Frequently Asked Questions

CleanMyCSV automatically cleans and standardizes your data files. The engine smartly fixes formatting, encoding, and structural errors to deliver a pristine CSV file that’s instantly ready for your software.

The algorithm targets several critical areas:

• Encoding & Structure: Universal conversion to UTF-8 to secure accents, plus automatic delimiter detection.
• Cleaning: Removal of completely blank rows and strict deduplication.
• Normalization: Standardizing dates, repairing zip codes (restoring missing leading zeros), and formatting email addresses and phone numbers.
• Finance: Normalizing financial amounts and automatically extracting currencies into a dedicated column.

CleanMyCSV adapts to your needs with four clear plans:

• Free Tier ($0): Perfect for testing. Includes 2 free cleans per day for files up to 5 MB.
• 24h Pass ($9): Unlimited cleaning for 24 hours. Supports files up to 25 MB.
• 7-Day Pass ($29): Unlimited cleaning for a week. Supports larger files up to 100 MB.
• Lifetime License ($99): Guaranteed unlimited lifetime access for files up to 100 MB, including all future updates.

The engine exclusively accepts .csv files. If your file is in any other format (like .xlsx), the system will immediately reject it.

Yes, security is our top priority. Your cleaned files are accessed via a secure link that automatically expires after 15 minutes. Once expired, the data cannot be recovered by anyone. For absolute privacy, we enforce a strict policy: all files are permanently deleted from our servers within 24 hours.

Upload blocks usually happen for the following reasons:

• The document is not a valid .csv file.
• The file size exceeds the limit of your current plan (5 MB, 25 MB, or 100 MB).
• You are on the Free Tier and have hit your limit of 2 files per day.
• To ensure server stability, we enforce a global rate limit of 300 requests every 5 minutes.

We guarantee total transparency. The tool generates a detailed diagnostic report alongside your cleaned file. This JSON audit shows the volume of modified data and provides a complete, line-by-line history of the actions taken on your columns.

The main difference lies in how we interpret ambiguous data. Our algorithm uses regional logic to guess the true intention behind poorly formatted data:

• Dates: The date "04/05/2023" is read as May 4th in FR mode (DD/MM), but as April 5th in US mode (MM/DD). If the engine spots "13/01", it automatically realizes the error in either mode and fixes it.
• Amounts: The number "1.234" is interpreted as 1,234.00 in FR mode (where dots often separate thousands), but as 1.23 in US mode (where dots mark decimals).
• Zip Codes: US mode triggers North American-specific rules, such as restoring leading zeros for the East Coast and standardizing the ZIP+4 format (e.g., 12345-6789).

Choosing a mode doesn't restrict the cleaning process (accents, dashes, and duplicates are fixed in both). It simply gives our AI engine the right cultural context.

Phone numbers are often typed in chaotically. Our algorithm standardizes them so they are immediately usable in your CRM or marketing tools:

• It strips out parasitic symbols (unnecessary spaces, parentheses, double zeros).
• It isolates and preserves internal extension numbers (ext / postes).
• If a cell contains multiple numbers (e.g., "06... / 01..."), it only keeps the first one to prevent import errors.

Selecting your region allows our algorithm to apply the correct formatting intelligence:

• US Mode: Local 10-digit numbers are automatically converted to the standard North American international format (e.g., +1 (555) 123-4567).
• Europe Mode: The algorithm is heavily optimized for the French market (converting numbers to +33 6...). For all other European countries (Germany, Belgium, Switzerland, etc.), we apply a precautionary principle: we perfectly strip out parasitic characters without forcing local formatting, ensuring the number remains completely valid (e.g., +4930123456).

No, your reference data is safe. Our phone detection is equipped with an "Anti-ID" safeguard. If a cell contains a very long string of raw numbers (over 13 characters, such as an EIN, a barcode, or a transaction ID) without any standard phone punctuation, the algorithm flags it as technical business data and completely ignores it.