Your forecast says you'll close $2.3M next quarter. Your CFO asks for the confidence level. Your VP of Sales squints at the pipeline, sees deals stuck in negotiation for three months, another that hasn't had a note since June, and a third with $500K attached but no contact name. The forecast isn't wrong—it's incomplete. And incomplete forecasts don't just miss targets; they trigger bad hiring, misspent budget, and missed board calls. The problem isn't that your reps are lazy. It's that pipeline updates feel like busywork when a deal is actually moving. A rep closes a call, sends a proposal, gets a thumbs-up from their contact's boss, and moves on to the next pitch. Updating the CRM feels like homework. So the deal lives in Proposal for six weeks, the rep knows it's closing in two, but the system doesn't. Your forecast model sees Proposal , applies a 50% close probability, and inflates the number. When the deal does close—or stalls—the forecast recalibrates too late. The fix isn't more nagging. It's visibility into where your forecast is breaking, automated nudges that interrupt before deals go dark, and mandatory field checks that make the CRM the source of truth instead of optional decoration. The four metrics that reveal forecast rot Before you can fix the pipeline, you need to see where it's broken. These four metrics expose deals hiding in stalled stages: 1. Deal age by stage How long has a deal been sitting in Proposal ? In Negotiation ? A deal that spends 60 days in Proposal when your average is 10 is either dead, forgotten, or genuinely waiting. You need to know which. Run a query against your CRM—group deals by stage, calculate the median dwell time per stage, then flag anything in the 90th percentile. A deal stuck for 120 days in a stage where the median is 14 days is the forecast blind spot. 2. Last activity timestamp A deal can look alive in the pipeline but be completely dead if the last email, call, or meeting note is six weeks old. Pull a report: for every deal above a certain size threshold (say, $25K ACV), show the date of the most recent activity. Anything older than 21 days is a candidate for a rep conversation. Anything older than 45 days is almost certainly forecast fiction. 3. Stage progression velocity Some reps move deals through stages fast; others let them marinate. This isn't necessarily bad—complex deals take longer—but it's a leading indicator of stalling. Compare average days-in-stage across reps. If rep A gets deals from Discovery to Proposal in 8 days, and rep B takes 45 days, the difference might be deal complexity or territory. Or rep B might just not be moving deals. Flag the outliers and ask the rep to explain. The conversation is more useful than the forecast adjustment. 4. Missing mandatory fields A deal with no decision date, no primary contact, or no next step is a deal that's forecast guesswork. Count deals by stage that are missing critical fields: close date, contact name, or a note from the last week. If more than 10% of deals in your Proposal stage have no close date, your forecast for that stage is unreliable by definition. Map your pipeline audit: a 30-minute template Run this audit once a quarter. It takes 30 minutes and exposes where your forecast is most vulnerable. Export or query your open deals. Include: deal name, amount, stage, created date, last activity date, close date (if filled), and rep name. Calculate stage dwell time. For each deal, subtract the stage entry date from today. A deal entered Proposal on Oct 1; today is Dec 15; dwell time is 75 days. Flag outliers by stage. What's the 90th percentile of dwell time for Proposal ? For Negotiation ? Any deal above that threshold gets a red flag. Cross-reference last activity. For flagged deals, check the last activity date. If it's more than 30 days old, the deal is likely stalled, not maturing. Audit mandatory fields. Count deals missing close date, contact name, or a recent note. If it's more than 10% per stage, your forecast model is downstream of dirty data. Interview the rep. For deals in the top 10% of dwell time and/or oldest last activity, ask the rep: Is this deal alive? When do you actually expect to close? What's the next step? The answer often doesn't match the CRM. A forecast is only as good as the data behind it. Incomplete data doesn't make the forecast conservative—it makes it wrong. Automated nudges that force pipeline hygiene Now that you've audited the pipeline, you need to stop it from rotting again. Automation is your best lever because it doesn't require enforcement—it just interrupts the rep at the moment of forgetfulness. Daily digest of deals needing updates Use your CRM's automation layer (or no-code automation tools ) to run a daily query: deals that are older than X days in a given stage with no activity in the past 7 days. Send the rep an automated daily digest flagging these deals. Not a nagging reminder—a factual list: Deal: Acme Corp | Stage: Proposal | Days in stage: 67