You open a CSV file and see a row of labels above all the data. The upload screen asks you to match columns, but two labels look almost the same. It is tempting to click through until the importer stops complaining.
Don't import yet. The header row is your chance to understand what the file contains before changing any records.
This guide shows a simple inspection you can do without uploading anyone's information. It is about preparing a file, not fixing subscriber-status counts after an import.
Protect the original before looking inside
Save an untouched copy somewhere your organization is authorized to store it. Work only on a copy and do not paste real contact records into a public troubleshooting chat or online example.
A spreadsheet program may automatically change phone numbers, dates, long identifiers or leading zeros. If the exact raw text matters, open a copy in a plain-text editor first. Avoid saving over the original after a program has interpreted its fields.
If the file came from an unfamiliar source, follow your normal security process before opening it. You do not need to run a macro or install a converter to inspect a CSV header.
Identify what the first line really says
A simple example CSV might begin:
email,first_name,permission_status,permission_source
[email protected],Alex,unconfirmed,signup_form
These values are fictional. The .invalid address is not a destination for sending a test message.
The first line names four columns; the second line shows how an individual row is arranged. A comma-separated file can also contain quoted fields, such as "Smith, Alex", where the comma inside quotes belongs to one value. Some exports use a different delimiter or a different text encoding. Do not assume each visible comma always marks a column boundary.
If the file opens as one enormous column, confirm the delimiter or import preview before changing the underlying data. If a label has strange characters at the start, record the exact appearance: it might be an encoding issue, not a misspelled field.
Make a four-column field map
Write the source header exactly, then decide what it means and whether the destination actually offers a matching field.
| Source header | Meaning you can verify | Destination field | Decision |
|---|---|---|---|
email |
Email address for this row | Provider email field, if supported | Map after validation |
first_name |
Given name, if actually collected | Given name, if supported | Optional |
permission_status |
Status defined by the exporting provider | Do not guess | Hold for provider-specific review |
permission_source |
Origin of the permission record | Do not guess | Preserve separately |
A column named permission_status is not proof that people consented to marketing. You need the source system's actual status definitions and the permission record. Never turn an unknown, unsubscribed or unconfirmed record into a subscribed contact just because the destination uses different words.
Empty columns may be normal. Duplicate-looking headers may hide different kinds of data. If you cannot explain a field, label it UNMAPPED — CHECK OWNER rather than invent a destination.
Check one sample without using real contact information
For a practice run, write a separate fictional file with the same column structure but no actual customer's details. Confirm that four headers produce four fields in your chosen preview, that a quoted comma stays inside one field, and that any empty value remains in its intended column.
A worked row could be [email protected],Alex,unconfirmed,signup_form. The expected interpretation is:
- Email:
[email protected] - First name:
Alex - Source-defined status:
unconfirmed - Source-defined origin:
signup_form
This is a format check, not permission to send an email or create a subscriber. If your actual provider offers a safe test import, use only an authorized test environment or account and verify how it handles statuses before a production change.
Know when to stop
Stop and ask the file owner or official provider documentation for help when a status field has unclear definitions, the file contains private information you lack permission to process, an importer proposes silently dropping columns, or the preview changes dates, identifiers or consent records.
Write down the header, suspected delimiter, number of fields, mapping decisions and unresolved questions. Keep the real export restricted. If the first row is interpreted correctly and each sensitive column has a documented decision, you have completed the pre-import header check—without risking a live contact change.
No real account import, subscriber update, delivery test or vendor UI behavior is claimed in this guide.