Most RevOps teams blame their forecasting model when numbers are off. Usually it's not the model — it's the data feeding it. Here are the 3 fields that cause the most damage, and the fastest fix for each.

1. Duplicate Company Records

Every duplicate splits a company's activity across two records. Your AE sees "$12k open pipeline" on one record and misses the $40k sitting on the duplicate. Fix: run a domain-match dedupe pass (not just name-match — "Acme Inc" vs "Acme, Inc." will never catch on name alone) before every forecast cycle, not after.

2. Stale "Close Date" fields

If close dates aren't getting pushed as deals slip, your pipeline coverage math is lying to you every single week. Fix: set a hygiene rule that any deal with a close date more than 7 days in the past auto-flags for AE review. Don't let stale dates silently roll into "this month's forecast."

3. Inconsistent Lead Source / Original Source tagging

When reps manually pick lead source from a dropdown, you get 6 versions of "referral" (Referral, referral, Ref, Client Referral, etc.). This kills your channel attribution and makes your CAC-by-channel numbers fiction. Fix: lock lead source to be set only by automation/integration, never manual entry.

The pattern across all three: hygiene isn't a one-time cleanup project, it's a recurring rule. Pick one of these, audit it this week, and see how much your pipeline number moves.

Happy to go deeper on any of these — what's the messiest field in your CRM right now?