sf-agent-framework

# Salesforce Data Quality Standards ## Overview Data quality standards ensure that Salesforce data is accurate, complete, consistent, and fit for its intended use, enabling reliable business decisions and operations. ## Data Quality Dimensions ### Accuracy **Definition**: Data correctly represents real-world entities and events **Standards**: - Email addresses follow valid format - Phone numbers include correct digits - Addresses validated against postal databases - Dates within reasonable ranges - Numeric values within expected bounds **Measurement**: - Error rate per field - Validation rule failures - Data verification audits - User-reported inaccuracies ### Completeness **Definition**: All required data is present **Standards**: - Mandatory fields always populated - Key business fields >95% complete - Related records properly linked - No orphaned records - Historical data preserved **Measurement**: - Field population percentage - Required field compliance - Relationship integrity - Missing data reports ### Consistency **Definition**: Data is uniform across the system **Standards**: - Same format for similar data types - Standardized picklist values - Consistent naming conventions - Uniform date/time formats - Matching data across objects **Measurement**: - Format compliance rate - Cross-object data matching - Duplicate detection rate - Standardization scores ### Timeliness **Definition**: Data is current and available when needed **Standards**: - Real-time updates for critical data - Batch updates within SLA - Last modified dates tracked - Stale data flagged - Timely synchronization **Measurement**: - Data age analysis - Update frequency - Synchronization lag - SLA compliance ### Uniqueness **Definition**: No unwanted duplicates exist **Standards**: - Unique identifiers enforced - Duplicate rules active - Matching rules configured - Merge procedures defined - Master record designation **Measurement**: - Duplicate rate - Merge frequency - Unique constraint violations - Matching rule effectiveness ### Validity **Definition**: Data conforms to defined formats and business rules **Standards**: - Field formats enforced - Business rules validated - Referential integrity maintained - Value ranges respected - Conditional logic applied **Measurement**: - Validation rule pass rate - Format compliance - Business rule violations - Constraint adherence ## Data Standards by Object Type ### Account Standards - **Naming Convention**: Legal entity name, no abbreviations - **Required Fields**: Name, Type, Industry, Owner - **Address Format**: Complete postal address - **Phone Format**: +1 (XXX) XXX-XXXX - **Website**: Valid URL with protocol - **Hierarchy**: Parent account relationships - **Duplicates**: Checked by name and domain ### Contact Standards - **Name Format**: First Name, Last Name (no nicknames) - **Email**: Valid format, verified domain - **Phone**: Primary number required - **Title**: Standardized job titles - **Account**: Must be associated - **Status**: Active/Inactive tracking - **Duplicates**: Email and name matching ### Opportunity Standards - **Naming**: Account - Product - Year - **Amount**: Required, positive values - **Close Date**: Future date for open opps - **Stage**: Valid progression only - **Probability**: Auto-calculated by stage - **Products**: Line items for won opps - **Competition**: Tracked for analysis ### Case Standards - **Subject**: Clear, descriptive title - **Priority**: Based on SLA rules - **Status**: Proper lifecycle flow - **Owner**: No unassigned cases - **Contact**: Associated customer - **Description**: Minimum 50 characters - **Resolution**: Required for closed cases ### Lead Standards - **Source**: Always populated - **Status**: Regular progression - **Rating**: Based on scoring model - **Company**: No personal emails - **Conversion**: Mapped fields defined - **Assignment**: Automated rules - **Duplicates**: Email and company check ## Data Entry Standards ### Field-Level Standards - **Text Fields**: Proper capitalization - **Picklists**: No "Other" without description - **Multi-Select**: Maximum selections defined - **Currency**: Correct currency code - **Percentages**: 0-100 range - **URLs**: Include protocol - **Lookups**: Valid relationships only ### Naming Conventions - **Accounts**: Legal name, suffix (Inc., LLC) - **Contacts**: Formal names, titles - **Opportunities**: Descriptive, searchable - **Campaigns**: Program - Type - Date - **Cases**: Issue - Product - Severity - **Products**: SKU - Name - Version ### Address Standards - **Format**: Consistent across records - **Validation**: Postal service verification - **Geocoding**: Latitude/longitude for mapping - **Billing/Shipping**: Both when applicable - **Country**: ISO codes used - **State/Province**: Standard abbreviations ## Data Validation Rules ### Validation Rule Categories - **Format Validation**: Email, phone, URL - **Range Validation**: Dates, numbers - **Conditional Validation**: Based on other fields - **Cross-Object Validation**: Related data - **Business Rule Validation**: Process requirements ### Common Validation Patterns ``` // Email validation REGEX(Email, "^[a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\.[a-zA-Z]{2,}$") // Phone validation (US) REGEX(Phone, "^\+?1?[-.]?$?([0-9]{3})$?[-.]?([0-9]{3})[-.]?([0-9]{4})$") // Required field combination OR( ISBLANK(Field1__c), ISBLANK(Field2__c) ) // Date range validation CloseDate > TODAY() && CloseDate < TODAY() + 365 // Conditional requirement AND( ISPICKVAL(Type, "Customer"), ISBLANK(Contract_Date__c) ) ``` ## Data Governance ### Data Stewardship - **Data Owner**: Business accountability - **Data Steward**: Operational responsibility - **Quality Monitor**: Regular audits - **Process Owner**: Workflow maintenance - **Technical Support**: System administration ### Quality Processes - **Initial Load**: Cleansing before import - **Ongoing Entry**: Validation at creation - **Batch Updates**: Quality checks - **Integration**: Error handling - **Periodic Review**: Scheduled audits ### Data Quality Tools - **Duplicate Management**: Native Salesforce - **Data Loader**: Batch operations - **DemandTools**: Advanced cleansing - **Cloudingo**: Deduplication - **Data.com**: Third-party enrichment ## Monitoring and Measurement ### Quality Metrics - **Completeness Score**: % fields populated - **Accuracy Rate**: Validated data percentage - **Duplicate Rate**: Duplicate records found - **Timeliness**: Data age distribution - **Compliance Rate**: Standards adherence ### Quality Dashboards - **Executive Dashboard**: Overall health - **Object Scorecards**: Per-object quality - **Trend Analysis**: Quality over time - **User Metrics**: Quality by creator - **Process Metrics**: Workflow effectiveness ### Audit Procedures - **Scheduled Reviews**: Monthly/quarterly - **Sample Testing**: Statistical sampling - **Full Scans**: Periodic complete review - **Exception Reports**: Anomaly detection - **User Feedback**: Quality issue reporting ## Data Cleansing ### Cleansing Strategies - **Standardization**: Format consistency - **Deduplication**: Merge duplicates - **Enrichment**: Add missing data - **Validation**: Verify accuracy - **Archival**: Remove obsolete data ### Cleansing Process 1. **Assessment**: Current state analysis 2. **Planning**: Cleansing approach 3. **Backup**: Data preservation 4. **Execution**: Cleansing operations 5. **Validation**: Quality verification 6. **Documentation**: Changes recorded ### Common Cleansing Tasks - Remove duplicate records - Standardize company names - Fix phone number formats - Validate email addresses - Complete missing fields - Update outdated information - Correct data relationships - Archive old records ## Integration Data Quality ### Integration Standards - **Field Mapping**: Documented and validated - **Data Transformation**: Consistent rules - **Error Handling**: Graceful failures - **Sync Validation**: Bi-directional accuracy - **Conflict Resolution**: Clear precedence ### API Data Standards - **Required Fields**: Enforced in API - **Format Validation**: Pre-submission - **Bulk Operations**: Quality preserved - **Error Reporting**: Detailed messages - **Retry Logic**: Transient failure handling ## Best Practices ### Prevention Strategies 1. **User Training**: Proper data entry 2. **Validation Rules**: Enforce at entry 3. **Page Layouts**: Guide users 4. **Automation**: Reduce manual entry 5. **Integration**: Single source of truth ### Quality Culture 1. **Executive Support**: Leadership commitment 2. **Clear Standards**: Documented rules 3. **Regular Training**: Ongoing education 4. **Accountability**: Quality ownership 5. **Recognition**: Quality achievements ### Continuous Improvement 1. **Regular Monitoring**: Track metrics 2. **Root Cause Analysis**: Understand issues 3. **Process Enhancement**: Improve workflows 4. **Tool Optimization**: Better technology 5. **Feedback Integration**: User input ## Compliance Considerations ### Regulatory Requirements - **GDPR**: Data accuracy and consent - **CCPA**: Consumer data rights - **HIPAA**: Healthcare data standards - **SOX**: Financial data integrity - **Industry**: Specific regulations ### Compliance Standards - **Audit Trail**: Change tracking - **Data Retention**: Defined periods - **Access Control**: Appropriate permissions - **Encryption**: Sensitive data protection - **Reporting**: Compliance documentation