.

Exclude or quarantine the row according to a documented downstream policy.

Release through ordered gates

Index only after distinct intent, current sources, functioning logic, counterexamples, answer-first content, metadata, schema, diagrams, links, privacy, similarity, mobile, backup, deployment, and live checks pass.

Search signals are measured later and are not required for release.

Keep a change log

Record mapping version, source fingerprints, counts by reason, threshold, decision, owner, correction, reviewer, monitoring trigger, and restoration reference.

Never erase the prior accepted mapping without supersession evidence.

Quarantine unresolved lines downstream

Keep missing, ambiguous, and stale cost joins in a visible exception population until the evidence owner resolves them. If a separate profit workflow excludes these lines, report the excluded count and revenue scope beside every total.

Do not silently calculate margin from the covered subset as though it represented all sold items.

Reconcile after every repair

After adding an alias, variation mapping, or dated cost record, rerun the complete closed-period packet rather than only the corrected line. Confirm that exact coverage increases, the original exception disappears for the intended reason, no new ambiguity is created, and the prior accepted mapping remains restorable.

A local correction is incomplete until population balance and counterexamples still pass.

Sources and further reading

Related Seller Profit Guard tools

Use the interactive tool

Enable JavaScript to open the calculator and process browser-local inputs. The explanatory content and source links remain available without JavaScript.

Related guide: Define sold-item grain, exact, missing, ambiguous and stale cost joins, thresholds, evidence, decisions, monitoring, and restoration.

This tool provides operating estimates, not tax, accounting, legal, financial, or marketplace-policy advice. Verify current official sources and your own records before changing prices or operations.