SfSkills Domains Docs GitHub

Home / data / data-loader-and-tools

data-loader-and-tools

Use this skill when selecting or configuring Salesforce data load tools: Data Loader, Data Import Wizard, Workbench, Salesforce Inspector, or Salesforce CLI for import/export. Trigger keywords: data loader configuration, import wizard, bulk data load, workbench SOQL, salesforce inspector, data export, csv import salesforce, data migration tool selection. NOT for mapping CSV columns to field API names — use data/data-loader-csv-column-mapping. NOT for coding a Bulk API 2.0 job and retrieving its failed records — use data/bulk-api-and-large-data-loads.

Open the skill folder on GitHub


Data Loader and Tools

Overview

Salesforce provides several tools for loading, exporting, and inspecting data. Choosing the wrong tool for the job is one of the most common causes of failed data migrations, governor limit errors, and security gaps. This skill covers tool selection, configuration, and safe operation — it does NOT cover writing Bulk API integrations in Apex or custom code.

Tool Comparison

ToolUse WhenMax RecordsAPI ModeOfficial?
Data Import Wizard<50K, supported objects, no automation50,000SOAPYes
Data Loader>50K, all objects, automation150MBulk API 2.0Yes
WorkbenchAd-hoc SOQL, REST ExplorerSmallREST/SOAPYes (sunset)
Salesforce InspectorQuick record lookupTinyRESTNo (open-source)
Salesforce CLIScripted/CI export-importVariableMultipleYes

Data Import Wizard

The Data Import Wizard is the preferred tool for small, standard-object loads (<50,000 records). It is browser-based, requires no installation, and includes field-mapping assistance. Supported objects are limited: Leads, Accounts, Contacts, Campaign Members, Cases, Solutions, and Person Accounts (when enabled). It does not support scheduling or command-line automation.

Data Loader (v66.0, Spring '26)

Data Loader is a desktop application available for macOS and Windows. It supports all standard and custom objects and scales to approximately 150 million records via Bulk API 2.0. Key configuration notes:

Workbench

Workbench (workbench.developerforce.com) is an officially provided browser-based tool useful for ad-hoc SOQL queries, REST API exploration, metadata retrieval, and small record operations. Salesforce has announced a sunset trajectory for Workbench. Official replacements are:

Use Workbench for exploratory, non-production tasks only. Do not build automated workflows that depend on it.

Salesforce Inspector / Inspector Reloaded

Salesforce Inspector and its community fork Inspector Reloaded are open-source Chrome browser extensions. They are not official Salesforce products and are not supported by Salesforce Support. They are useful for quick single-record inspection, field API name lookup, and ad-hoc SOQL in the browser. They should never be used for bulk operations, production data changes, or in environments with strict security review requirements (ISV packages, regulated industries).

Salesforce CLI

The Salesforce CLI (sf / sfdx) supports data operations via sf data import, sf data export, sf data query, and related commands. It is the right choice for CI/CD pipelines, scripted migrations, and developer sandbox seeding. It supports JSON and CSV formats and can target specific API versions.

Security Considerations

  1. Estimate record count and identify target objects. If under 50,000 records and the object is in the supported list (Leads, Accounts, Contacts, Cases, etc.), use the Data Import Wizard and stop here.
  2. Determine operation type. Insert, update, upsert, and delete are supported by all tools. Hard delete requires the Bulk API Hard Delete permission — verify it is granted before proceeding.
  3. Select the tool. For >50K records, all objects, or automated/scheduled runs, use Data Loader in Bulk API 2.0 mode. For scripted/CI contexts use Salesforce CLI. Reserve Workbench for exploratory ad-hoc queries only.
  4. Configure credentials securely. For Data Loader headless mode, use encrypt.sh/encrypt.bat to encrypt the password or configure OAuth JWT. Never store plaintext passwords in process-conf.xml.
  5. Run a sample load in sandbox. Execute against a representative 1,000-record subset first. Inspect the success and error CSV outputs before scaling up.
  6. Execute full load and monitor. For Bulk API 2.0 jobs, monitor job status in Setup > Bulk Data Load Jobs. Capture the error CSV and resolve all failures before closing the migration window.
  7. Validate and archive. Run post-load record counts and spot-check field values. Archive the success/error CSVs and process-conf.xml (with no credentials) for audit purposes.

Official Sources Used

Package files