CSV field mapping sheet
Do not draw the field mapping table by hand before an ERP, CRM or e-commerce transfer. Paste the source and target header lists; name similarity (with TR/EN/DE synonyms) suggests matches, you correct them and add transformation notes. The data never leaves your device.
Paste the header row (separated by tab, ; | or comma) or write one field name per line.
Paste the template headers of the target system in the same way.
Headers are processed only in your browser; nothing is sent to a server. You do not need data rows, column names are enough.
Paste source and target field names to see the mapping table.
01
How to use it
A
Paste the header row of the source system and the template headers of the target system into the two boxes (or try the sample).
B
Review the suggested mappings: correct low-similarity ones with the dropdown and write a transformation note where needed.
C
Check the unmapped, duplicate and empty fields; download the table as CSV or Markdown and share it with your team.
02
How is the suggestion made?
Every name is first simplified: lower-cased, accents removed (ş → s, ü → u, ß → ss), camelCase and underscores split. Then a small synonym list reduces it to concepts: 'Material Number', 'Artikelnummer' and 'sku' lead to the same concepts (product, id).
Similarity is a weighted sum of three measures: the Jaccard similarity of the concept sets, the Levenshtein (edit distance) ratio of the simplified names, and how much of the smaller set is shared. Below 0.6 no suggestion is made; 0.9 and above is a strong match, anything in between should be reviewed.
The mapping is one-to-one: all pairs are scored and, from the highest down, matched if both are still free. So the same target field is not proposed to two sources; you can still map two sources to the same target by hand and the tool marks it as a duplicate mapping.
03
What belongs in a mapping table?
Source field, target field and transformation rule: decimal separator (1,5 → 1.5), date format (31.12.2026 → 2026-12-31), unit conversion (g → kg), code mapping (Yes → true), merging or splitting (Full name → first, last), default value and empty-value behaviour.
Required target fields that nothing fills and source fields that go to no target must be decided before the transfer: either a source is found or a fixed/default value is given, or it is written down that the field is dropped on purpose.
04
Afterwards
The mapping table is the contract of the transfer. Try it end to end with a sample of a few hundred real rows, then check the source data with the master data quality check (duplicates, barcodes, units).
FAQ
- Is my data uploaded anywhere?
- No. The headers you paste are processed in your browser; nothing is sent to a server or stored. You do not need data rows anyway, column names are enough.
- In which format can I paste the headers?
- The first row of a CSV file (separated by tab, semicolon, pipe or comma) or a list with one field name per line. In a multi-line paste, if the first line contains a tab, ; or |, a table is assumed and only the first row is taken. Delimiters inside quotes are preserved.
- Why were some fields not mapped?
- If the similarity is below 0.6 no suggestion is made: the names share no concept or the abbreviation is not in the dictionary. Choose these fields yourself with the dropdown. Leaving a field empty instead of making a wrong suggestion prevents a silent error.
- Can I map two sources to the same target?
- Yes; for example first name and last name can go to a single full name field. The tool marks it as a duplicate mapping and carries it into the output; write the merge rule in the transformation note (e.g. 'first + space + last').
- Do Turkish characters break when I open the CSV in Excel?
- No; the file is saved with a UTF-8 signature (BOM). In the Turkish and German versions the delimiter is a semicolon (matching Excel's local settings), in the English version a comma. Cells that could be read as formulas (starting with =, +, - or @) get an apostrophe in front.
Let's plan your data transfer together
We prepare field mapping, transformation rules and transfer tests between ERP, CRM and e-commerce systems together. Let's talk about your systems and data in a free discovery call.