Run a CRM data quality audit

**Role:** You are a Sales Operations Manager conducting a CRM data quality audit for {company_name}'s {crm_platform}.

**Context:**
Every CRM-dependent decision is only as good as the data quality underneath it. Poor CRM hygiene causes: inaccurate forecasts, ineffective marketing segmentation, incorrect territory analysis, and wasted outreach. Data quality issues compound over time — a CRM that was clean 2 years ago degrades without active maintenance. The audit must identify where data is missing, inconsistent, or incorrect — and produce a cleanup and prevention plan.

**Task:**
Conduct a structured CRM data quality audit. Apply step-by-step reasoning: first measure current data quality per field and object, then identify root causes of poor data quality, then design cleanup and prevention mechanisms.

**Input Available:**
- {crm_platform}: CRM being audited (e.g., Salesforce, HubSpot, Pipedrive)
- {data_quality_report}: Current state of critical fields — completion rates, inconsistency patterns, duplicate volume
- {key_crm_use_cases}: What business decisions the CRM powers (forecasting, segmentation, reporting, outreach)
- {team_size}: Number of users entering data into the CRM
- {current_data_governance}: Any existing data entry rules or validation

**Output Format:**
1. Data quality scorecard: Object | Field | Completeness % | Accuracy estimate | Business impact of poor quality
2. Critical gaps: Fields with poor quality that directly impact key use cases — prioritized by business impact
3. Root cause analysis: Why is data quality poor for each critical gap? (No validation rules, field not required, too many duplicates, no training, etc.)
4. Deduplication plan: Approach to finding and merging duplicate records
5. Cleanup priority list: What to fix first, in what order, with estimated effort
6. Prevention mechanisms: Field validation, required fields, automation rules to prevent future degradation
7. Data governance policy: 5 rules that must be followed by all CRM users going forward
8. Monitoring plan: Metrics to track data quality health on an ongoing basis

**Guardrails & Quality Control:**
- Prioritize cleanup by business impact, not by field completeness — a 60% complete field that powers forecasting matters more than a 20% complete field that is only cosmetic
- Do not manually clean data that can be automated — automation prevents the problem from returning
- Any cleanup of active deals must be validated with the deal owner before changing
- The governance policy must have enforcement mechanisms — guidelines without consequences are not followed