MRR, ARR, CAC, and LTV: Unit Economics Every SaaS Buyer Must Understand
When you buy a SaaS business, you're not just buying current revenue — you're buying a prediction about future revenue. Whether that prediction is accurate depends entirely on the unit economics underneath the headline numbers. Sellers know how to present MRR attractively. Buyers need to know how to interrogate it.
This guide breaks down the four metrics that determine whether a SaaS acquisition is a compounding asset or a leaking bucket.
MRR: Monthly Recurring Revenue
MRR is the normalized monthly revenue from all active subscriptions. It's the most fundamental metric in any SaaS business — and also one of the most commonly misrepresented in acquisition listings.
What counts as MRR: active paid subscriptions billed monthly or normalized to monthly from annual plans (annual subscription ÷ 12).
What does NOT count as MRR: one-time setup fees, professional services revenue, lifetime deals, non-recurring payments, or revenue from churned customers that hasn't been removed yet.
In due diligence, always ask the seller to provide a cohort export from their payment processor (Stripe Dashboard → Revenue → MRR). If they can't produce this, or if their reported MRR doesn't match what you see in the Stripe export, that's a serious red flag.
MRR movement breakdown
A healthy MRR report shows five components:
- New MRR: Revenue from new customers this month
- Expansion MRR: Upgrades and add-ons from existing customers
- Contraction MRR: Downgrades from existing customers
- Churned MRR: Revenue lost from cancellations
- Net New MRR: New + Expansion − Contraction − Churned
A business with positive Net New MRR is growing. A business relying entirely on New MRR with high Churned MRR is running a leaky bucket — growth is masking retention failure.
ARR: Annual Recurring Revenue
ARR is simply MRR × 12. It's used to size the business and anchor valuation multiples. SaaS businesses typically sell at 3–6× ARR for smaller deals, and up to 10× ARR for high-growth, low-churn businesses.
One important nuance: if a business has a mix of monthly and annual plans, confirm how ARR is calculated. Annual subscriptions should be counted at face value (not annualized monthly), and monthly subscriptions should be multiplied. Sellers sometimes inflate ARR by counting active annual subscriptions at full annual value before confirming they'll actually renew.
Churn: The metric that determines everything
Churn is the percentage of revenue (or customers) lost each month from cancellations. It is the single most important unit economics metric in a SaaS acquisition — because it determines how long your revenue lasts.
Monthly revenue churn benchmarks:
| Churn Rate | Assessment | Revenue half-life |
|---|---|---|
| Under 1% | Excellent | ~6 years |
| 1–2% | Good | 3–6 years |
| 2–5% | Acceptable | 1–3 years |
| 5–10% | Concerning | Under 1 year |
| Over 10% | Red flag | Under 6 months |
At 5% monthly churn, you lose half your customer base in 13 months. At 10%, in just 7 months. A business with high churn isn't recurring revenue — it's a treadmill where you constantly need new customers just to stay flat.
Ask for a cohort retention table: how many customers who signed up in month X are still active 6, 12, and 24 months later. If the seller can't produce this, that's the answer.
CAC: Customer Acquisition Cost
CAC is the total cost to acquire one new paying customer. It includes all marketing spend, sales salaries, tool costs, and attribution.
Formula: CAC = Total acquisition spend ÷ New customers acquired in the same period
For most bootstrapped SaaS businesses, CAC comes from organic channels (SEO, word of mouth) and is very low — sometimes under $10. For businesses running paid ads, CAC can be $50–$500+. The number itself isn't the issue. The issue is the ratio of CAC to LTV.
During due diligence, ask: "How do new customers find you?" If the answer is "mostly Google" or "we don't really do marketing," that's a double-edged sword — low CAC is great, but you need to understand if that organic traffic is sustainable or at risk.
LTV: Customer Lifetime Value
LTV is the total revenue you expect from a customer before they churn. It's the counterweight to CAC and the foundation of whether the business's growth economics are sustainable.
Formula: LTV = Average MRR per customer ÷ Monthly churn rate
Example: If customers pay $49/month on average and monthly churn is 2%, LTV = $49 ÷ 0.02 = $2,450 per customer.
The LTV:CAC ratio
This single ratio tells you whether the business's acquisition economics are healthy:
- Under 1:1: Business is losing money acquiring customers
- 1:1 to 3:1: Marginal — not enough buffer for overhead and growth
- 3:1 to 5:1: Healthy — the benchmark for well-run SaaS
- 5:1+: Strong unit economics — pricing power or very low CAC
A bootstrapped SaaS with no paid acquisition and 2% churn might have LTV:CAC of 20:1 or higher. That kind of business is worth paying a premium for — it prints money with minimal ongoing investment.
NRR: Net Revenue Retention
NRR (sometimes called Net Dollar Retention) measures what percentage of last year's revenue you still have this year — from the same customer cohort — after accounting for churn, downgrades, and upgrades.
NRR above 100% means existing customers are paying you more over time even as some churn — this is the hallmark of a compounding SaaS business. NRR below 80% means the business is shrinking from within, and new customer acquisition is just patching the leak.
What to verify in due diligence
- Request a Stripe/Paddle MRR export — reconcile against seller-reported MRR
- Ask for a cohort retention table (month of acquisition vs. % still active at 6/12/24 months)
- Calculate actual monthly churn from the data yourself — don't rely on the seller's calculation
- Check how MRR has trended over the last 24 months — growth, flat, or declining?
- Identify the top 10 customers by MRR — what % of total revenue do they represent?
- Ask what the leading cause of churn is — do they track it?
- Confirm whether any lifetime deals are included in MRR (they shouldn't be)