Harbor & Pine Goods (fictional) · Sources: supplier_a.csv, supplier_b.csv · Destination: Shopify-style import · Mapping HPG-map-v1 · Generated 2026-09-29 08:45
These SKUs are not in import_ready.csv. No value was converted, rounded, guessed or copied from a sibling.
| SKU | Reason | Field | Supplier A | Supplier B | Detail | Decision needed |
|---|---|---|---|---|---|---|
| HPG-BTL-12-CLR | unit conflict | capacity | 12 US fl oz | 350 ml | suppliers state capacity in different units. For reference only: 12 US fl oz = 354.88 ml, so the two statements are not equal. Not converted, rounded or chosen. | Which capacity is printed on the product and should be listed? |
| HPG-MUG-16-CRM | missing required | capacity | (blank) | (blank) | capacity is required for Drinkware and is blank in supplier A; supplier B is also blank. Not inferred from the SKU code or sibling variants. | Confirm the capacity of this SKU. |
| Case | What the files show | Treatment |
|---|---|---|
| True duplicate | Supplier A lists HPG-TWL-OAT (row 10) and hpg-twl-oat (row 22): same SKU once spacing and case are normalised, every other field identical (barcode SYN-29009). | Collapsed to one record; logged in change_log.csv (M1, M2). |
| Look-alike, not a duplicate | HPG-TWL-OAT and HPG-TWL-OLV: same title, price, weight and description. Different SKU, barcode and colour (Oat vs Olive). | Kept as two variants of Linen Tea Towel (M3). |
| Product | Variants ready / total | Held SKUs |
|---|---|---|
| Glass Travel Bottle | 3 / 4 | HPG-BTL-12-CLR |
| Enamel Camp Mug | 3 / 4 | HPG-MUG-16-CRM |
| Linen Tea Towel | 3 / 3 | - |
| Soy Candle | 6 / 6 | - |
| Ceramic Pour-Over Dripper | 2 / 2 | - |
| Wool Throw Blanket | 2 / 2 | - |
| Cotton Canvas Tote | 4 / 4 | - |
| Stoneware Pasta Bowl Set of 4 | 1 / 1 | - |
| Walnut Cutting Board | 2 / 2 | - |
| Beeswax Food Wraps 3-Pack | 1 / 1 | - |
| Check | Result |
|---|---|
| supplier a rows | pass |
| supplier b rows | pass |
| unique skus | pass |
| true duplicates collapsed | pass |
| import ready count | pass |
| held | pass |
| import ready skus | pass |
| product handles in output | pass |
| price values normalised | pass |
| true duplicate sku | pass |
| look-alike pair kept as 2 distinct variants | pass |
Not tested here: loading into a live store. A real job adds the client's own ten-SKU test import before the full file is used.
| Where | SKU | Field | Before | After | Rule |
|---|---|---|---|---|---|
| supplier_a.csv row 22 | HPG-TWL-OAT | sku | hpg-twl-oat | HPG-TWL-OAT | M1: normalised SKU |
| supplier_a.csv row 22 | HPG-TWL-OAT | (whole row) | duplicate of row 10 | (collapsed) | M2: true duplicate: same SKU and identical facts as row 10 |
| supplier_b.csv row 11 | HPG-THR-GRY | retail_price | $89.00 | 89.00 | M6: normalised price |
| supplier_b.csv row 10 | HPG-THR-RST | retail_price | $89.00 | 89.00 | M6: normalised price |
Full rules and field map: mapping_spec.md. File list and fingerprints: handoff_manifest.md. This is a demonstration on fictional data. It shows the checks performed; it is not evidence of results for any real business.