What is the operator checklist for a Salesforce health audit � specific field-by-field signals that the CRM is rotting from the inside?
Operator Checklist: Salesforce Health Audit — Field-by-Field Rot Signals
A rotting Salesforce org has six categories of decay: Opportunity field completeness, stage integrity, account/contact data quality, duplicate proliferation, automation debt, and adoption collapse. Each category has hard signals you can pull today with native reports. Treat anything below the thresholds below as a P1 fix.
---
THE DETAIL
Without regular assessments, orgs face "org decay" — a gradual decline in performance, user adoption, and data quality from ad-hoc changes and accumulated technical debt. Here's how to find it, object by object.
---
🔴 CATEGORY 1 — Opportunity Object (Pipeline Rot)
- Close Date integrity — Run a report: Opportunities > Close Date < Today, Stage ≠ Closed. More than 5% of open pipeline with a past close date means your forecast is fiction.
- Stage history gaps — Run reports checking for critical blank fields, specifically Stage History on Opportunities. If deals skip stages (e.g., jump from Prospecting to Negotiation with no stage-progression timestamp), MEDDPICC/Force Management has no data to run on.
- Amount volatility — Large deal size swings suggest scope reduction or expansion that reps haven't explained in the record.
- No Next Steps field — Missing Next Steps creates pipeline blind spots. If >20% of active opps have a blank Next Steps field, you're running a forecast off vibes.
- Forecast slippage — Repeated close date changes signal process gaps and future revenue loss. Flag any opp with 2+ close date pushes in a single quarter.
- Negative stage movement — Deals that move backward in the sales process are a stage-definition problem, not a rep problem. >3% backward movement = broken stage map.
---
🔴 CATEGORY 2 — Account / Contact Object (Master Data Rot)
- Duplicate accounts — If reps are creating multiple versions of the same contact or account, it signals a lack of proper deduplication processes. Run a duplicate report. >2% dupe rate = immediate dedup job.
- Blank Industry on Accounts — Kills segmentation, territory planning, and every ICP filter you run. Target: <5% blank.
- Stale contacts — 70% of CRM data deteriorates and becomes inaccurate annually. Any contact not touched in 18 months with no activity log is dead weight distorting your TAM math.
- Missing Lead Source — Missing Lead Source creates pipeline blind spots that make attribution models worthless. If Marketing can't tie pipeline to source, CAC payback calculations break.
---
🔴 CATEGORY 3 — User Adoption Signals
- Shadow spreadsheets — When team members maintain parallel records outside the CRM, it's a loud signal that something has gone wrong with your Salesforce data.
- Login cadence — Login History shows up to 20,000 records of user logins for the past six months. Any quota-carrying rep logging in fewer than 3x/week is not using SFDC as their system of record.
- Report graveyard — Archive or delete anything that hasn't been viewed in the past 90 days and is not tied to a scheduled report or workflow. 50+ orphaned reports = no reporting governance.
---
🔴 CATEGORY 4 — Automation & Config Debt
- Zombie flows/process builders — Review installed packages and integrations — are they actively used and up-to-date? Remove unused components to reduce complexity.
- Security Health Check score — Scores in the 90–100% range indicate strong alignment. 80–89% signals emerging gaps. Below 80% reflects multiple high-risk findings that could violate SOX §404 or GDPR Art 32 controls.
---
BENCHMARK TABLE
| Signal | Healthy | Yellow | 🔴 Red (Act Now) |
|---|---|---|---|
| Open opps w/ past close date | <3% | 3–8% | >8% |
| Blank Next Steps on active opps | <10% | 10–25% | >25% |
| Duplicate account rate | <1% | 1–3% | >3% |
| Blank Lead Source | <5% | 5–15% | >15% |
| Reps not logging in 3x/week | 0 | 1–2 | >2 |
| Security Health Check score | 90–100 | 80–89 | <80 |
| Pipeline coverage ratio | 4–5x | 3–4x | <3x — a typical coverage value is 3–4x to hit quota; anything below signals a problem |