6.1 Data Import Wizard, Data Loader, Export, and Backup

Choosing the right data tool

Data and Analytics Management is the largest current Platform Administrator domain, so treat data movement as an operational control problem rather than a button-clicking topic. The admin must protect trust in the org while still helping business teams move quickly. A sales operations user may only ask to upload a spreadsheet, but the real admin question is broader: which object is affected, how many records are changing, what fields are required, which automations fire, how duplicate rules behave, and how the team will reverse the change if the load is wrong.

The Data Import Wizard is the guided, browser-based option. It is appropriate when the import is supported by the wizard, the record volume is within the wizard range, and the admin wants a safer mapping experience for accounts, contacts, leads, campaign members, person accounts where enabled, or custom objects. It can help avoid duplicates by using available matching choices during import. It is not the right tool for every object, very large files, complex updates, scheduled jobs, or delete operations.

Data Loader is the administrator workhorse for larger and more technical jobs. It supports insert, update, upsert, delete, hard delete where permitted, export, and export all. It can use record IDs or external IDs, save mapping files, run through the user interface, and support command-line automation in environments where that is approved. Because Data Loader can change many records quickly, the admin should use it with a tested file, explicit permissions, and a clear backup.

Setup paths matter in practice. Data Import Wizard is reached from Setup > Data Import Wizard. Data Export is reached from Setup > Data Export, where an admin can request an export or schedule recurring exports if the org edition and settings allow it. Data Loader is installed locally or launched through the supported Salesforce distribution, and it requires API-enabled access. Users also need object and field permissions for the work they are doing.

Safe import and backup workflow

A realistic import starts before Salesforce. Profile the spreadsheet first. Look for blank required fields, inconsistent picklist values, state and country formatting, duplicate names, phone number patterns, invalid email addresses, date formats, currency columns, owner names that do not match Salesforce users, and lookup references that need IDs. If the source file cannot identify existing Salesforce records reliably, pause and build a matching strategy. Updating by name alone is risky when two accounts share similar names.

Use a small test file in a sandbox or Trailhead Playground whenever the change has meaningful risk. The test should include clean rows, bad rows, duplicates, missing lookups, picklist edge cases, and records owned by different users. The goal is not only to prove that successful rows load. The goal is to see how failed rows are reported and whether validation rules, flows, triggers, duplicate rules, assignment rules, and required fields behave as expected.

Recommended workflow:

1. Define the business outcome, affected objects, record count, and required timing.
2. Export the target records before changes, including Salesforce record IDs and important lookup fields.
3. Clean and normalize the source file, then freeze the version that will be loaded.
4. Decide whether the action is insert, update, upsert, delete, or export.
5. Map every CSV column to a Salesforce field and remove columns that should not change.
6. Test with a small file and review success and error files row by row.
7. Run the production load during an approved window and monitor automation, errors, and user impact.
8. Save the original file, mapping, success file, error file, and rollback notes.

External IDs are important in integrations and migrations. An External ID field stores a key from another system, such as an ERP customer number. With Data Loader upsert, Salesforce can update a matching record when the external ID already exists or insert a new record when it does not. This avoids fragile name matching and reduces duplicate risk. The external ID field should be unique when the business process requires one source record to match one Salesforce record.

Exports are not just for migration. An admin may export data before a bulk owner change, before a deduplication cleanup, after a release that changed automation, or as part of a scheduled data retention process. Data Export produces compressed files that can include data, attachments, files, and related content depending on options. It does not automatically prove that the organization can restore quickly. A mature admin documents who downloads export files, where they are stored, how long they are retained, and how sensitive data is protected outside Salesforce.

Study trap: do not choose the biggest tool just because the job sounds important. A marketing coordinator importing 2,000 campaign members may be better served by a guided wizard or campaign member import flow. A system administrator updating 300,000 contacts with an external system key likely needs Data Loader. A compliance team asking for a periodic copy of all data needs Data Export or a backup product, plus retention and access controls. The right answer is the tool that fits object support, volume, operation, repeatability, and risk.