A Practical Guide to SaaS Revenue Forecasting
Revenue forecasting is one of those things every SaaS founder knows they should be doing but few do well. It sits in that uncomfortable gap between “I have a spreadsheet somewhere” and “our FP&A team runs a full model every quarter.” If you are anywhere between those two extremes, this guide is for you.
We will walk through why forecasting matters, the most common methods, the inputs you need to get right, and the mistakes that quietly wreck your projections.
Why Forecasting Matters More Than You Think
Revenue forecasting is not just a fundraising exercise. Yes, investors want to see projections, but the real value is operational. A reliable forecast lets you answer three questions that come up constantly:
- Can we afford to hire? Headcount is the largest expense for most SaaS companies. Without a forecast, you are either hiring too aggressively (burning runway) or too conservatively (missing growth).
- When do we hit profitability? Or more precisely, when do we need to hit profitability given our current cash position? This question only has an answer if you can project revenue forward with some confidence.
- Where should we invest? Marketing budget, engineering resources, support hiring: these all depend on knowing what revenue looks like in three, six, and twelve months.
A forecast does not need to be perfect. It needs to be directionally right and updated regularly. The goal is to make better decisions, not to predict the future with precision.
Three Forecasting Methods Worth Knowing
1. Simple Run-Rate Projection
This is the most basic approach: take your current MRR, multiply by 12, and apply a growth rate. If you are at $50k MRR and growing 8% month-over-month, your run-rate projection gives you a rough annual figure.
When it works: Early-stage companies with limited historical data and relatively consistent growth. It is a reasonable starting point when you have fewer than 12 months of data.
When it breaks: As soon as growth rates start to fluctuate, seasonality kicks in, or churn becomes meaningful. A simple run-rate assumes the future looks like this month, which is rarely true for more than a quarter or two.
2. Cohort-Based Forecasting
Cohort-based forecasting breaks your customer base into groups based on when they signed up (or another meaningful attribute like plan type or acquisition channel). You then model each cohort’s behavior over time: retention, expansion, contraction, and churn.
This approach is more work to set up but dramatically more accurate because it accounts for the reality that not all customers behave the same way. A customer who signed up 18 months ago has a different retention profile than one who signed up last week.
The basic process:
- Group customers by signup month (or quarter)
- Track each cohort’s revenue over time
- Calculate average retention and expansion curves
- Apply those curves to current and projected new cohorts
- Sum it all up to get your forecast
Cohort-based models are particularly useful for understanding how improvements to retention or onboarding compound over time.
3. Bottoms-Up Forecasting
A bottoms-up forecast builds revenue from the components that drive it: leads, conversion rates, average deal size, and expansion patterns. Instead of starting with a top-line number and projecting forward, you start with the inputs and let the math produce the output.
Example inputs for a self-serve SaaS:
- Monthly website visitors
- Trial signup rate
- Trial-to-paid conversion rate
- Average starting MRR per customer
- Monthly expansion rate (upgrades, seat additions)
- Monthly churn rate
This method is powerful because it connects your forecast directly to things you can measure and influence. If you want to grow faster, the model shows you exactly which lever to pull.
Getting Your Inputs Right
Regardless of which method you use, the quality of your forecast depends on the quality of your inputs. Here are the numbers that matter most:
MRR Growth Rate. Track net new MRR (new customers + expansion - contraction - churn) as a percentage of starting MRR. Use at least 6 months of data to establish a trend, and be honest about whether the trend is accelerating, decelerating, or flat.
Gross Churn Rate. The percentage of revenue lost each month from cancellations and downgrades, before accounting for expansion. This is the number most founders underestimate. A 3% monthly churn rate sounds small, but it means you lose roughly 31% of your revenue base annually.
Net Revenue Retention. Revenue from existing customers this period divided by their revenue last period. If your NRR is above 100%, your existing customer base is growing without any new sales. This is the single most important metric for forecasting because it tells you how much of your current revenue you can count on keeping.
Expansion Revenue. Upgrades, additional seats, usage overages, and add-on purchases. For many SaaS companies, expansion revenue is the difference between good and great growth. Track it separately from new customer revenue.
Tools like Subdash can help here by pulling these metrics directly from your Stripe data, so you are working with actuals rather than estimates when building your model.
Common Forecasting Pitfalls
Ignoring Seasonality
Many SaaS businesses have seasonal patterns they do not recognize. B2B companies often see slower signups in December and August. E-commerce tools spike in Q4. If you are projecting based on your best month, you will overshoot.
Look at 12+ months of data and check for patterns. If you do not have 12 months yet, talk to others in your space about what they see.
Conflating Gross and Net Churn
Gross churn tells you how much revenue you are losing. Net churn (or net revenue retention) accounts for expansion. Both matter, but they tell you different things. Using net churn in your forecast while ignoring the expansion assumptions baked into it is a common mistake.
Assuming Linear Growth
Growth rates change. Early traction often looks exponential, then tapers as you saturate initial channels or move upmarket. Build your forecast with the assumption that growth rates will evolve, and revisit your assumptions monthly.
Not Accounting for Contraction
Customers do not just churn or stay. Many downgrade. If 5% of your customers drop to a lower plan each month, that revenue loss is real and needs to be in your model.
Over-Indexing on New Business
It is tempting to build a forecast that is mostly about “how many new customers will we sign.” But for any SaaS company past its first year, the behavior of existing customers (retention, expansion, contraction) typically has a larger impact on total revenue than new sales.
Why Scenario Modeling Matters
A single-point forecast is a bet. Scenario modeling turns that bet into a range.
At minimum, build three scenarios:
- Conservative: Assumes growth slows, churn increases slightly, and expansion stays flat. This is your “things get harder” case.
- Base: Uses your current trends with modest improvements. This is what you plan against.
- Optimistic: Assumes your recent improvements to conversion, retention, or expansion continue compounding. This is the case you hope for but do not staff against.
The value of scenarios is not in picking the right one. It is in understanding the range of outcomes and making decisions that work across multiple scenarios. If a hire only makes sense in the optimistic case, it is probably too early. If an investment pays off even in the conservative case, it is a safer bet.
Subdash’s forecasting features let you run these scenarios against your actual Stripe data, so you can see how different assumptions about churn, growth, and expansion play out against real numbers rather than a hypothetical spreadsheet.
Putting It Into Practice
If you are starting from scratch, here is a practical sequence:
- Export your MRR data for the last 6-12 months, broken down by new, expansion, contraction, and churn
- Calculate your trailing averages for each component
- Build a simple spreadsheet that projects forward 12 months using those averages
- Create three scenarios by adjusting your key assumptions up and down by 20-30%
- Review and update monthly as actuals come in
The first version will feel rough. That is fine. The act of building the forecast forces you to understand your business at a level that intuition alone cannot match. And each month, as you compare projections to actuals, the model gets better.
Revenue forecasting is not about being right. It is about being less wrong over time, and making sure the decisions you make today still look reasonable when tomorrow’s numbers come in.