Supplier file preparation
A checked supplier file for the import template you already use
Two supplier files disagree on units, option names or required attributes. We map them to your store's existing import template, hold back every product with a conflict or a missing fact, and hand you the file with a list of what still needs a decision. Your team keeps control of the import.
Who it is for
Ecommerce agencies onboarding a new supplier into a client store that already has an import template, and merchants with a blocked supplier batch and a named person who owns the product data. It is a bounded batch, not a whole-store migration.
If your importer already handles the files once it is configured, you do not need us.
What you send
- Two supplier files, up to 100 SKUs and ten agreed attributes. Each variant SKU counts toward the 100.
- Your destination template or schema.
- A way to test: the validator or test environment you already use, for a ten-SKU dry run.
- One product-fact approver, and confirmation that we may use the data.
What you get
- A mapping specification: which supplier field goes where, and the rules applied.
- A normalised file in your template's columns.
- Your original files, preserved unchanged.
- A change log of every edit.
- A register of missing and conflicting values: SKU, field, what each supplier said, the decision needed, and whether the row is held back.
- A ten-SKU dry run through your validator or test environment, then checks of the full file.
- Handoff instructions.
- One consolidated correction round. Our errors are corrected.
$1,500
For the first project, up to 100 SKUs. It sets up the mapping.
$500 for each later batch of up to 100 SKUs in the same supplier formats, attributes and approved mapping.
Paid in full, in one payment, before work starts. We send the payment link once the scope is agreed in writing.
Lead time: 15 business days after we have complete inputs and payment.
Tell us about the supplier, the store's import template and the number of SKUs. A redacted sample helps. You get a fixed scope back, or a plain no if it is not a fit.
What it does not include
- Filling in missing product facts. A missing or conflicting value is held back for your approver, never guessed, and an unresolved row is never labelled import-ready.
- Product copywriting or descriptions.
- Running the production import. Your team does that.
- Live inventory changes or platform rebuilds.
- Regulated product advice or product safety certification.
- Original product research at scale, or scraping restricted sources.
- Ongoing emergency support.
- New suppliers, new fields, a new taxonomy or a new output format. Each needs a new quote.
A partial file is delivered only if you agree which rows are excluded and why.
Synthetic example, fictional data
What the checks look like
This example was built on invented data for a fictional home goods supplier: a product master with 30 rows, a distributor price and spec sheet with 29 rows, and a store template using 17 columns. Together they cover 29 unique SKUs with size, colour, scent and capacity variants. Problems were planted on purpose.
| Planted problem | What happened |
|---|---|
| A glass bottle is 12 US fl oz in one file and 350 ml in the other | Held as a unit conflict, with both values copied exactly. Not converted, rounded or chosen. 12 US fl oz is 354.88 ml, so the two statements are not the same product fact. The approver is asked which capacity is printed on the product. |
| A mug with no capacity in either file | Held as a missing required value. Not inferred from the SKU code or from sibling variants. |
| Two towels identical except for SKU, barcode and colour | Kept as two separate variants, not merged as duplicates. |
| The same towel SKU entered twice, once in lowercase with stray spaces | Collapsed to one row, because it is a true duplicate with identical facts. |
The result was 27 rows ready for the template and 2 held back with reasons. Every price, barcode, weight and option value in the output traces to a raw supplier cell. A next-batch file with the price column renamed stopped the run and wrote nothing.
The example shows the checks on invented data. It was not loaded into a real store or importer, which is why a paid project includes a ten-SKU test through your own validator. It is not a client result.