Your Pipedrive forecast looked clean at 20 deals. Then you hit 50. Now your VP of Sales is asking why forecast was 30% too high for three quarters straight, and your CFO is rebuilding the revenue model in a spreadsheet because your pipeline forecast has become indistinguishable from noise. This is not user error. Pipedrive's forecast logic—multiplying deal value by stage probability—assumes deal progression is linear, independent, and honest. At 50+ deals, that assumption collapses. Shared deals, stage-to-stage friction, and reps inflating pipeline all compound into systematic over-forecast. If you're planning a migration to a more mature CRM , or you just need your forecast to tell you the truth, here's where the math breaks and what to fix. The weighted forecast formula breaks on three inputs Pipedrive's forecast is straightforward: Deal Value × Stage Probability = Forecast Revenue . On paper, it works. In practice, three patterns break it by 25–35%. 1. Shared deals hide duplicate probability A deal assigned to two reps (or a team) counts its full value into both forecasts. In Pipedrive, that deal is counted twice. A ₹50 lakh deal split between your Account Executive and your Solutions Engineer appears as ₹100 lakh in forecast. This is not a bug—it's a design choice—but it's lethal when shared deals represent 15–25% of your pipeline at scale. Audit step: Export your deals and count records where assigned_to_user_id is an array or where the same deal appears in multiple reps' forecasts. If more than 10% of your pipeline value is shared, your forecast is already 10%+ too high before any other adjustments. 2. Long cycle deals age without stage movement A deal in 'Negotiation' (70% probability) for four months is not a 70% close. It's a deal stuck. Pipedrive's forecast engine doesn't measure deal age—it only reads stage. So a deal at 70% probability stays at 70%, whether it's been there two weeks or four months. Real close rates drop sharply after 60 days without movement. Audit step: Pull the last_activity_date for each deal. Deals inactive for more than 60 days in any stage should be discounted by 20–30%. A ₹50 lakh deal in Negotiation for 90 days is not a ₹35 lakh forecast; it's closer to ₹10–15 lakh. 3. Rogue stage definitions inflate early-stage deals Sales reps optimize for optics. If your 'Qualified' stage carries 40% probability and a rep has no visibility, they move deals there to keep forecast healthy. Over time, early-stage probability creeps upward. Deals that used to stay in 'Prospect' for two months now move to 'Qualified' in two weeks, padding forecast without improving close rates. Audit step: For each stage, measure the actual close rate of deals that entered that stage in the past 12 months. Compare this to your assigned stage probability. If your 'Qualified' stage shows 40% assigned probability but only 25% historical close rate, your forecast is 60% too optimistic at that stage alone. The 30% disconnect: where historical close rates diverge from stage probability Run this comparison: take deals closed in the past 12 months, group them by the stage they were in 30 days before close, and calculate the actual percentage that closed. Then compare that number to the probability you've assigned to that stage in your forecast settings. In mature Pipedrive deployments we've audited, the divergence is consistent: Prospect (20% assigned): Historical close rate 8–12% (30–40% over-forecast) Qualified (40% assigned): Historical close rate 25–28% (30–40% over-forecast) Negotiation (70% assigned): Historical close rate 45–55% (25–35% over-forecast) Proposal (85% assigned): Historical close rate 60–70% (20–30% over-forecast) The root cause is not rep dishonesty—it's that stage progression is not uniform. A deal in Negotiation has a 70% probability of closing if it moves . But 30–40% of Negotiation deals stall, age out, or revert. Pipedrive's formula doesn't penalize age; it only penalizes the reps who don't move deals. Three cleanup steps before you export and migrate If you're moving to a more configurable CRM like Orin or Salesforce, or you just want forecast to tell the truth, run these audits first: Step 1: De-duplicate shared deals Identify deals assigned to multiple team members. For each, decide: does it belong to the AE, the team, or should it be split? Once you decide, remove it from secondary reps' pipelines. Export a list of these deals—keep it for your migration checklist. Step 2: Age-adjust forecast Pull add_time (deal creation date) and last_activity_date for each deal. Calculate days in current stage. Apply a discount: Days in stage 0–30: 100% of forecast value Days in stage 31–60: 80% of forecast value Days in stage 61–90: 50% of forecast value Days in stage 90+: 20% of forecast value This single adjustment typically reduces forecast by 15–20% and makes trend-spotting real: you'll see which stages have aging problems. Step 3: Recalibrate stage probability Pull closed deals from the past 12