How to Clean a CSV Before Uploading to Salesforce (Avoid Every Import Error)

You have a CSV full of leads, contacts, or accounts ready to go into Salesforce. You click Import, wait for the progress bar, and then watch in horror as Salesforce spits back hundreds of errors. Bad dates. Missing required fields. Phone numbers that do not match. Duplicate records flooding your CRM. This is one of the most common frustrations in sales operations, and it is almost entirely preventable. The key is cleaning your CSV before it touches Salesforce. This guide walks through exactly what Salesforce expects, the seven most common import failures, and how to fix each one — manually or automatically with NoSheet.

What Salesforce Expects from Your CSV

Salesforce Data Import Wizard and Data Loader are surprisingly strict about data formatting. Before you upload anything, make sure your file meets these requirements:

• Field mapping: Your CSV column headers must match Salesforce field names or API names. "Phone Number" will not automatically map to the Phone field unless you manually map it during import.
• Required fields: Every Salesforce object has required fields. For Leads, you need at minimum LastName and Company. For Contacts, you need LastName. Any row missing a required field will fail.
• Date format: Salesforce expects dates in YYYY-MM-DD format (ISO 8601). Dates in MM/DD/YYYY, DD-Mon-YY, or other regional formats will be rejected or misinterpreted.
• Phone format: While Salesforce stores phone numbers as text, integrations like Salesforce SMS, CTI adapters, and Einstein Activity Capture expect consistent formatting. E.164 is the safest bet.
• Email validation: Salesforce validates email format on import. Any malformed email (missing @, double dots, spaces) will cause the row to fail.
• Picklist values: If your CSV contains values for picklist fields (like Lead Source, Industry, or Status), those values must exactly match the picklist options defined in your Salesforce org. "Referral" will not match "referral" or "Referred" if your picklist says "Referral".
• Field length limits: Text fields have character limits. The standard Name fields allow 80 characters. A company name that exceeds 255 characters will be truncated or rejected.
• UTF-8 encoding: Your CSV must be UTF-8 encoded. Files saved from older versions of Excel in ANSI or Windows-1252 encoding can produce garbled characters for names with accents, umlauts, or non-Latin characters.

The 7 Most Common Salesforce Import Failures

After helping thousands of users clean their data for Salesforce, these are the seven failures we see over and over again:

1. Duplicate Leads and Contacts

You import 10,000 leads and suddenly have 3,000 duplicates. This happens when your CSV contains records that already exist in Salesforce, or when the CSV itself has internal duplicates. Salesforce Data Import Wizard has a basic duplicate-matching feature, but it only catches exact matches on email. Fuzzy duplicates (different formatting of the same phone number, slight name variations) slip right through.

2. Bad Date Formats

Your CSV has dates like 3/15/2026 or 15-Mar-26. Salesforce wants 2026-03-15. If even one date in your file is in the wrong format, the entire column can fail to map, or worse, dates get silently misinterpreted (is 03/04/2026 March 4th or April 3rd?).

3. Missing Required Fields

A surprising number of CSVs have blank LastName or Company fields. Sometimes the data was collected through a web form that did not enforce those fields. Sometimes rows got corrupted during export. Either way, Salesforce will reject every single row where a required field is empty.

4. Phone Format Mismatches

Your CSV has phone numbers in a dozen different formats: (415) 555-1234, 415.555.1234, +14155551234, and 4155551234. While Salesforce will technically store all of these, inconsistent formatting makes it impossible to deduplicate on phone number, breaks telephony integrations, and creates a mess for your sales reps trying to dial from Salesforce.

5. Invalid Email Addresses

Typos are everywhere. john@gmial.com, sarah@company..com, mike at outlook.com. Salesforce validates email format and will reject rows with syntactically invalid addresses. Even emails that pass syntax validation might be dead domains or disposable addresses that will bounce the first time you send.

6. Picklist Values That Do Not Match

Your CSV says "Web" for Lead Source, but your Salesforce org has "Website." Your CSV says "Technology," but the picklist has "Tech." These mismatches cause the field to be left blank on import or the entire row to fail, depending on your import settings. This is especially painful with custom picklists that have been modified over time.

7. Field Truncation

Long company names, addresses, or description fields that exceed Salesforce character limits get silently cut off. You might not notice until a sales rep sees a truncated company name months later. The standard Name field is 80 characters. The Description field allows 32,000. Custom text fields default to 255 unless configured otherwise.

Step-by-Step CSV Cleaning Checklist

Before every Salesforce import, run through this checklist:

1. Remove blank rows. Delete any completely empty rows or rows that contain only whitespace. These create phantom records in Salesforce.
2. Fill required fields. Check that every row has values for LastName, Company (for Leads), and any other required fields in your org. Flag rows with missing values and either fill them or remove them.
3. Standardize dates to YYYY-MM-DD. Convert all date columns. Watch for ambiguous dates where day and month could be swapped.
4. Format phone numbers to E.164. Strip all punctuation, add country codes, prepend the + sign. See our E.164 conversion guide for details.
5. Validate email addresses. Check for syntax errors, fix obvious typos (gmial to gmail, outlok to outlook), and remove clearly invalid entries.
6. Match picklist values. Export your Salesforce picklist values and cross-reference them with your CSV data. Replace any mismatches.
7. Deduplicate. Remove internal duplicates within your CSV based on email address or phone number. Then check against existing Salesforce records if possible.
8. Check field lengths. Identify any values that exceed Salesforce character limits and truncate them intentionally (rather than letting Salesforce do it silently).
9. Save as UTF-8 CSV. Ensure your file is saved with UTF-8 encoding, not ANSI or Windows-1252.

Salesforce Import Failure to NoSheet Fix Mapping

Here is how NoSheet handles each of the seven common failures automatically:

How NoSheet Makes Salesforce Imports Painless

Instead of manually running through the checklist above for every import, NoSheet handles the entire cleaning workflow in a single pass. Upload your CSV, select the operations you need (format phones, standardize dates, validate emails, remove duplicates), preview the results, and download a Salesforce-ready file. Everything runs in your browser, so your sensitive lead data never touches a third-party server.

The typical workflow takes under 60 seconds regardless of file size. NoSheet handles files with hundreds of thousands of rows without slowing down, which is critical for large-scale data migrations and quarterly list imports.

If you are doing regular imports — weekly lead uploads from marketing, monthly list purchases, or ongoing CRM migrations — NoSheet saves hours of manual data cleaning every single time. Your Salesforce admin will thank you.

Related Resources