Product Unit Price Audit

Compare supplier prices without confusing a pack price with a per-unit price. The report preserves exact SKUs, shows effective unit-cost changes, and flags declarations that cannot be compared.

Launch price: US$0.10 per completed audit report, covering both snapshots within the limits below. One unit-price-audit-completed event per report; no start fee, per-row fee or separate platform-usage charge. Malformed input fails before the report event. Reports with review findings or incompatible products are completed reports and are charged. A new run creates another report and charge. The Store Pricing tab is authoritative.

Compare previous_json and current_json JSON text arrays. Each product has exactly seven string fields: sku, price, price_basis, pack_quantity, unit, currency, tax_basis. Exact unique nonblank SKUs preserve leading zeros, case and whitespace (128 characters maximum). Other fields: 64 characters maximum.

price and pack_quantity use plain decimal text, maximum 24 characters, 15 integer digits and 8 fractional digits. Positive amounts only. Explicit price_basis is pack or unit; positive pack_quantity is required for both. Currency is uppercase three-letter syntax only. Unit labels must match exactly. tax_basis is inclusive or exclusive. The caller must establish meaning and accuracy of all declarations.

Pack effective unit price equals price divided by pack_quantity; unit pricing uses price directly. Different currency, unit or tax basis is not compared. Calculations use exact fractions; display values use eight decimal places and half-even rounding. Every result is review_required. Missing SKUs mean absent from this snapshot, not discontinued.

Malformed JSON, duplicate properties, invalid Unicode, missing/extra fields, non-string values, blank/duplicate SKUs reject the input. Other invalid declarations produce explicit row-not-comparable findings. Each snapshot: 2 MiB UTF-8, 10000 rows, nesting 16 and 200000 JSON nodes. Report: 4 MiB; split large batches when necessary. No report is written when the cap is exceeded.

No fetching, FX conversion, tax calculation, fuzzy matching, store writes or safety certification. Only submit authorised non-sensitive data; Apify stores input/output.

Apify SDK 4.1.0. Use 256 MiB, timeout 60 seconds, restart-on-error and standby off. Existing dataset reports are retained on restart; no exactly-once concurrency guarantee. The implementation passed 46 shared core/boundary/adapter tests. Prior synthetic cloud output confirmed unchanged pack/unit cost, a 100% unit-cost increase from pack shrinkage, and an excluded currency mismatch. This does not establish production reliability or customer adoption.

Quick start

Run the prefilled Input example. SKU001 changes from AUD30 per pack of 12 to AUD2.50 per unit, so its effective cost is unchanged. Added and missing SKUs remain explicit. For pack shrinkage, compare the same SKU at AUD30 per pack with quantity 12 before and quantity 6 after: unit cost rises from 2.50 to 5.00, a 100% increase.

Who it fits

Purchasing and catalogue workflow builders who already have two structured snapshots and confirmed product identities, units, currency and tax treatment. Use previous_json/current_json in Apify Console or through the API. This does not fetch supplier files, parse Excel/PDF, infer units or change a shop.