Most SaaS forecasts fail by Q2, and it's almost always because churn and expansion aren’t modeled with the same rigor as new bookings. Leadership teams project new revenues with confidence while applying rough retention assumptions that miss critical dynamics. The result? Hiring plans that outpace growth, cash runway surprises, and investor conversations that expose financial blind spots.
The recurring revenue model that many SaaS businesses run on demands a different approach to forecasting than more traditional businesses. Small shifts in churn or expansion rates compound over time, creating outcomes that diverge dramatically from straight-line projections. Scenario-based forecasting that explicitly accounts for these forces gives leadership the visibility needed to make confident decisions.
For many companies, revenue churn is the single most important variable in a SaaS business forecast. A company with 5% monthly gross revenue churn will see a customer cohort decline by 46% over twelve months without expansion.
Consider a $100,000 cohort acquired in January with 5% monthly churn and no expansion:
Now, add 2% monthly expansion from remaining customers:
The cohort still contracts, but expansion slows the decline significantly. This is why Net Revenue Retention (NRR) matters more than any other metric. NRR shows whether your base grows or shrinks: starting MRR minus churn plus expansion, divided by starting MRR.
2025 data from Benchmarkit shows a median NRR of 101% among SaaS companies, with companies in the 75th percentile reporting an NRR of 110%. Broadly speaking, companies reported higher retention rates the larger their average contract value; retention was highest among enterprise contracts and lowest among companies with smaller annual contract values.
If your NRR sits below these ranges, growth depends almost entirely on new customer acquisition. That’s an expensive, unpredictable strategy that indicates your business doesn’t have a strong product-market fit and puts a lot of stress on your sales team to deliver new revenues.
[CTA - SaaS Performance Benchmarking Tool]
If you apply a single expansion rate across your customer base, you're not forecasting: you're guessing.
Enterprise customers behave nothing like SMB customers. Early cohorts may show different patterns than recent ones. Customers acquired through inbound, outbound, or partner channels exhibit distinct retention and expansion characteristics.
Build your model around meaningful segments: customer size tiers, acquisition cohorts, and channel source at minimum. For acquisition cohorts, track customers by the period they joined. A cohort closed during a discount campaign will behave differently than one acquired through enterprise sales. Cohort analysis reveals these patterns and prevents you from projecting behavior that no longer applies.
Your baseline should reflect current performance using trailing three to six-month averages for churn and expansion. This smooths seasonal variation. Structure your model as a monthly waterfall: Starting MRR + New Bookings + Expansion - Churn = Ending MRR.
Use driver-based inputs, not hard-coded revenue. Tie projections to customer count, ARPU, retention percentages, and expansion rates by segment. When you adjust assumptions, the model recalculates automatically. Create multiple default scenarios so you can build an understanding of how your numbers hold up under different business conditions.
Start with your base case, and then create an optimistic scenario that explores what happens if initiatives succeed. Your customer success team's new onboarding might reduce early-stage churn by 20%. A recently launched feature could drive expansion among 30% of your mid-market segment. Model these improvements realistically: customer behavior shifts gradually over quarters, not weeks.
Because of compounding, even a 1% improvement in monthly churn extends cohort value significantly:
You should also create a pessimistic scenario that allows for external pressures or underperformance due to internal issues. Increased competition might elevate churn. Economic uncertainty could drive downgrades rather than expansions. This scenario determines minimum cash runway requirements and triggers predetermined responses: reduce burn, shift from acquisition to retention, or adjust pricing.
Companies with usage-based or hybrid pricing need three separate forecasts:
Usage creates volatility and billing lag. A customer might increase consumption in March, but you invoice in April and collect in May. Model these timing differences explicitly, or cash flow projections will miss by months.
High-usage customers often expand faster but show different churn patterns: they're operationally locked in but more price-sensitive. Low-usage customers churn more easily but might respond well to activation campaigns. Segment your usage cohorts and track consumption trends separately from customer retention, or you'll miss the behavioral drivers behind your numbers.
Sophisticated investors expect multiple scenarios. They're looking for specific signals:
NRR trends over time. Is it improving or deteriorating? A company growing from 105% to 110% NRR tells a different story than one declining from 115% to 108%.
Other important facts to your investors include:
Understanding CAC payback periods alongside churn rates reveals whether you're acquiring customers profitably. A 12-month payback period with 5% monthly churn means you're barely breaking even on nearly half of your customers before they leave.
Mistake #1: Not updating assumptions when GTM strategy changes. Shifting from sales-led to product-led fundamentally changes conversion rates, customer profiles, and retention patterns. Using legacy data after significant changes produces misleading projections.
Mistake #2: Ignoring involuntary churn. Failed payments account for a considerable amount of revenue churn at many SaaS companies. This can be addressed with payment retry logic and card update workflows, yet most models don't separate involuntary from voluntary churn.
Mistake #3: Modeling expansion as a percentage of base. Real expansion comes from specific mechanisms: seat growth, tier upgrades, feature add-ons, usage increases. Model these drivers explicitly rather than applying blanket rates.
Mistake #4: Over-optimizing for precision. Models with dozens of variables become unmaintainable. Focus on the five to seven drivers that create 80% of the financial impact.
Build triggers around your scenarios. If monthly churn exceeds baseline by more than 1 percentage point for two consecutive months, investigate immediately. If expansion rates fall, examine which segments are slowing.
The best leading indicators vary by business but commonly include:
Track these by segment and set thresholds that trigger account reviews. Catching churn risk 45 days early often means the difference between saving and losing the account.
The real value isn't prediction: it's strategic clarity. Your model reveals which initiatives deliver the highest ROI. A pricing change that reduces churn by 1% may create more value than a new feature that drives 10% expansion in a small segment.
Small improvements compound dramatically. A company that reduces monthly churn from 3% to 2.5% and increases expansion from 2% to 3% might see minimal Q1 impact, but the compounding effect over twelve to eighteen months substantially changes trajectory.
Most SaaS teams don't need a new forecasting model. They need a model they can trust and the insight to act on it. G-Squared Partners specializes in building financial forecasting frameworks that drive better decision-making for SaaS companies.
Whether you're preparing for a fundraising round, managing cash runway uncertainty, or need to understand what's actually driving retention and expansion dynamics, our team brings the specialized expertise that growth-stage companies require. Contact us to learn how we can help you build forecasting models that give you confidence in your financial future.