SaaS unit economics: the numbers to calculate before you build

Before building a SaaS product, calculate four numbers: revenue per customer per month (ARPU), how many customers leave (churn), what it costs to acquire a customer (CAC) and how much a customer brings over their lifetime (LTV). If LTV is not several times CAC, the product loses money on every new customer, however well it is built.

Here is how to calculate unit economics at the idea stage, which assumptions to test and how the numbers shape what you build.

What are SaaS unit economics?

SaaS unit economics calculate how much a company earns or loses on one customer over their whole lifetime. For a subscription business this is the key viability measure: revenue arrives a little each month, while acquisition costs are paid upfront.

The core metrics:

Metric	What it means	How to calculate
MRR	Monthly recurring revenue	Sum of all subscription payments in the month
ARPU	Average revenue per customer per month	MRR / paying customers
Churn	Share of customers cancelling in a month	Left during the month / at the start of the month
Gross margin	Share of revenue after direct costs	(Revenue − hosting, support, payment fees) / revenue
LTV	What a customer brings over their lifetime	ARPU × gross margin / monthly churn
CAC	Cost of acquiring one customer	Sales and marketing spend / new customers

How do you calculate LTV and CAC before launch?

Before launch, LTV and CAC are calculated from assumptions rather than data — and that is fine: the goal is not a precise figure but an understanding of the conditions under which the business works. The calculation shows which assumptions are critical, and those are the ones to test first.

Assumptions for a B2B request-tracking service
  plan price (ARPU)             $50 a month
  gross margin                  80%
  monthly churn                 4%
  CAC                           $300

LTV = 50 × 0.8 / 0.04 = $1,000
LTV / CAC = 1,000 / 300 ≈ 3.3
CAC payback = 300 / (50 × 0.8) ≈ 7.5 months

The industry commonly aims for an LTV to CAC ratio of at least 3 and CAC payback within a year. It is not a law but a widespread benchmark: below it, growth consumes cash faster than it comes back.

Which assumptions to test first

1. Churn — the most sensitive number: 8% churn instead of 4% halves LTV.
2. CAC — the cost of a customer from real channels, not "they'll just come".
3. Price — whether customers will actually pay what the model assumes.

Photo: Unsplash

How do unit economics shape what you build?

Unit economics show which features pay off: anything that reduces churn, raises the average price or lowers acquisition cost affects the business more than features added "for completeness". That is why the calculation should come before the MVP backlog.

In practice:

• high churn — prioritise onboarding, reminders and anything that gets customers to value in the first days;
• low average price — you need team plans, add-on modules and annual billing;
• expensive acquisition — referral mechanics, a free tier or trial, and integrations that bring customers in matter;
• low margin — revisit infrastructure, manual operations and expensive third-party services.

We covered the metrics used to track these effects in product metrics after launch.

Which calculation mistakes are common?

The most common mistake is calculating LTV without margin and churn, assuming "the average customer pays for three years". The second is leaving sales and marketing salaries out of CAC and counting only ad spend.

Other mistakes:

• taking churn from the first loyal customers, who came through personal connections;
• mixing annual and monthly subscriptions in one churn figure;
• leaving payment fees, taxes and support costs out of the margin;
• averaging across all customers although segments behave very differently;
• not updating the model once real data arrives.

How do you test the economics before major development?

Test the economics before major development with cheap experiments: a landing page with a price and a sign-up form, manual sales to the first customers, a pilot with limited features. The goal is real data on price, conversion and acquisition cost.

The order we recommend:

1. Build the model on assumptions and find the most sensitive ones.
2. Test price and demand: a landing page, calls with prospects, pre-orders.
3. Launch a minimal version and measure churn in the first cohorts.
4. Recalculate the model on real data and decide where to invest next.

We covered the scope and budget of a minimal version in how much an MVP costs.

In short

• SaaS unit economics are ARPU, churn, margin, LTV and CAC.
• LTV = ARPU × margin / churn; a common benchmark is LTV/CAC of at least 3 and CAC payback within a year.
• Churn is the most sensitive assumption; test it and price first.
• The numbers set development priorities: onboarding, pricing plans, acquisition channels.
• Test the economics with cheap experiments before major development, and recalculate on real data.