You're reviewing next quarter's forecast in Pipedrive. The number looks solid: ₹52M pipeline, weighted at 35% average stage probability, suggests ₹18.2M expected close. Three weeks later you land ₹9.8M. The ₹8.4M gap isn't market slip—it's in the math. Pipedrive's weighted forecast is conceptually correct: stage probability × deal value = weighted revenue . A ₹1M deal in 80% probability (closing stage) contributes ₹800K. A ₹1M deal in 20% probability (early pipeline) contributes ₹200K. The system forces discipline: it penalizes stage creep and rewards closure velocity. But at ₹50M ACV, three structural breaks corrupt the math so badly that your forecast becomes a fiction: Shared deals appear in multiple reps' pipelines and get weighted twice (or three times) Deal stage bloat—reps stuck deals in high-probability stages artificially—goes undetected Manual probability overrides ("this deal is 90% even though it's in our 40% stage") erase the stage-probability link that makes the formula work The result: forecast inflation of 15–40% depending on your deal routing discipline. Here's how to see it, measure it, and fix it. How Pipedrive's Weighted Math Actually Works Pipedrive calculates expected revenue per deal as: Expected Revenue = Deal Value × Stage Probability Your pipeline summary aggregates these: sum all expected revenues, divide by total deal value, and you get a weighted average probability. That's the 35% or 45% you see at the top of your forecast view. The math is sound because it assumes two things: No deal appears twice in the pipeline. Each ₹1M deal belongs to one rep, one pipeline, one probability weighting. Stage probability is mechanical, not overridable. A deal in the "Negotiation" stage inherits Negotiation's 70% close rate. A deal in "Prospecting" inherits Prospecting's 10% close rate. The stage defines the probability. Both assumptions break at scale. Here's why. Break #1: Shared Deals Get Weighted in Two Pipelines You close enterprise deals with multiple stakeholders. Sales closes the IT contract; your partner channel rep closes the MSP side deal; your customer success team locks in the renewal commitment. In Pipedrive, these are often three separate deals—but they reference the same customer and the same final close date. Worse: all three appear in their respective owner's pipeline. The weighted forecast sees three ₹500K deals at 60% probability (₹900K expected), when the true deal value is ₹1.5M at 60% probability (₹900K expected). The final number is the same, but your deal count is inflated—and more dangerously, your rep-by-rep forecast allocation is wrong. You think Rep A has ₹900K at risk; she actually has ₹300K, and ₹600K is shared with two other reps. When you're chasing ₹50M in quarterly closes, three-way splits across enterprise accounts become the rule, not the exception. Each rep thinks their deal is at-risk; the forecast suggests all three are equally exposed. In reality, when the customer decides to buy, they buy once, and one rep gets partial credit—if any. The forensic question: How many deals in your pipeline reference the same customer, same close date, same expected contract value? If you find five such clusters in a 50-deal pipeline, you're overstating pipeline by 8–12%. At ₹50M, that's ₹4M–₹6M of phantom revenue. Break #2: Deal Stage Bloat and the Probability Trap Your Pipedrive stage structure probably looks like this: Prospecting (10% probability) Qualified (25%) Demo Scheduled (40%) Proposal (60%) Negotiation (75%) Closing (90%) This is sane. The problem: reps hate moving deals backward. A deal in Negotiation that hits a compliance objection should slide back to Proposal. It doesn't—reps leave it in Negotiation and add a note: "Waiting on legal." Now you have a deal at 75% probability that has a 25% true close rate (blocked by legal review, no ETA). Multiply this across a 50-deal pipeline. Eight deals stuck in high-probability stages (Closing, Negotiation) due to hidden blockers. Your weighted forecast inflates by the gap between stated probability (75%) and true probability (35%). An ₹1M deal in this state contributes ₹750K to your forecast instead of ₹350K—a ₹400K phantom. At ₹50M ACV, if 15–20% of deals are stuck in elevated stages, you're inflating forecast by ₹7.5M–₹10M. The audit: For every deal in Negotiation or Closing, ask when it last moved stages. Deals stuck for 30+ days with no stage change are warning signs. Pull the deal notes and look for blockers (legal, procurement, budget review, board approval) that should have pushed the deal to an earlier stage. Break #3: Manual Probability Overrides Erase Stage Discipline Pipedrive lets you override stage probability. A deal in Prospecting (10%) can be manually set to 50% probability if the rep argues the customer is further along than the stage suggests. This is where forecast becomes opinion. One rep's 50% override is another rep's 15% reality. And Pipedrive can't see the difference. The math breaks because t