Akeneo CSV Format Fixer

When an Akeneo import silently drops rows or throws “column not found,” the cause is almost always the file, not the product data. This Akeneo CSV import fix tool checks your export against Akeneo’s import rules — delimiter, enclosure, encoding, header naming, and locale and scope suffixes — then repairs the mechanical problems and gives you a fixed file, so you can resolve the problem before you queue another failed import job.

Akeneo CSV Format Fixer

Runs in your browser · no upload

Paste the import CSV, or upload it

Check & fix Upload a file Try an example Clear

Identifier attribute code Output delimiter (match your import profile) Semicolon ; (Akeneo default) Comma ,

Rules follow Akeneo's product CSV format documentation.

How Akeneo will read each column

Row-level findings

Download fixed CSV Copy UTF-8 without BOM, " as enclosure. Unfixable rows are kept as-is and listed above.

Checks structure against Akeneo's documented CSV conventions; it cannot see your PIM, so it does not know which attribute, family, category or channel codes actually exist there. See how Claro maps supplier files to your Akeneo attributes →

What it checks

• Delimiter and enclosure — confirms the file uses a consistent separator and that the enclosure character (default ") is balanced, so a stray comma in an MRO description or a furniture dimension field does not shift every following column.
• Character encoding — flags non-UTF-8 byte sequences and a leading byte-order mark (BOM), the two most common reasons accented brand names or the ² in mm² arrive as mojibake.
• Header integrity — detects duplicate columns, leading or trailing whitespace, and headers that do not resolve to a valid attribute code (lowercase, no spaces, no reserved characters).
• Locale and scope suffixes — validates the attribute-locale-scope pattern (for example description-en_US-ecommerce) so localizable and scopable attributes land in the right channel instead of being rejected.
• Required identifier column — verifies a sku (or your configured identifier) column exists and that no row is missing its value.
• Row width consistency — counts fields per row and reports any line where the count does not match the header, the classic cause of a “ragged” CPG price file failing partway through.
• Quoting and line breaks — finds unescaped quotes and raw newlines inside fields that break naive parsers used in some import pipelines.
• Property and value formats — enabled as 1/0, categories as comma-joined category codes, prices in price-EUR-style columns as plain numbers with a dot, and metric values paired with a filled -unit column.
• A fixed file — header renames (en-US to en_US, suffix order, case, stray spaces), enabled and category separators, price formats, encoding and BOM are fixed in a UTF-8 download. Problems that need a decision, such as a missing SKU or a duplicate identifier, are listed instead of guessed.

How the Akeneo CSV import fix works

Akeneo’s CSV importer expects a flat file that follows a small set of conventions: one header row of attribute codes, optional -locale-scope suffixes for localizable or scopable attributes, an identifier column, and UTF-8 encoding with a configurable delimiter and enclosure. When any of those assumptions breaks, the importer either rejects the file or, worse, imports a misaligned version of it.

This tool parses your file the same way a strict importer would and reports each deviation in plain language with the offending row or column. It cannot see your PIM, so it does not know which attribute, family, category, or channel codes exist in your instance, and it does not reformat opinions about your taxonomy — it only checks structural correctness against the documented CSV contract, which is why the same file behaves identically whether you run a 200-row industrial fastener sample or a 150,000-row distribution catalog.

Fixing the file is the cheap part. The expensive part is mapping a supplier’s columns to your Akeneo attribute codes in the first place, and keeping that mapping stable as suppliers change their export formats. If you are wiring CSV onboarding into a repeatable pipeline rather than fixing files one at a time, Claro’s product data layer handles schema mapping, identity resolution, and validation upstream so clean records reach Akeneo by default.

Related resources

FAQ

Why does my Akeneo CSV import fail with no error message? +

Silent failures usually trace to encoding or row-width problems rather than the data values. A BOM at the start of the file, a mismatched enclosure, or a single ragged row can make the importer skip records or stop early without a clear message. Run the file through the checks above to find the exact line before re-importing.

What delimiter and encoding does Akeneo expect for CSV imports? +

Akeneo CSV imports default to a semicolon or comma delimiter with a double-quote enclosure, and they expect UTF-8 encoding. The exact delimiter is configurable in the import profile, but it must be consistent across every row. Files exported from spreadsheets in a regional locale often switch the delimiter unexpectedly, which this tool flags.

How do I format localizable attributes in an Akeneo CSV? +

Localizable and scopable attributes use the attributecode-locale-scope header pattern, for example name-en_US-ecommerce. If an attribute is localizable but not scopable, you include only the locale; if it is neither, you use the plain attribute code. Mismatched suffixes are a frequent cause of rejected columns.

Will fixing the CSV solve a duplicate or mismatched product problem? +

No. This tool fixes file structure so the import runs. It does not deduplicate records or resolve which supplier SKU maps to which of your products — that is an identity and matching problem that happens before the file is built. For that, see the schema mapping playbook and Claro’s identity resolution layer.

Is there a row or file-size limit? +

There is no imposed limit. Because all processing runs locally in your browser, the only constraint is your device’s available memory. Large distribution catalogs of 100,000-plus rows are checked the same way as a small sample file.