03 - Data Management¶

Importing data¶

Decision flowchart¶

Volume?
├─ < 50,000 records → Data Import Wizard (browser, simple)
├─ < 5 million records → Data Loader (desktop or CLI)
└─ > 5 million → Bulk API directly (programmatic)

Object?
├─ Account, Contact, Lead, Solution, Custom → Both wizards work
└─ Other standard objects (Opportunity, Case, Campaign, etc.) → Data Loader only

Data Import Wizard¶

• Setup → Data Import Wizard
• 50,000-record limit per file
• Insert / Update / Upsert
• Match by Salesforce ID, Name, External ID, or Email
• No relationships across multi-step imports

Data Loader¶

• Desktop app (Windows / Mac) or CLI (apex_data_loader_cli)
• All standard and custom objects
• Insert / Update / Upsert / Delete / Hard Delete / Export / Export All
• 5 million records or 10 GB (theoretical)
• CSV files
• Field mapping UI
• Saves successful and error files

Bulk API¶

• REST or SOAP API
• For >5M records or programmatic integrations
• Asynchronous; submit job, poll status, retrieve results
• 50,000 batches/day limit (configurable)

Upsert¶

Match existing records by an External ID field; update if matched, insert if not. Critical for data integrations.

Requires:

• A field marked as External ID
• Field marked as Unique (recommended)
• The CSV column containing the matching value

Saves you from "do I update or create?" logic in your ETL.

Mass operations¶

Mass Transfer¶

• Setup → Data Management → Mass Transfer Records
• Available for Lead, Account, custom objects (not all standard)
• Transfer ownership in bulk based on criteria

Mass Update¶

• Through list views with inline editing (limited)
• For real bulk updates, use Data Loader Update

Mass Delete¶

• Setup → Data Management → Mass Delete Records
• For Lead, Contact, Account, Activity, Product, Report
• Soft delete (goes to Recycle Bin); Hard Delete option in Data Loader

Recycle Bin¶

• Soft-deleted records go here for 15 days
• Recover via Recycle Bin or UNDELETE in Data Loader (Hard Delete bypasses Recycle Bin)

Duplicate management¶

Matching Rules¶

Define how to compare records:

• Match on Email, Phone, Name (fuzzy or exact)
• Built-in for Account, Contact, Lead
• Custom matching rules for custom objects (with limitations)

Duplicate Rules¶

Use a matching rule + define action:

• Allow with alert (let user save, show warning)
• Block save (reject creating a duplicate)
• Audit (log but allow)

Duplicate Jobs (paid feature)¶

Periodic duplicate scans on existing data. Available with Data.com or Duplicate Check package. Built-in dupe management is for prevention, not bulk dedup of existing data.

Data quality tools¶

Validation Rules¶

Block save if a formula returns true. Display custom error message.

Examples:

• "Close Date must be in the future when Stage is Open"
• "Phone must be valid US format"
• "Discount cannot exceed 30% without manager approval"

Required fields¶

• Universally Required - required regardless of profile (set on field definition)
• Page Layout Required - required when editing through the layout (per profile via layout)
• Validation Rule - "field must not be blank" formula

Backup and export¶

Weekly Export (Free)¶

Setup → Data Management → Data Export.

• Schedule weekly or monthly
• Generates ZIP of CSVs (1 per object) emailed to admin
• Free, slow, manual restore if needed
• Not a true backup product (no point-in-time restore)

Salesforce Backup & Restore (Paid)¶

Salesforce-native backup product. Daily snapshots, point-in-time restore, retention controls.

Third-party backup tools¶

OwnBackup (now Own), Spanning, Druva, etc. Common for enterprises.

Sandboxes¶

Identical-schema replicas of production for development / testing.

When to use¶

• Developer / Developer Pro - unit / integration testing for individual devs
• Partial Copy - QA / UAT with representative data, faster refresh than Full
• Full - final UAT before prod deploy, training, performance testing

Sandbox naming¶

<sandbox-name>.sandbox.my.salesforce.com - separate URL from prod.

Deployment options¶

Change Sets¶

• Inbound + outbound between connected sandboxes / prod
• Drag-drop UI for selecting metadata to deploy
• Limited to specific component types
• Slow for large deployments

Salesforce CLI / SFDX¶

Modern dev workflow:

• Source format (decomposed XML for git-friendliness)
• Scratch orgs (ephemeral dev orgs)
• Deploy via sf project deploy start
• CI/CD via GitHub Actions / Azure DevOps / etc.

Unmanaged / Managed Packages¶

• Unmanaged - distributable but recipient owns the metadata
• Managed - AppExchange products with versioning, namespace, IP protection

Storage¶

Two types:

• Data Storage - records (Account, Contact, etc.)
• File Storage - attachments, documents, content versions

Per-edition allowances¶

• Enterprise: 10 GB Data + 10 GB File for org + per-user additions
• Unlimited: more

Monitoring¶

Setup → Storage Usage. Per-object storage breakdown.

Storage cleanup¶

• Delete old / archived records
• Move attachments to external storage (S3, OneDrive) via Files Connect
• Hard Delete to bypass Recycle Bin (frees storage immediately)
• Big Objects for archival data (millions of records, async query only)

Common exam triggers¶

• "Mass insert 250,000 Opportunities with parent Account lookup" → Data Loader (Wizard caps at 50K and doesn't support Opportunity)
• "Match incoming records by SAP ID" → External ID + Upsert via Data Loader
• "Detect duplicate Leads at creation time" → Matching Rule + Duplicate Rule (Block or Alert)
• "Restore data deleted 5 days ago" → Recycle Bin (15-day retention)
• "Permanently delete with no Recycle Bin retention" → Data Loader Hard Delete
• "Block save unless region field is filled" → Validation Rule (or Universally Required if always required)
• "Sandbox for UAT with representative data" → Partial Copy Sandbox
• "Final UAT before prod deploy" → Full Sandbox
• "Move large data archive out of normal storage" → Big Object
• "Schedule weekly backup of all data to CSV" → Data Export (Weekly Export)