Audit a Messy CRM for Missing Required Data

You are a CRM data quality auditor who specializes in cleaning up the chaos at agencies that scaled faster than their ops.
Agency: [AGENCY_NAME] | CRM: [CRM_TOOL] | Records to audit: [RECORD_COUNTS] | Known problems: [KNOWN_PROBLEMS] | Critical fields per object: [CRITICAL_FIELDS] | Top 3 reports leadership trusts (or doesn't): [REPORTS_TRUST]
Produce a structured CRM data audit. Identify (1) missing required fields, (2) stale data, (3) inconsistent picklist usage, (4) orphaned records, (5) duplicate detection. For each finding, recommend whether to fix manually, bulk-update, or ignore.

- Audit must be runnable as a list of filtered views/queries in [CRM_TOOL] — give me the exact filter logic
- Prioritize findings by impact on forecasting and AM handoff (highest first)
- Distinguish "fix in next 7 days" from "fix in next quarter" from "living with it"
- Include a count target per finding (e.g., "deals missing close_date in open pipeline")
- Recommend automation/validation to prevent recurrence for the top 3 issues

Section 1: audit query plan (filter view name | object | filter logic | what it surfaces). Section 2: findings prioritization table (Finding | Records Affected | Impact | Fix Type | Owner | Deadline). Section 3: top 3 prevention rules to add.

• [AGENCY_NAME] — Your agency's name
• [CRM_TOOL] — HubSpot, Pipedrive, Salesforce, etc.
• [RECORD_COUNTS] — Approx counts (e.g., 1200 contacts, 340 deals, 280 companies)
• [KNOWN_PROBLEMS] — Anything you already know is broken
• [CRITICAL_FIELDS] — Fields that MUST be populated per object (MRR, close date, source)
• [REPORTS_TRUST] — Reports leadership uses or refuses to use

Agency: Beacon Digital | CRM: HubSpot | Records: 4500 contacts, 850 deals, 600 companies | Known: source field is free text with 80+ values, half of closed-lost deals have no reason | Critical fields: deal MRR, deal source, close date, AM owner | Reports: forecast (don't trust), source ROI (broken), AM capacity (manual spreadsheet)

**Audit Queries (HubSpot views to build)**
| View | Object | Filter | Surfaces |
|---|---|---|---|
| Open deals missing MRR | Deal | Pipeline=Sales AND MRR is empty AND Stage != New Lead | Forecast killers |
| Closed-lost missing reason | Deal | Stage=Closed Lost AND Lost Reason is empty | Win/loss blind spots |
| Deals missing AM owner | Deal | Stage=Closed Won AND AM is empty | Handoff failures |
| Source = free-text mess | Contact | Source is in [list of bad values] | Attribution chaos |
| Stale deals | Deal | Last activity >45 days AND Stage = Open | Ghost pipeline |

**Findings**
| Finding | Records | Impact | Fix | Owner | Deadline |
|---|---|---|---|---|---|
| Open deals missing MRR | 180 | Forecast 22% off | Bulk update via export+import | Sales Mgr | 7 days |
| Source field chaos | 3200 contacts | Source ROI report broken | Map free text to new picklist via workflow | Ops | 14 days |
| Stale deals >45 days | 95 | Pipeline coverage fiction | Manager review, close-lost the dead ones | Reps | 7 days |
| Closed-lost no reason | 120 | No learning loop | Backfill via rep interview, sample 30 | Sales Mgr | 30 days |

**Prevention:** (1) make MRR required to enter Proposal stage. (2) Convert Source to dependent picklist with no free text. (3) Workflow: if deal idle 30 days, alert owner + manager.

• Don't try to fix everything. Pick the 3 findings that break the report leadership actually uses and fix those first.
• Bulk-fix by export-edit-reimport for anything over 50 records. Manual fixing burns 10x the time.
• Every cleanup project must end with a validation rule, or you'll be doing the same audit in 6 months.