MRR Forecasting Methods for SaaS Teams

Why Forecast MRR?

MRR forecasting turns today's metrics into tomorrow's decisions. A reliable forecast tells you when you can hire, how much you can spend on acquisition, and whether you will hit the targets you promised investors. Without a forecast, every spending decision is a guess; with one, it is a plan with explicit assumptions you can revisit. Start experimenting with scenarios in our MRR Growth Projector.

Method 1: Run-Rate Extrapolation

The simplest forecast: Future MRR ≈ Current MRR × (1 + recent growth rate)^periods. If MRR grew 8% monthly for the last three months, a 12-month run-rate forecast projects MRR × 1.08^12. This method is fast and works reasonably for stable, mature businesses — but it is dangerous for startups because growth rates decay as the base grows. A startup growing 20% monthly will almost certainly not sustain that for a year, and run-rate math will produce fantasy numbers. Use run-rate only for short horizons (1–3 months) or as a sanity check, never as your official plan.

Method 2: MRR Movement Build-Up

The workhorse forecast for most SaaS companies builds the future from the five MRR movements:

• New MRR: forecast from pipeline — deals × close rate × average contract value, normalized monthly.
• Expansion MRR: forecast from historical expansion rates applied to the growing base, adjusted for known large upgrades.
• Contraction and churned MRR: forecast from historical churn and contraction rates, ideally segmented by cohort age and plan.

Forecasted MRR = Starting MRR + New + Expansion − Contraction − Churn, rolled forward month by month. This method forces you to make every assumption explicit, which is its real value: when the forecast misses, you know exactly which assumption broke. For the churn inputs, use the Churn Rate Calculator to get clean historical rates.

Method 3: Cohort-Based Forecasting

Cohort forecasting is the most accurate method for businesses with enough history. Group customers by signup month, measure each cohort's retention curve (what % of MRR remains after 1, 2, 3… months), and apply the stabilized curve to newer cohorts. Older cohorts' curves predict younger cohorts' futures.

This method naturally captures the reality that churn is highest early and stabilizes over time — something flat-rate forecasts miss entirely. It also isolates the impact of product or onboarding changes: if the March cohort retains better than the January cohort, something you changed in February worked. Our Cohort Analysis Guide walks through building these curves step by step.

Method 4: Pipeline-Weighted Sales Forecast

For sales-led SaaS, new MRR comes from the pipeline: Expected new MRR = Σ (deal value × stage close probability). A $10k MRR deal at 50% probability contributes $5k of expected new MRR. Weight each pipeline stage by its historical close rate, apply sales cycle timing (deals close when they close, not when you want them to), and sanity-check against historical pipeline-to-close conversion. The classic failure mode is happy-ears forecasting — reps inflate close probabilities. Calibrate stage probabilities against actual history quarterly.

Choosing the Right Method for Your Stage

• Pre-product-market fit: do not forecast far — use 1–3 month run-rates and focus on learning, not precision.
• Early growth ($10k–$100k MRR): MRR movement build-up with conservative churn assumptions; cohorts are still thin.
• Scaling ($100k–$1M MRR): cohort-based forecasting for the base plus pipeline-weighted new sales — this is the gold standard combination.
• Mature ($1M+ MRR): segmented cohort models by plan, region, and channel, reconciled monthly against actuals.

Whatever method you choose, forecast in scenarios — base, upside, and downside — and review monthly. A forecast you never revisit is fiction; a forecast you update is a management tool. Tie your forecasts back to unit economics with SaaS Pricing analysis.

Conclusion

Start with movement build-up, graduate to cohort-based forecasting as your data matures, and layer pipeline weighting on the new-sales side. Keep assumptions explicit, review monthly, and always present scenarios instead of single numbers. Good forecasting does not predict the future perfectly — it makes your decisions about the future explicit and testable.