Data strategy = governance, quality, integration, retention, analytics — all coordinated.
Components:
1. Data ownership.
Per object / domain: who owns the data? Data Steward role per domain.
2. Master Data Management.
- Single canonical version per entity (Customer, Product, Employee).
- Salesforce as MDM, or external MDM tool.
- Cross-system reconciliation.
3. Data quality.
- Validation rules.
- Duplicate management.
- Cleansing schedule.
- Quality dashboards / scorecards.
4. Data lifecycle.
- Retention per data type.
- Archive strategy (Big Object / external).
- Deletion / anonymisation workflows.
- Compliance with regulations.
5. Data integration.
- System-of-record per object.
- Sync patterns.
- Conflict resolution.
- Audit trail.
6. Data classification.
- Sensitivity levels.
- Compliance categories.
- Drives encryption / sharing decisions.
7. Analytics architecture.
- Operational reporting in Salesforce.
- Cross-system in warehouse + BI.
- Real-time vs batch.
- AI / ML inputs.
8. Data governance.
- Council with data stewards.
- Standards documentation.
- Change request process for schema.
- Data quality reviews.
9. Talent.
- Data engineers for warehouse.
- Data architects for strategy.
- Salesforce admins for org-level.
Architectural artefacts:
- Data dictionary — every metric / field defined.
- Lineage diagram — where data flows.
- Data model diagrams — structures.
- Quality scorecard — metrics.
- Retention matrix — what's kept how long.
Common pitfalls:
- No ownership — data drifts without accountability.
- Quality unmeasured — can't improve what you don't measure.
- No retention policy — data accumulates indefinitely.
- Operational and analytical mixed — Salesforce trying to do both poorly.
- Manual reconciliation — should be automated.
Senior insight: data strategy is foundational architecture. Without it, every project is fighting symptoms. With it, projects build on a strong base.
Treat data strategy as a separate workstream, not a sub-component of any project.