| Metric | Definition or formula | Reference value | Planning use |
|---|---|---|---|
| MRR | Normalized (amortized) monthly subscription revenue; one-time revenue such as professional services or pilots is excluded | Trend and components matter more than a fixed target | Review monthly and split into new, expansion, contraction and churned revenue |
| ARR | MRR x 12 | Always 12 times MRR | Use ARR for annual contracts and MRR for monthly plans |
| Customer churn | Customers who churned in a period / customers at period start x 100 | Logo retention: enterprise 90-95%, mid-market 85%, small business 70-80% | Compare against your own customer size, not a blended average |
| Net revenue retention (NRR) | Revenue kept and grown from existing customers | B2B SaaS median 82%, top quartile 97% (Dec 2025, 3,500 software companies); aim for over 100% | Below 100%, new sales must first replace leakage before growth shows |
| CAC and payback | CAC = sales and marketing expense in the preceding period / new customers in the current period; payback months = S&M spend / (MRR x gross margin) | Judge against LTV rather than in isolation | Decide how fast acquisition spend can scale |
| LTV | Total gross profit expected from the average customer | At least 3x original CAC over time | Raise prices, improve retention or cut CAC when the ratio falls short |
| Gross margin | Share of revenue left after delivering the product | At least 75% long term | Sets how much is left for sales, product and hiring |
Figures come from ChartMogul’s SaaS metrics cheat sheet and the Substack essay The SaaS Metrics That Matter.
The top SaaS metrics are MRR, ARR, churn, net revenue retention, CAC, LTV and gross margin. Track them monthly, compare each against your own customer segment, and treat reference values as direction rather than fixed targets.
Founders weighing how to grow a startup will find these numbers decide where effort goes: retention, pricing, acquisition spend or cost of delivery.
What the numbers change in your decisions
Only a handful of SaaS metrics really drive decisions, according to the Substack essay on SaaS metrics. Many of them are also absent from a traditional P&L, as Maxio’s SaaS metrics guide points out, so they need a tracking system rather than a spreadsheet afterthought.
The practical effect of each reading looks like this:
- Dollar retention under 100% a year: the essay calls this a leaky bucket and a problem. Fix onboarding, product value or pricing before adding acquisition spend, because new customers will mostly replace those lost.
- Dollar retention near 120% or higher: the best SaaS companies reach this level, per the same essay. It suggests expansion revenue outweighs losses, which supports spending more on sales.
- LTV below 3x CAC: the company cannot easily justify growth spend. Choose between raising prices, improving retention and lowering acquisition cost, whichever your cohort data shows is cheapest to move.
- Gross margin below 70%: Maxio flags this as a sign that infrastructure costs are eating into profit. Review hosting and third-party costs before pushing harder on sales.
- ARR before a funding round: in a 2021 post, the essay said the Series A threshold had previously been $1M ARR and had recently been around $500k as rounds got preempted and happened earlier. Treat that as a dated market observation, not a rule.
With those decisions in view, the mechanics of each metric are worth getting exactly right, starting with revenue.
Revenue metrics: MRR and ARR
MRR is the base number every other metric hangs on. ChartMogul defines monthly recurring revenue as normalized, amortized monthly subscription revenue. A customer on a longer plan interval is spread across the months it covers. Maxio gives a simple case: 200 active customers each paying $50 per month produce $10,000 MRR.
Keep non-recurring revenue out. The Substack essay is explicit that one-time revenue, such as professional services or pilots, does not count towards MRR or ARR. Mixing it in inflates the trend line and hides how much of the business is truly repeatable.
ARR is MRR multiplied by 12, and the essay stresses it is always 12 times MRR. Because the conversion is mechanical, the choice between them is about contract type: use ARR for annual-contract businesses and MRR for monthly ones. Quoting ARR for a business on monthly plans makes the number look steadier than the contracts behind it.
Break MRR into new, expansion, contraction and churned movements each month. That split shows whether growth comes from winning customers or growing existing ones, which is the first clue to whether the product is pulling customers in. It also supports planning for Product‑Market Fit Stages: A Practical 3‑Stage Roadmap To Find Fit Fast (2026) – Startalliance, where retention and expansion patterns help show whether fit is holding.
Churn and retention
Churn tells you how much of the base you keep, and it can be measured by customers or by revenue. Read both, because a few large accounts leaving can hurt more than the customer count suggests.
Customer churn and logo retention
ChartMogul gives the customer churn formula: divide the customers who churned in a period by the customers at the start of that period, then multiply by 100. The essay’s logo retention ranges show why segment matters: 90-95% is common for enterprises, 85% for mid-market and 70-80% for small businesses. A small-business product whose logo retention sits inside that range is on track, while the same level in enterprise would be a concern.
Revenue churn and net revenue retention
Gross revenue churn adds churned MRR and contraction MRR, then divides by starting MRR, per ChartMogul. A worked example from Growth Equity Interview Guide starts at $10,000, adds $3,000 of new revenue and ends at $12,000, which means $1,000 of revenue churn and a 10% churn rate.
For net revenue retention, ChartMogul’s data on 3,500 software companies as of December 2025 puts the B2B SaaS median at 82% and the top quartile at 97%. It still advises B2B SaaS companies to aim for over 100%. Consumer and AI-native products run much lower, with median NRR around 48-49%. That makes the benchmark a matter of business model: a B2C product at the median is not failing by B2B standards, but it should not borrow B2B targets.
Read these numbers as separate ideas. The median shows what is typical, the top quartile shows what strong performers reach, and over 100% is the point at which existing customers alone grow revenue.
Unit economics: CAC, payback, LTV and gross margin
These four metrics decide whether growth spending pays back. They interlock: gross margin feeds both payback and LTV, and churn feeds LTV through customer lifespan.
CAC and payback
The essay calculates CAC by dividing sales and marketing expense in the preceding period (a month or a quarter) by the number of new customers in the current period. Using the prior period reflects that spend precedes the customers it wins. Payback months come from sales and marketing spend divided by MRR x gross margin. Maxio’s example: a CAC of $1,200 and $200 of MRR per customer gives a payback period of six months.
LTV
Breaking Into Wall Street defines LTV as the total gross profit a SaaS company can expect from the average customer. Its worked case multiplies 5 years by $100 by an 80% gross margin to get $400. Customer lifespan comes from churn: lifespan equals 1 divided by the average cancellation rate, so a 10% rate implies 10 years.
Use the ratio with care. The same source treats 3x as better than 2x, but calls a 10x LTV to CAC unrealistically high in most cases, so an extreme result more likely points to a flawed input, such as an unrealistically low churn assumption, than a triumph. The essay’s guide is that LTV should grow to at least 3x original CAC over time, which is a floor to aim for rather than a target to exceed at any cost.
Gross margin
Maxio’s formula is total revenue minus cost of goods sold, divided by total revenue, multiplied by 100. With $200,000 of revenue and $50,000 of costs, the margin is 75%. The essay says SaaS companies should reach at least 75% long term, and Maxio treats anything under 70% as a signal of infrastructure cost pressure. A young company may sit below the long-term level, but a falling trend is a prompt to look at hosting and vendor costs.
Putting the metrics into a monthly routine
Reviewing the full set on one page each month beats chasing individual numbers, because each one is easiest to interpret next to the others.
- Segment before judging: split churn and retention by customer size and signup cohort, since the benchmarks above differ widely between enterprise, mid-market and small business.
- Pair efficiency with quality: read CAC and payback alongside NRR and gross margin. A short payback on customers who churn quickly is not a healthy result.
- Tie each metric to an owner and an action: retention to product and customer success, CAC to go-to-market, gross margin to engineering and finance.
- Link spend to channel evidence: acquisition cost depends on how you reach customers, which the Launch Faster, Win Customers: A Practical Go-To-Market Strategy For Tech Startups (2026 Playbook) – Startalliance guide covers in more depth.
Frequently asked questions
When should a founder use ARR instead of MRR?
Use ARR for annual-contract businesses and MRR for monthly-plan businesses. Since ARR is MRR x 12, the choice only changes how the same revenue is presented.
Does a high LTV to CAC ratio always mean healthy growth?
No. A ratio of 3x is regarded as better than 2x, but 10x is unrealistically high in most cases, so check the churn and margin assumptions behind it.
Why do benchmarks for NRR vary so much?
Business model changes the baseline. ChartMogul’s median is 82% for B2B SaaS, against around 48-49% for consumer and AI-native products.
What does churn tell you that MRR growth does not?
MRR growth can hide leakage when new sales are strong. Churn and contraction show how much existing revenue is being lost, and dollar retention under 100% a year signals a leaky bucket.
Do one-time fees count towards recurring revenue?
No. Revenue such as professional services or pilots is excluded from both MRR and ARR.
