What's the difference between expansion ARR and net new ARR for forecasting?

Question

Pulse RevOps · The Machine · Accepted Answer

**Net new ARR = new customer ACV + expansion from existing base. Expansion ARR = upsells + cross-sells + price increases only.** They move on different timelines: net new closes in weeks (forecast 2 months out), expansion takes months (forecast 4–6 months out). Confusing them kills your forecast.

**Net new ARR mechanics:**
- New logo closes, goes live this month → books this month, but **CAC payback needs 6–12 months** of cash.
- If you forecast on a monthly or even quarterly basis, new logos create lumpy revenue. Expect ±30% variance.
- Most common mistake: forecasting net new linearly. A single $500K deal hitting in March doesn't mean $500K every month.

**Expansion ARR mechanics:**
- Customer already live. Seat expansion, tier upgrade, add-on module → books **2–4 weeks after signature**, not 3 months.
- More predictable than net new. **Net retention (NRR) % tells you expansion run rate.** At 120% NRR, a $10M cohort generates $2M in expansion ARR annually.
- Expansion has **70–90% close rate** (existing relationship); net new closer to 20–30%.

**Forecast split in practice:**

1. **Net new forecast:** Pipeline (weighted by stage %) × historical win rate. Update monthly. Assume new logos slip 2–3 weeks.
2. **Expansion forecast:** NRR % × ARR from customers 3+ months old. Carve out multi-year deals (less predictable expansion).
3. **Mix ratio matters.** At $5M ARR, maybe 60% is net new (early growth). At $50M, maybe 60% is expansion (machine working).

| Driver | Velocity | Variance | Predictability |
|--------|----------|----------|----------------|
| Net New | 2–6 weeks | ±30% | 20–30% win |
| Expansion | 2–4 weeks | ±10% | 70–90% win |

**Action:** Build two separate pipelines. Run net new forecast as Bayesian (buckets by stage). Run expansion as cohort-vintage model (" what % of 12-month cohort expanded by month 6?").

```mermaid
graph LR
  A[ARR Pipeline] --> B[New Logos]
  A --> C[Expansion]
  B --> B1[Weeks to Close: 6-10]
  B --> B2[Win Rate: 20-30%]
  B --> B3[Payback: 6-12m]
  C --> C1[Weeks to Close: 2-4]
  C --> C2[Win Rate: 70-90%]
  C --> C3[Run Rate: NRR %]
  B1 --> X[Variance: High]
  C1 --> Y[Variance: Low]
```

TAGS: expansion-arr, net-new-arr, forecasting, nrr, pipeline-management

Driver	Velocity	Variance	Predictability
Net New	2–6 weeks	±30%	20–30% win
Expansion	2–4 weeks	±10%	70–90% win

What's the difference between expansion ARR and net new ARR for forecasting?

What does the score mean?