Your forecast is wrong. Not slightly wrong—wrong enough to miss board guidance, stall hiring plans, and blow commission calculations. At ₹50M ACV and above, three systematic forces inflate pipeline by 30–50%, and most sales leaders never catch them until the quarter closes. We worked with a ₹200M ARR company that discovered ₹100L in attributed pipeline. After a brutal audit, ₹60L evaporated: shared deals counted twice, three-month-old stalled deals still forecasted, and unqualified leads sitting at 60% probability. What remained—₹40L of real, owned, active deals—was half the reported number. Their forecast error was 60%. This isn't incompetence. It's structural. Here's what corrupts large forecasts and how to fix it. Force 1: Shared pipeline counts deals at 200% Large deals don't live in one rep's pipeline. An enterprise deal involves an account executive, a sales development rep who sourced it, a sales engineer, and often a partner channel rep. When your CRM lets each person own the deal independently, you don't get four views of one deal—you get one deal counted four times. A ₹3Cr ARR deal sits in the AE's pipeline at 100% forecast. The SDR also marks it as 100% (they 'own' the early stage). The SE tags it 100% (their project). The partner counts it 100% (their pipeline). Forecast shows ₹12Cr. Reality shows one deal. The damage compounds in deal-registration workflows. A partner brings a customer to you; you mark them as partner-sourced but also create a house account. The deal is now in two pipelines, each forecasted independently. Your CRM doesn't know which rep's count is canonical. How to fix shared ownership Implement a single source of truth for deal ownership. In Orin's CRM , this means: One primary owner per deal. Not shared ownership. Not collaborative deals. One AE owns ₹50M+ deals; that AE's forecast is the only forecast that counts. SDRs, SEs, and partners are linked contacts, not co-owners. Deal-split rules for partner deals. If a partner sourced 50% of value, the deal splits: 50% to the partner's pipeline, 50% to your house account. No deal appears at 100% in two places. Use Orin's pipeline tracking to enforce the split at entry, not in a spreadsheet after close. Deal-routing discipline. The moment an SDR qualifies a lead, it transfers to the AE. The deal leaves the SDR's pipeline. No co-ownership, no double-counting. This feels slow; it's not. Clarity beats consensus forecasts every time. Shared ownership doesn't improve forecast accuracy—it inflates it. A deal owned by four people is owned by none of them. Force 2: Phantom deals—stalled for 90 days but still forecasted A deal enters your pipeline at 40% probability. Three months pass. No discovery call. No follow-up. No activity. The deal sits in 40% forecast, month after month. You report ₹10Cr in forecast. The rep tells you it's 'still active.' Your board thinks ₹10Cr is closing this quarter. Nothing closes. This happens because your CRM defines probability (or stage) but not recency. A deal can be 'in negotiation' with zero activity in 90 days. It's phantom—still on the books, but not real. Large deals amplify this. A single ₹5Cr deal that's stalled is hard to admit to and easy to rationalize ('the buyer's on vacation,' 'budget cycles in Q3'). That one deal inflates forecast by 2%. Multiply by 15 phantom deals and you're 30% high. How to fix phantom deals Define 'active' by activity, not stage. A deal is active if it had an activity (call, meeting, email, update) in the last 30 days. If not, it moves to 'stalled' automatically, and forecast drops to 0% or a floor you set (5%, if you want to be conservative). Use Orin's automation rules to enforce this: no manual override, no exception pleading. Set deal-age thresholds by deal size. A ₹50L deal can sit quiet for 45 days; a ₹5Cr deal should trigger a hard conversation at 30 days inactive. Create automations that flag old, high-value deals for review or closure. Compress quarterly forecast to days, not months. At close, include only deals with activity in the last 14 days. Everything else is 'pipeline' and forecasted separately. This forces reps to actively manage their forecast and kills the 'maybe in Q4' excuse. Archive wins and losses weekly, not at month-end. The moment a deal closes or is lost, it leaves the active pipeline. This keeps forecast clean and forces the team to fill the pipe with new deals, not carry dead weight. Force 3: Unqualified deals weighted at 60%+ probability A rep adds a prospect to the pipeline at 60% because they 'sounded interested' on a cold call. No budget check. No timeline. No champion identified. The prospect is a title-only contact at a company of 5,000. This deal is nowhere near close; it's barely a lead. But at 60%, it's forecasted at ₹50L (40% × ₹1.25Cr deal size). Multiply that across a team: 30 reps × 3 half-qualified deals × ₹50L = ₹45Cr forecast on prospects who've never been qualified. Real close rate: 1–2%. Real expected value: ₹45–90L. The root