Original worked guide / 2026-10-05
Fix CSV import errors: a worked contacts example
A CSV import fails when the destination cannot interpret the file structure or its business values. First separate quoting and column-count errors from destination rules. This example fixes a synthetic contact export without silently dropping extra cells or changing customer IDs.
By Hasnain Qureshi / Karachi, working remotely with UK and US teams.
Try the original source
Download the deliberately inconsistent source CSV and open it in the CSV Cleaner & Import Checker. It has six data records, four intended columns and synthetic example.com addresses. Keep the untouched source. Record numbers below include the header; physical line numbers can differ for multiline fields.
| Source record | Finding | Decision |
|---|---|---|
| 2 | Whitespace around Alice | Trim only after review |
| 3 | Company is Field, Ltd | Keep the quoted comma |
| 4 | Duplicate of record 2 after trimming | Remove the exact duplicate |
| 5 | All fields empty | Remove the blank row |
| 6 | Missing Company field | Pad only if Company is optional |
| 7 | Unexpected fifth field | Retain in removed-rows download for review |
Choose the rules deliberately
- Enable source headers and keep automatic comma detection. Inspect the preview before choosing fixes.
- Choose Trim whitespace, Remove blank rows and Remove exact duplicate rows. Trimming makes records 2 and 4 identical. Without trimming they are different strings.
- Choose Pad short rows. The extra-cell record still blocks download: padding cannot decide what an unexpected value means.
- Correct record 7 in the source or remove it from this test input and save it separately. Recheck the five remaining records. Alternatively, Exclude inconsistent rows retains both short and long rows separately, giving a different reviewed output.
- For the expected download in this example, retain the padded Cara row and exclude only Dan by editing the test source. Confirm the proposed three-row output before downloading.
Reconcile the result
The expected result contains Alice, Bob and Cara. Alice retains ID 001; Bob retains the comma in his company name; Cara has an empty optional Company cell. Dan is unresolved, not a verified contact. Download the expected reviewed CSV and compare the values, not just the row count. A three-row result is correct only for the decisions stated here.
If you choose the tool’s general Exclude inconsistent rows rule instead, Cara is also excluded. Download removed rows and reconcile all source records. Do not substitute the sample output for your business file.
What remains before importing
A four-column file can still have an invalid email, missing consent, an unknown custom field or an ID that already exists in the destination. The cleaner does not connect to your CRM, validate email deliverability or infer those rules. Test a small destination import and compare created, updated and rejected records. If the same cleanup recurs, custom data import and integration development can put validation and review into the actual application.
Discuss a recurring import workflow
Share the current workflow, constraints and desired outcome. We can review scope and agree what needs implementation.