The quality of an import is decided before the file is uploaded. A well-shaped spreadsheet produces a short exception queue; a poorly shaped one produces hundreds of ambiguous rows that all trace back to two or three fixable habits.
What the File Needs
There is no single mandatory schema, because supplier exports never share one. What matters is that the file contains enough to identify each product, and that the identifying columns are distinguishable from the descriptive ones.
| Column | Needed? | Notes |
|---|---|---|
| Manufacturer | Strongly recommended | Narrows candidates more than any other single field. |
| Model or part number | Strongly recommended | The primary matching signal. Clean this column first. |
| Product type | Helpful | Disambiguates similar model strings across categories. |
| Price | Optional | Carried through; never used for matching. |
| Condition | Optional | Kept as unit-level commercial data, separate from device facts. |
| Internal SKU | Optional | Keep it in its own column. Never merge it into the model column. |
manufacturer,model,product_type,condition,price,internal_sku
Acme Medical,AM-4200,Infusion pump,Refurbished,4250.00,SKU-11842
Acme Medical,AM-4200XL,Infusion pump,Demo,5100.00,SKU-11843Clean the Identifier Column
- Trim leading and trailing spaces. They are invisible in a spreadsheet and they defeat exact matching.
- Remove internal prefixes and suffixes that your business added, such as a branch code appended to the model.
- Normalize separators. Decide whether the model is 'AM-4200' or 'AM 4200' and apply it consistently.
- Split combined cells. A cell containing 'AM-4200 (refurb, 2019)' is three facts in one field.
- Check for auto-formatting damage, such as model numbers that look like dates or that lost a leading zero.
The Leading-Zero Problem
Import, Then Read the Failures First
Upload the File
GearDock reads the sheet and shows the detected columns before anything is committed.
Map the Columns
Confirm which column is the manufacturer, which is the identifier, and which are descriptive. Mapping is per import, so a differently shaped supplier file does not require changing anything permanent.
Review Validation Results
Look at the rejected and flagged rows before the accepted ones. Failures almost always cluster around a single upstream cause worth fixing at the source.
Resolve Ambiguous Matches in Groups
Sort the ambiguous rows by the pattern causing the ambiguity. One decision usually clears dozens of rows.
Re-Import the Corrected Rows
Fixing the source file and re-importing is normally faster than correcting rows individually, and it improves the file you will use again next quarter.
Do Not Include Patient Information