Section 3 · Data Modeling and Management · 20% of exam
3.3 Import, Export, and Connect External Data
Exam objective
Given a scenario, determine options and considerations for importing and exporting data, including external data sources. Salesforce Certified Platform App Builder | SU ·26 guide alignment 10
Choose a data movement tool based on volume, object support, matching needs, repeatability, and user skill. Data Import Wizard is a guided browser tool suited to common imports and supported objects at moderate scale. Data Loader supports larger or more controlled insert, update, upsert, delete, export, and bulk operations. Exact limits and supported objects can change; verify current Salesforce Help before planning a production migration. A one-time import and a repeatable integration are different operating problems.
Before importing, profile the source: identify required fields, formats, duplicates, invalid references, ownership, and relationships. Create a mapping from source columns to Salesforce fields. Use external IDs to match records across systems and to support upsert when a reliable key exists. If importing related records, load parents first or use reference IDs that correctly resolve relationships. Decide how to handle blank values, rejected rows, duplicate rules, validation rules, and automation that may run during the operation. Back up or export affected data before bulk changes. Exports can support analysis, backup, migration, and reconciliation. Standard exports and Data Loader have different capabilities, formats, and volumes. A file export is sensitive data: store it securely, limit access, and delete it according to retention policy. Validate counts and sample records after a load. Keep the import success and error files so rejected rows can be corrected without accidentally duplicating successful rows.
External objects expose data held outside Salesforce through an external data source and an adapter. This can reduce copying and help users work with remote records, but behavior depends on the adapter and external system. Consider latency, authentication, availability, supported relationship and query behavior, field mappings, and access controls. External data is not automatically identical to locally stored Salesforce data, and some platform features or reporting behaviors may differ. Use external objects when real-time access to remote data is more valuable than local ownership or full platform feature parity.
Applied scenario: For a migration of customer accounts and related contacts, first clean the source and assign stable external IDs. Import accounts, review errors and counts, then import contacts mapped to their parent account through the external ID relationship. Test duplicate rules and validation behavior in a sandbox using a representative file. Reconcile totals and spot-check addresses, ownership, and record relationships. For future nightly synchronization, evaluate an integration or an external data pattern instead of manually repeating a one-time import workflow. Design and risk review: The most damaging errors are loading children before parent keys resolve, treating a display name as a durable matching key, and rerunning a successful insert file as if it were an upsert. Automation and duplicate rules can change outcomes during a bulk load. A data export can include confidential fields that the recipient does not need. For external data, do not promise offline availability or native reporting behavior without verifying the adapter and product documentation. Salesforce Certified Platform App Builder | SU ·26 guide alignment 11
Exam lens: First translate the stated requirement into a measurable outcome. Identify the actor, record or object in scope, timing, access boundary, and exception path. Eliminate choices that solve only the visible interface while missing the underlying data, security, or lifecycle requirement. Then compare the remaining supported features by fit, operational complexity, maintainability, and any dependency named in the scenario. A strong answer satisfies the stated need with the least fragile configuration and preserves the platform’s security and governance model. These are study heuristics, not additional official exam rules.
Extra concept: Treat configuration as a system rather than a list of settings. A field may participate in a formula, validation rule, flow, report, permission set, Lightning page, and deployment package at once. Before changing it, identify those dependencies, decide which are intentional, and test the full user journey. This dependency view explains why a locally correct change can still fail after deployment or for a different user. Use Setup dependency tools where available, source comparison, and focused regression checks; no single dependency view is guaranteed to reveal every runtime relationship.
Verification exercise: Turn the lesson scenario into a small test plan before calling the design complete. Record the starting data, the user permissions, the expected outcome, and the evidence that would confirm it. Test one ordinary path, one boundary condition, and one invalid or unauthorized path. Compare the result with the business requirement, then repeat after deployment in the target environment. This catches differences caused by data, activation, page assignment, sharing, or dependencies. Save the test cases with the release notes so another administrator can reproduce the result.
Key takeaways
- Given a scenario, determine options and considerations for importing and exporting data, including external data sources
- Choose the platform feature that fits the requirement, data, access, and operating context.
- Test realistic users, exceptions, dependencies, and lifecycle effects before release.
- Official references: https://help.salesforce.com/s/articleView?id=005298964&language;=en_US&type;=1 https://help.salesforce.com/s/articleView?id=sf.data_import_wizard.htm&type;=5 https://help.salesforce.com/s/articleView?id=sf.external_data_sources.htm&type;=5 Salesforce Certified Platform App Builder
- SU ·26 guide alignment 12
