Skip to main content

Prepare an import file

An import connects spreadsheet columns to an existing asset or relationship type. Preparing its identifiers and values helps you update the intended records without creating unwanted duplicates.

Use a separate file for each asset type or relationship type. Importing data does not create new types or properties in the metamodel.

Start from a template​

You need an Administrator or Owner account in the organization to access imports.

  1. Open the organization settings and select Imports.
  2. Under Download a blank template:, choose an asset or relationship type.
  3. Open How to fill to read the expected columns, choices and constraints.
  4. Close the help and select CSV or XLSX.
  5. Open the downloaded file and add your data below the header row.
  6. Save a working copy, preserving the original identifiers.

The asset template includes Name, Owner, editable properties and relationship columns. These two header names remain English even in the French interface. Calculated and hidden properties are excluded.

A relationship template includes the relationship's Boldo ID, its source and target, its editable properties, and Archived. Leave the relationship ID empty when creating a new instance. For an update, use the existing instance's identifier.

Choose identifiers deliberately​

An asset's name describes it; its Boldo ID identifies the record. A unique-key property can identify it across systems when your organization has configured one.

Intended operationPreparation
Create assetsProvide names and the properties you want to populate.
Update by namePreserve the existing names. A different name can create another asset.
Rename an assetUse its existing Boldo ID or stable unique key, together with the new name.
Update one relationshipPrefer its Boldo ID, especially when parallel relationships exist.
Create relationshipsIdentify one source and one target per row, with properties that distinguish parallel instances.

Keep identifiers as text in Excel. This preserves leading zeros and prevents long values from being converted to scientific notation. Do not invent Boldo IDs for new records.

The identifier selected during import must be populated. In an asset import, a row with an empty matching key is skipped even if its name is present. Read the asset matching rules before combining creations and updates.

Arrange the spreadsheet​

Boldo accepts CSV and XLSX. Convert older Excel formats to XLSX before uploading them. Put headers on the first row and give every useful column a distinct name.

For XLSX, Boldo reads the first worksheet. Move the data to that worksheet or save a separate workbook. Blank header columns are ignored, repeated headers are renamed, and completely empty rows are omitted. Avoid relying on those corrections: inspect the resulting columns before proceeding.

For CSV, UTF-8 preserves accented characters. Boldo can detect commas, semicolons, tabs and vertical bars as column separators. If the columns look wrong, set Delimiter explicitly. If accents are corrupted, check Encoding; ISO-8859-1 and windows-1252 are also available.

The delimiter between columns is different from the delimiter inside a cell containing several values. For example, a semicolon-separated CSV can contain SaaS,Mobile in one multiple-choice cell. Use your spreadsheet's CSV export to quote cells containing separators or line breaks.

Format values​

Use How to fill for the selected type: maximum lengths, score bounds and available choices depend on your organization's configuration.

ValueExample and expected format
Date2026-10-01 or 2026/10/01; day/month or month/day forms follow the column’s Date format.
Number or currency1234.50 or 1234,50, without a currency symbol or unit.
ScoreAn integer within the property's configured range.
Choice or yes/noUse consistent values, then check their mapping to Boldo's choices.
Multiple choicesSeparate values with the configured cell delimiter, a comma by default.
Rich textMarkdown, such as **Important**, rather than pasted page markup.
OwnersOrganization members' email addresses, separated by the configured cell delimiter.
Related assets in an asset importNames, IDs or unique keys matching the mode selected for that column.

For dates, open the column options and check Date format: dd/mm/yyyy or mm/dd/yyyy. Slashes and hyphens are accepted. Boldo proposes month first only when unambiguous values support it and none indicate day first. Otherwise, it proposes day first: check ambiguous dates such as 03/04/2026. Dates written year first keep the same meaning in either mode. This applies to asset and relationship properties.

For yes/no fields, Boldo automatically maps true, vrai, oui, yes, 1 and x to true. It maps false, faux, non, no and 0 to false, ignoring case, accents and surrounding spaces. This applies to asset and relationship properties. Check the proposed mapping and assign any other values yourself.

For a standalone relationship import, each source or target cell identifies one asset. It is not a list of targets.

For example, an asset file can contain:

Name;Editor;Yearly Licence;TSA End date
Customer portal;Example supplier;1234.50;2026-10-01

These property names are examples. Use the names from your own template and inspect the detected mapping.

Review before uploading​

  1. Confirm the file contains the intended type and first worksheet.
  2. Check identifiers, duplicate rows, dates and choice values.
  3. Remove irrelevant columns, especially an unwanted name column when updating by ID or unique key.
  4. Decide whether empty cells should preserve or clear existing property values.
  5. Decide whether relationship columns add information or replace existing links.
  6. Start with a small representative file and inspect its preview.

A recognized name column can still supply a name or matching fallback during an ID or unique-key asset import. Removing an unwanted column is safer than assuming an unmapped name column has no effect.

Continue with importing assets or importing relationships. For large files, use manageable batches and inspect each result before continuing.