mapryva.Prepare supplier CSV files for ERP imports: mapping checklist
Renaming columns is easy. Keeping the meaning of item IDs, prices and units intact is the real work. Use this repeatable workflow to prepare supplier spreadsheets for an ERP or store import.
By the Mapryva team · Updated 26 October 2026
A practical guide from Mapryva. No software purchase or signup is needed to use it. Mapryva is in development; there is no app download yet.
1. Define the destination before editing the source
Start with the destination system’s current import template. For every target column, record its name, required status, type, allowed values, unit and blank-value behaviour. An empty stock value could mean unknown, leave unchanged, or zero. Agree on its meaning before importing.
Keep the original file unchanged. Work on a copy and record the supplier, date and template version. That gives you something to compare against when a supplier changes a header next month.
2. Write an explicit mapping table
This is a fictional example, not an import template for a particular ERP.
| Source column | Target column | Rule | Check |
|---|---|---|---|
| Supplier SKU | supplier_sku | Text; retain leading zeros | Required, unique per supplier |
| Item name | description | Trim outer whitespace | Check encoding |
| Net EUR | unit_cost | Parse agreed numeric locale | Confirm currency and price basis |
| Units per pack | pack_size | Positive whole number | Confirm units in each pack |
| Available | stock_units | Confirm stock unit | Flag blanks for review |
Download the free mapping template (CSV). It contains fictional column names and review rules only. Use a text/CSV importer with UTF-8 encoding and a comma delimiter.
3. Treat SKUs as identifiers, not quantities
0007 does not necessarily mean 7. Explicitly import SKUs, barcodes and other identifiers as text. If Excel has already discarded zeros or digits, changing the cell format afterwards cannot reliably recover them. Reimport the untouched original with the correct type.
In Excel, start with Data → From Text/CSV and transform before loading. Review automatically inferred types. Power Query automatically detects types for unstructured sources, so explicitly set identifier columns to text. See Microsoft’s Power Query data type documentation and text/CSV import instructions.
4. Validate prices and units separately
Renaming 12,50 € to “price” does not make it a valid number. Specify numeric locale, currency and precision. Confirm whether the amount is per item, pack, kilogram or hundred units, and whether it includes tax.
For example, EUR 30 per pack of 12 is EUR 2.50 per item only when the pack size and price basis are confirmed. AI should not guess either from a vague header or a similar product.
5. Run these checks before every import
- Schema: Verify required columns. Flag missing, new or renamed headers.
- Keys: Quarantine missing and duplicate SKUs. Use supplier plus supplier SKU when IDs are not globally unique.
- Row counts: Reconcile input rows with accepted plus held-back rows. No silent drops.
- Values: Check currency, unit, sign, decimal convention and allowed values. Keep blank and zero distinct.
- Changes: Separate new items, missing items and unusually large price changes for review. An absent SKU is not automatically discontinued.
- Dry run: Test a small reviewed sample in a sandbox or import preview. Approve the complete import only after checking the result and preparing a recovery path.
6. Choose the simplest tool that covers the workflow
Excel can be enough for a small one-off file. For recurring files with a stable structure, Power Query is a sensible first option: keep transformations explicit and refresh them for the next file. The ERP’s own importer may already cover your required fields and error reports.
A separate mapping tool becomes useful when many sources send different schemas or files need to stay on the device. Still keep three jobs separate: column mapping connects fields, record matching connects items, and business validation checks whether the values make sense. A plausible column match cannot replace the other two.
Where Mapryva fits
We are developing Mapryva as a desktop app that proposes CSV, Excel and JSON field mappings using local AI, with human review before export. These are planned capabilities, not an available ERP integration. An initial macOS test is the focus; timing, requirements and pricing are not confirmed.
If supplier imports are a recurring part of your work, you can join the free waitlist for news about test access. Please do not send confidential supplier or customer files.
Join the early access waitlist
Free waitlist. No confirmed launch date. No download yet.