Skip to content
New: simulate any flow against real records before it goes live. See what's new

Import contacts and companies from CSV

Last updated 3 August 2026

Solstral imports a CSV into any object — the three defaults or one you defined yourself. The importer maps your columns onto that object's fields, so there is no fixed template you have to conform to.

1. Prepare the file

One row per record, with headers in the first row. Save as UTF-8 if your data contains accented characters, currency symbols or anything outside plain ASCII — this is the single most common cause of mangled names after an import.

A few habits that save time: keep dates in a consistent format, put one value per column rather than combining "City, State", and strip trailing spaces if your export is untidy.

2. Import companies first

If you are importing both companies and contacts, do companies first. Contacts can then be linked to an existing company during mapping. Import them the other way around and you will have to relate them afterwards, one at a time or by re-import.

3. Map the columns

The importer lists each CSV column beside a dropdown of the object's fields. Custom fields appear in that list alongside the standard ones, so a column such as "Account manager" or "Renewal date" can go straight into the field you made for it.

Two kinds of field are deliberately absent from the list: formula fields and reference (lookup) fields. Both are computed when a record is read rather than stored on it, so there is nothing to import into. If a column in your file corresponds to one of these, leave it unmapped — the value will be calculated for you.

4. Review and merge duplicates

After the import finishes, open Duplicates. Solstral surfaces likely matches for you to confirm rather than silently merging anything, because an incorrect automatic merge is far more expensive to unpick than a duplicate is to combine.

Merging keeps both records' history — activities, emails and notes are preserved on the surviving record — and lets you choose the winning value field by field where the two disagree.

5. Check the audit log

Every imported record is written to the audit log, so if something looks wrong you can see exactly what the import created and when. That makes a bad import straightforward to identify and reverse.

Importing a large file? Run a 20-row sample first and check the mapping produced what you expected. Fixing a mapping mistake on 20 records takes a minute; on 20,000 it does not.