Cohort-Based SaaS Revenue Forecasting: Break Revenue Into the Four Fundamental Engines

Moving Beyond Simple Growth Rates in SaaS Revenue Forecasting

A simple month-over-month growth rate can feel like a solid indicator of progress, but it often masks the underlying dynamics of a SaaS business. That steady 10% growth might hide rising customer churn, which is being offset by a heroic, and perhaps unsustainable, sales effort. This lack of clarity is why so many revenue projections swing wildly and misinform critical hiring or fundraising decisions. For a more durable and insightful approach to how to forecast SaaS revenue by customer cohort, founders should turn to cohort-based analysis.

This method becomes particularly valuable around the $500k to $1M ARR mark, or once a company has 50 to 100+ customers, providing a stable base for analysis. It separates the performance of new customers from existing ones, offering a true picture of business health. For integrating CRM data into your projections, see the Sales & Pipeline Forecasting Frameworks for a broader view.

Why Aggregate Growth Metrics Are Misleading

Projecting future revenue by simply extrapolating a historical monthly growth rate is a common but flawed starting point. This single metric blends the performance of newly acquired customers with the behavior of your entire existing base, offering a distorted view. A blended rate cannot answer the most important questions for a SaaS business: Are we keeping the customers we win? Are they spending more with us over time? Answering these requires detailed customer retention analysis, not just a top-line growth percentage.

This is the critical distinction between the health of your new customer acquisition and the health of your existing customer base. Cohort-based forecasting separates these two stories. By grouping customers into cohorts based on when they signed up, such as the “January 2024 cohort,” you can track the lifecycle of each group independently. This customer segmentation forecasting reveals trends in retention, expansion, and churn that are invisible in an aggregate growth number, forming the basis for reliable SaaS revenue modeling.

Part 1: How to Forecast SaaS Revenue by Customer Cohort Using the Four Fundamental Engines

To build a reliable forecast, you must first understand that your Monthly Recurring Revenue (MRR) is driven by the four fundamental engines of SaaS growth. Breaking down your revenue changes into these components provides a transparent view of your business momentum and operational performance.

1. New MRR: This is the simplest engine. It represents the total recurring revenue generated from brand new customers acquired during a specific period, typically a month. This engine is a direct measure of your sales and marketing effectiveness.
2. Expansion MRR: This is revenue growth from your existing customers. It includes upgrades to higher-priced plans, the purchase of additional seats, or cross-sells into new product lines. Strong expansion MRR is a powerful sign of a healthy, valuable product that customers are integrating deeply into their operations.
3. Contraction MRR: The opposite of expansion, this is the decrease in recurring revenue from existing customers who downgrade their plans, reduce their seat count, or remove add-ons. It often signals a partial misalignment between your product's value and a customer's needs.
4. Churned MRR: This represents the total MRR lost from existing customers who cancel their subscriptions entirely during the month. This is the most direct measure of value destruction and a critical metric to manage.

Together, these engines determine your Net Revenue Retention (NRR), a key metric for measuring the health and momentum of your existing customer base. It is calculated by taking your starting MRR, adding expansion MRR, then subtracting contraction and churned MRR. A strong NRR is often benchmarked at over 100% for successful SaaS companies, indicating that revenue growth from existing customers outpaces any losses.

The importance of expansion revenue tracking cannot be overstated. According to a 2021 report by OpenView, companies with best-in-class NRR generated over half of their new ACV from expansion. This demonstrates that your happiest customers are your most significant and efficient source of growth. For formal guidance on recognizing this revenue, consult the principles within IFRS 15.

Part 2: A Practical Guide to Your First Cohort Analysis for Startups

The primary challenge for founders is that customer data is often fragmented. It lives across billing tools like Stripe or Chargebee, a CRM like HubSpot, and various product analytics platforms. However, you can build your first powerful cohort analysis with just the essentials from your billing system. For most early-stage startups, the reality is pragmatic: a well-structured spreadsheet is the perfect place to start.

Here’s how to assemble your first analysis:

1. Export Your Subscription Data: Log into your billing system, such as Stripe, and export a list of all subscriptions. You need three key fields for each subscription: a unique Customer ID, the Subscription Start Date, and the MRR value for each month that customer has been active.
2. Assign Each Customer to a Cohort: In your spreadsheet, create a new column called "Cohort." For each customer, use their Subscription Start Date to assign them to a monthly cohort. For example, a customer who signed up on January 15, 2024, belongs to the "Jan-2024" cohort.
3. Structure Your Cohort Table: Create a table where each row represents a distinct cohort (e.g., Jan-24, Feb-24). Each column should represent the number of months since that cohort was acquired (Month 0, Month 1, Month 2, etc.). The cells will contain the sum of MRR from all customers in that cohort for that specific month in their lifecycle. Month 0 represents the New MRR for that cohort.

If you sell digital services to consumers in the United Kingdom, remember to check the UK VAT rules for compliance when processing billing data.

Your resulting table should look similar to this structure, which can be formatted in a plain text or code block to avoid spreadsheet errors:


| Cohort | M0 (New) | M1 | M2 | M3 | M4 |
| :-------- | :--------- | :--------- | :--------- | :--------- | :--------- |
| Jan-2024 | $10,000 | $9,800 | $9,700 | $9,900 | $9,850 |
| Feb-2024 | $12,500 | $12,300 | $12,100 | $12,200 | |
| Mar-2024 | $11,000 | $10,900 | $10,950 | | |
| Apr-2024 | $14,000 | $13,800 | | | |
| May-2024 | $13,200 | | | | |


This structure immediately visualizes customer churn and revenue retention patterns, providing the critical historical data needed for forecasting SaaS growth with confidence.

Part 3: From Historical Analysis to Prediction

With your historical cohort table built, you can now move from analysis to building a predictive financial model. The process involves two distinct methodologies: one for your existing customers and another for the future customers you have not yet acquired.

Forecasting Existing Customer Cohorts

Your existing cohorts are a known quantity. You can project their future revenue by calculating their historical retention and expansion rates. From the table above, you can see how each cohort's MRR changes from one month to the next. By averaging these month-over-month percentage changes across several mature cohorts, you can develop a reliable retention curve. For example, you might find that on average, cohorts retain 98% of their revenue from Month 2 to Month 3, but expand to 101% from Month 3 to Month 4 due to upgrades.

Here is a simplified numerical example of these ARR prediction methods in action. The February 2024 cohort had $12,100 in Month 2. Your historical data shows that cohorts, on average, expand to 100.8% of their MRR between Month 2 and Month 3. You can therefore forecast the February cohort's Month 3 MRR as $12,100 multiplied by 1.008, which equals approximately $12,200.

Forecasting Future Customer Cohorts

Projecting revenue from future cohorts who have not yet signed up requires a different approach that relies on your sales and marketing assumptions. You must estimate the New MRR you will acquire in upcoming months. This figure is typically derived from your sales pipeline in a CRM, factoring in historical data on lead generation, conversion rates, and average contract values. Once you have a monthly New MRR forecast (e.g., you expect to close $15,000 in New MRR in June 2024), that new cohort's future performance can be projected using the same historical retention curve you calculated for your existing cohorts.

This dual approach provides a robust foundation for your financial model, separating the predictable revenue from your existing base from the more variable growth dependent on new customer acquisition.

Part 4: Graduating From Spreadsheets to Scenario Planning

Maintaining a multi-cohort forecast in a manual spreadsheet is effective initially, but it quickly becomes time-consuming and prone to errors as your business scales. The static nature of a spreadsheet makes it difficult to ask dynamic “what-if” questions, which are essential for strategic planning. When you spend more time updating formulas than analyzing outcomes, this is the trigger to graduate to more sophisticated tooling.

Scenario planning is the ability to model the impact of different business decisions on your forecast. For instance: What happens to our runway if we increase marketing spend by 20% and it boosts New MRR by only 10%? How does a 2% improvement in churn affect our cash position in six months? Answering these questions confidently is difficult in a spreadsheet but is the primary function of dedicated financial planning tools.

Startups typically progress through a few stages of tooling. After spreadsheets, many adopt analytics platforms like Baremetrics or ChartMogul, which automate cohort analysis directly from billing data. As forecasting needs become more complex, especially around Series A and B fundraising, companies often graduate to dedicated FP&A platforms like Vareto, OnPlan, or Pigment. These tools are built specifically for dynamic, multi-scenario SaaS revenue modeling, linking financial projections directly to operational drivers. These insights can be integrated with your real-time sales dashboards for continuous monitoring and adjustment.

Practical Takeaways for Founders

For founders running finance in the early stages, moving from a simple growth rate to a cohort-based forecast can feel daunting. However, the clarity it provides for making decisions about hiring, managing runway, and truly understanding customer retention is invaluable.

First, start today with what you have. You do not need a perfect, automated system from day one. Export the necessary subscription data from Stripe or QuickBooks and build your initial cohort table in a spreadsheet. This simple act will provide more insight than any aggregate growth metric ever could.

Second, focus intensely on Net Revenue Retention. This single metric tells you if you have a product customers value and are willing to pay more for over time. An NRR consistently above 100% is a powerful signal of product-market fit and the most efficient engine for sustainable growth. You can use your cohort-based forecast to directly inform headcount planning and justify hiring decisions with data.

Finally, use your forecast as an active decision-making tool. A good forecast is not just a report for investors; it is a strategic guide. Model how improving retention by a single percentage point could extend your runway or fund a key hire. When your questions become more complex than your spreadsheet can handle, you will know it is time to invest in a dedicated tool. This structured approach, integrated with broader Sales & Pipeline Forecasting Frameworks, turns your financial model from a reactive report into a proactive guide for growth.

Frequently Asked Questions

Q: How much historical data do I need for a reliable cohort analysis?
A: Ideally, you should have at least 12 to 18 months of subscription data. This provides enough history to identify meaningful trends in retention and expansion, especially seasonality. However, you can start with as little as 6 months of data to begin building your initial SaaS revenue modeling framework.

Q: What is considered a good Net Revenue Retention (NRR) rate?
A: For venture-backed SaaS companies, an NRR over 100% is considered good, as it means expansion revenue from existing customers outweighs all churn and contraction. Best-in-class companies, particularly those selling to mid-market or enterprise customers, often target NRR of 120% or higher. For startups serving SMBs, achieving over 100% can be more challenging.

Q: Why is cohort analysis better than just tracking customer churn rate?
A: A simple churn rate averages the behavior of all customers, old and new. Cohort analysis is a superior form of customer retention analysis because it reveals how churn behavior changes over the customer lifecycle. You might discover, for instance, that churn is high in Month 2 but very low for customers who stay past Month 6.