Bulk import contacts

Pour :Administrateur

↗Open in Unisoft

View of the 'bulk import' page

Bulk import lets you load a list of contacts from an Excel (.xlsx, .xls), OpenDocument (.ods) or CSV (.csv) file, with a 4-step wizard: upload, mapping, validation, import.

Access it from the Contacts menu → Import, or via the URL /app/contact/import.

⚠

The import is IRREVERSIBLE

Once launched, the import creates or updates contacts directly in your base. There is no automatic undo. Pay attention to the chosen deduplication strategy and back up your base before any important import.

Step 1 — File upload

1. 1

   (Optional) Download the template

   The Download the template button offers a sample Excel file with the recommended columns. Useful to prepare a file from scratch.

2. 2

   Drag and drop or click

   Drop your file in the upload area, or click to select. Accepted formats: .xlsx, .xls, .csv, .ods. Recommended size: up to 50,000 rows.

3. 3

   Choose the sheet (if multiple)

   If your Excel file contains multiple tabs, select the one that contains your contacts.

4. 4

   Confirm the header row

   The wizard automatically detects if your first row is a header (short non-numeric text). Confirm or adjust: indicate Yes or No, and the header row number (or the first data row if no headers).

A raw preview of the first 10 rows is displayed. Empty rows are automatically ignored.

★

Preamble (title, logo) before data

If your file starts with title or logo rows, indicate the correct row number for headers (or the first data row). The wizard will ignore everything above.

Step 2 — Column mapping

This is the key step: you indicate which contact field corresponds to each column of your file.

The wizard displays a table with, for each column:

• The column name in your file (header or Column A, B, C…).
• Up to 3 examples of non-empty values from the file.
• A Target field selector.
• (If AI mapping used) The confidence score in percentage.

Available target fields

Standard field	Type	Notes
Last name, First name	Text	Basic identity
Primary email, Secondary email, Email (3)	Email	Multiple columns possible to emails[]
Primary phone, Secondary phone, Phone (3)	Phone	Multiple columns possible to tels[]
Date of birth	Date	Format detected or imposed at step 3
Gender (Male yes/no)	Boolean	Accepts Male/Female, M/F, H/F, Yes/No…
Address, Postal code, City, State/Region, Country	Text	Primary address
Address 2 (Label, Street, Postal code, City, Country)	Text	Secondary address
Archived, Deceased, Head of family	Boolean	Contact states

In addition, you can map to:

• Tag / Category — the cell becomes a category applied to the contact. If the cell contains multiple separate categories (by comma, semicolon, pipe or slash), they will be split according to the separator chosen at step 3.
• Existing custom fields — your tailor-made fields already created in your organization (Text, Number, Yes/No, Date, Date & Time).
• + Create a new custom field — to create on the fly a field that doesn't exist yet.
• — Ignore this column — — for useless columns (internal ID, header duplicates…).

ℹ

A minimum identity field is mandatory

At least one column must be mapped to Last name, First name, Primary email or Primary phone to be able to continue. Without an identifier, impossible to create a contact.

Automatic AI mapping

The Automatic AI mapping button asks an AI model to analyze your columns (headers + examples) and propose a mapping for each. The result displays a confidence score:

• ≥ 80% (green) — reliable, can be validated as-is.
• 50–80% (yellow) — to check.
• < 50% (orange) — uncertain, to correct manually.

Hover over the score to see the reason invoked by the AI.

The AI remains a suggestion: you remain free to adjust everything manually before continuing.

Create a new custom field

If a column doesn't match any standard or existing custom field, choose + Create a new custom field. A modal asks:

• The label displayed to the user (e.g.: "Car").
• The technical key (without spaces or accents, e.g.: car).
• The type: Text, Number, Yes/No, Date, Date & Time.

The field is created on the fly and becomes available for subsequent imports.

Step 3 — Validation and options

This step checks each row and displays the individual status: valid, warning (internal duplicate in the file), error (invalid format, missing identity), ignored (manually excluded).

Import options

Three option sections to set:

Duplicate management & filters

Option	Values	Behavior
If the contact already exists	Update / Ignore / Create a duplicate	If an existing contact has the same email or phone: enrich (update), skip (skip), or create anyway (create).
Require a contact method	Yes / No	If enabled, rows without email AND without phone are rejected. Recommended.

⚠

Duplicate strategy: choose carefully

Update is generally the right choice: we enrich existing contacts with new info without creating duplications. Create a duplicate is dangerous and should be reserved for particular cases (for example two distinct legal entities sharing a business email).

Data format

Option	Description
Default country (phones)	Country used to interpret numbers without area code. E.g.: France (+33) transforms 0612345670 into +33612345670.
Date format	Format imposed to parse dates of birth. Auto-detection tries several common formats. To force if your dates are ambiguous (12/03/2024 = March 12 or December 3?).
Tag separator	Character that separates categories in a Tag-type cell (comma, semicolon, pipe, slash).

Global tags

Add tags that will be applied to all imported contacts. Ideal to trace the source (IMPORT_2026_APRIL, RACHIDESS_LIST, PASSOVER_EVENT). Enter the tag then click on Add (or press Enter). Click on the cross of a tag to remove it.

Verification table

Below the options, the table lists all rows with their status. You can:

• Filter by status (All / Errors / Warnings / Ignored).
• Edit a cell directly in place — click on it, modify, validate with Enter. The modifications do not affect your original file; they are kept in memory until import.
• Ignore a row — click on its row number (first column) to switch it to ignored. Click again to reactivate.

The summary at the top displays: N valid, N warnings, N errors, N ignored out of N rows in total.

ℹ

To continue, you need zero errors

The next step (Import) is only accessible if all errors are corrected (or the rows in error are ignored) AND there remains at least one valid row.

Step 4 — Launch the import

A final confirmation displays before launch:

• Summary of the number of rows to import, with warnings and ignored.
• Reminder of the chosen duplicate strategy.
• Reminder of the contact requirement filter if it is active.
• List of applied global tags.
• "I understand that this action is irreversible" checkbox — mandatory before being able to click on Launch the import permanently.

During the import

The import runs in batches of 200 contacts. A progress bar indicates the current batch and percentage. Live counters display:

• +N created (green) — new contacts.
• ↻N updated (blue) — enriched existing contacts.
• N ignored (gray) — skipped according to the duplicate strategy.
• N errors (red) — rejected by the server.

After the import

A result screen displays the final summary:

Status	Meaning
Import successful	No errors on the server side.
Import completed with errors	At least one row was rejected by the server.

Two buttons are offered:

• Return to contacts — takes you back to the contact catalog.
• Download the error report — generates a CSV with, for each rejected row, the row number in the original file, the type of error (validation or server), the reason, and all the original values. Ideal to correct and then re-import only the rows in error.

★

Successive imports

If you re-import the same file with the Update strategy, the contacts already created on the first pass will not be duplicated — they will simply be re-overwritten with the same data. It is harmless, and it is even useful to retry after correcting errors.

Limitations and best practices

• Recommended max size: 50,000 rows per file. Beyond, split into multiple files.
• No direct import of family links from this wizard — the parent-child links are then created manually or via the API.
• Back up first your base if you import on a base containing already contacts (update = overwriting of certain values).
• Do a test on 10 rows before a big import — extract the first 10 rows in a separate file, launch the import, check the result in the catalog, then launch the real import.