Free · No signup · Runs in your browser

Map CSV columns to the schema you need.

Load the export you have and the import template you need. Deterministic header-only suggestions connect the two; you review every field, transform, default, and issue before downloading a ready-to-import CSV.

01 · How it works

Three steps, then done.

Imports fail when two systems name the same field differently. This workspace makes that translation explicit, repeatable, and testable without sending customer or operational data to a server.

i. define

Load source and target

Use CSV, TXT, XLS or XLSX for the source; upload a target template, paste field names, load the sample, or arm a saved recipe.

ii. map

Review every decision

Accept deterministic header matches, choose source columns, set output order, types, required flags, defaults, and transforms.

iii. validate

Export only when ready

Validate the complete transformed output, inspect row-level issues, then download mapped CSV and aggregate evidence.

02 · Why ours

An import mapping that shows its work.

No black-box confidence score and no silent remapping: every suggestion, unavailable recipe source, default, transform, and validation failure remains visible.

  • 01

    Deterministic suggestions

    Header normalization, aliases, token overlap, and bounded edit distance produce explainable exact, strong, possible, or unmatched candidates.

  • 02

    Complete-output checks

    Required and type rules run after defaults and transforms across every row, not just the preview.

  • 03

    Reusable recipes

    Save or export versioned mapping definitions without filenames, source rows, previews, suggestions, issues, or results.

  • 04

    Safe evidence

    Aggregate HTML and Markdown omit raw values and defaults; the raw-value issue CSV requires an explicit disclosure.

"The import is ready only when every required target and transformed output value passes."
— the mapping receipt
03 · FAQ

column mapper questions.

How are source columns matched to target fields?
Suggestions use header text only: Unicode and punctuation normalization, case and separator handling, a bounded alias dictionary, token overlap, and bounded edit distance. Only unambiguous exact or strong matches are applied automatically; possible matches stay suggestions.
No. The interactive mapping is one-to-one: choosing a source already used elsewhere clears its prior mapping and announces that change. You can use an explicit default for a separate target field.
Ready means a source and target exist, every required target is mapped or has an explicit default, target names and source assignments are unique, and every transformed output value passes its selected required and type rules.
Only a version, recipe name, and ordered target fields with configured source names, type, required flag, explicit default, and transform. It excludes filenames, rows, previews, suggestions, unavailable-source findings, issues, and results.
Configured source names resolve exactly after documented header normalization. Missing sources are listed as unavailable and are never silently replaced by a new suggestion.
No. Mapping, transformation, validation, recipe storage, and exports run locally in your browser. Normal use does not send source or output rows to csvtodashboard.
Related tools
CSV Import Mapping Guide
Design, validate and reuse a defensible field map.
Open →
CSV Data Dictionary
Document source meaning, ownership and sensitivity.
Open →
CSV Profiler
Audit types and data quality before import.
Open →
CSV Validator
Repair delimiter, quote and header structure.
Open →
Clean CSV Headers
Normalize header names directly.
Open →
All CSV tools
The full local-first toolkit.
Open →