How do we measure and improve forecast accuracy beyond activity metrics?

Question

Pulse RevOps · The Machine · Accepted Answer

## Direct Answer Forecast accuracy = deal age + rep history + pipeline composition. Track **3 tiers**: rep forecast vs actual (**65%+ target**), deal velocity (days-to-close), stage conversion rates. Red-line reps missing **75% attainment** for **2 quarters**—they're guessing. ## Operator Approach Activity metrics (dials, meetings, proposals) are vanity unless tied to **revenue outcome**. Real forecast accuracy requires **deal-level intelligence**. **Tier 1: Individual Rep Forecast Accuracy** Measure: **rep forecast in CRM vs actual close** per quarter - Target: **65–75%** (overforecasting is more common than underforecasting) - Red flag: **> 20% miss** indicates guessing or bad data - Ramp consideration: ramping reps typically forecast **± 30%**; mature reps **± 15%** Action: Reps missing forecast by **> 25%** for **2 consecutive quarters** → coaching or termination discussion **Tier 2: Deal Velocity & Stage Progression** Measure by segment (enterprise, mid-market, SMB): | Deal Stage | Enterprise (Days) | Mid-Market (Days) | SMB (Days) | Conversion Rate | |---|---|---|---|---| | Discovery | 14–21 | 7–14 | 3–7 | 40–50% | | Demo | 21–35 | 10–21 | 5–10 | 35–45% | | Proposal | 35–60 | 14–28 | 7–14 | 25–35% | | Negotiation | 21–45 | 7–14 | 3–7 | 60–80% | Red flags: - Deals stuck in demo for **> 6 weeks** (losing momentum) - Proposal acceptance **< 20%** (pricing/product mismatch) - Negotiation taking **> 6 weeks** (deal fragility) **Tier 3: Pipeline Composition Quality** Score pipeline by risk: - **Committed:** reps report **95%+** confidence, deal **< 14 days** out = **100% weight** - **Probable:** reps report **75–95%** confidence, **15–60 days** out = **50% weight** - **Possible:** reps report **40–75%** confidence, **> 60 days** out = **10% weight** - **Pipeline:** reps report **< 40%** confidence = **0% weight** Forecast = (Committed × 1.0) + (Probable × 0.5) + (Possible × 0.1) **Forecast accuracy diagnosis tree:** | Symptom | Root Cause | Fix | |---|---|---| | Large misses (±25%+) per rep | Rep guessing or data stale | Deal review coaching, daily updates | | Aggregate forecast wrong by 5–10% | Bad stage distribution | Validate stage rules, clean CRM | | Deals slipping between quarters | Velocity trending slower | Pipeline health review, resource gaps | | High closing rates but late discovery | Reps under-capturing early-stage | Sales process audit, activity enforcement | ## Mermaid: Forecast Accuracy Diagnostic Loop ```mermaid graph TD A["Weekly Forecast Review"] --> B{"Rep Forecast vs Actual > 20% Miss?"} B -->|Yes| C["Examine Pipeline: Deal Age"] C --> D{"Deals < 14d & Confirmed?"} D -->|Yes| E["Rep issue: coaching needed"] D -->|No| F["Pipeline health issue"] B -->|No| G{"Stage Conversion Rates Normal?"} G -->|No| H["Data quality: stage rules off"] G -->|Yes| I["Forecast trending OK"] E --> J["1-on-1 deal review with rep"] F --> K["Review deal velocity vs historical"] H --> L["Clean CRM; retrain on stages"] J --> M["Re-forecast after coaching"] K --> N["Identify bottleneck stage"] L --> O["Validate forecast accuracy"] ``` **Sources:** Pavilion Forecast Accuracy Benchmarks, Bridge Group Pipeline Health Study, SaaStr Sales Operations Best Practices TAGS: forecast-accuracy,pipeline-quality,deal-velocity,rep-coaching,stage-conversion,pipeline-composition,deal-intelligence

Deal Stage	Enterprise (Days)	Mid-Market (Days)	SMB (Days)	Conversion Rate
Discovery	14–21	7–14	3–7	40–50%
Demo	21–35	10–21	5–10	35–45%
Proposal	35–60	14–28	7–14	25–35%
Negotiation	21–45	7–14	3–7	60–80%

Symptom	Root Cause	Fix
Large misses (±25%+) per rep	Rep guessing or data stale	Deal review coaching, daily updates
Aggregate forecast wrong by 5–10%	Bad stage distribution	Validate stage rules, clean CRM
Deals slipping between quarters	Velocity trending slower	Pipeline health review, resource gaps
High closing rates but late discovery	Reps under-capturing early-stage	Sales process audit, activity enforcement

How do we measure and improve forecast accuracy beyond activity metrics?

Direct Answer

Operator Approach

Mermaid: Forecast Accuracy Diagnostic Loop

What does the score mean?