SaaS Metrics for Startups in 2027: The Few Numbers to Track, With Worked Examples

SaaS metrics for startups come down to a handful of numbers you can compute by hand. This guide explains MRR and churn, net revenue retention, CAC payback and the LTV to CAC ratio with worked examples, and flags the formula mistakes that are easy to make, so you do not build a dashboard on a wrong definition.

What to know before you track SaaS metrics

The formulas behind SaaS metrics are simple, but they are easy to mix up: net new MRR with MRR, a cost per customer divided by an annual figure and called months, or NRR counted with new customers included. Rules such as "3 to 1 is good" or "payback under 12 months" circulate widely as fact, and early-stage reality, such as what to do with 20 customers, free trials and annual plans, is often skipped.

So the examples here are small, labelled hypothetical and arithmetic-checked, and each formula says which assumptions it needs.

Why fewer numbers is better

At the start you have few customers, little history and little time. A dashboard of 40 metrics produces noise. Pages in this space list from 4 to 40 or more. A better question is: what decision will this number change?

For most early SaaS products there are three decisions:

1. Is anyone paying, and is that growing? (revenue and growth)
2. Are they staying? (churn and retention)
3. Can I afford to find more of them? (acquisition cost against what they pay)

Every metric below answers one of those. If you cannot name the decision, skip the metric for now.

A worked month

All numbers here are hypothetical, chosen so the arithmetic is easy to follow. Suppose a small SaaS has a single $30-a-month plan.

Ending MRR = 3,000 + 300 + 60 − 30 − 120 = $3,210. Net new MRR = 3,210 − 3,000 = $210.

Those are two different numbers. MRR is the total at a point in time. Net new MRR is the change over a period. Mixing them up is a common error.

Revenue metrics

MRR and ARR

MRR (monthly recurring revenue) is the monthly value of all active subscriptions. Annual plans count as one-twelfth each month. ARR is MRR × 12. In the example, ARR = 3,210 × 12 = $38,520.

Count only recurring revenue. One-off fees, setup charges and usage you cannot predict should be listed separately. If you offer a free trial, do not count trial users until they pay.

Growth rate

Monthly MRR growth = net new MRR ÷ starting MRR. Here, 210 ÷ 3,000 = 7%.

It helps to see what compounding means. If 7% held for twelve months, MRR would multiply by 1.07 twelve times, which is about 2.25×. That is an illustration of arithmetic, not a forecast. Early growth is lumpy, and one large customer can swing it. Look at a few months together, not one.

Retention metrics

Customer churn

Customer (logo) churn = customers lost ÷ customers at the start of the period. Here, 4 ÷ 100 = 4%.

Use the starting count as the base, and the same period for every month. Saying "total customers" without stating whether that means start or end makes numbers hard to compare.

The annual effect is bigger than it looks. If 4% leave each month, the share who remain after a year is 0.96 to the twelfth power, about 0.61. So roughly 39% of the starting customers are gone after a year. For comparison, 3% monthly is about 31% a year and 2% monthly is about 21.5% a year. Keep this in mind when someone says churn is "only a few percent".

Revenue churn

Gross revenue churn includes lost revenue from both cancellations and downgrades: (churned MRR + contraction) ÷ starting MRR = (120 + 30) ÷ 3,000 = 5%. Counting only cancellations gives 4%. Different pages use different versions, so write down which one you use.

Net revenue retention

NRR shows what happens to revenue from the customers you already had, including upgrades. The formula is:

NRR = (starting MRR + expansion − contraction − churned MRR) ÷ starting MRR

In the example: (3,000 + 60 − 30 − 120) ÷ 3,000 = 2,910 ÷ 3,000 = 97%.

Two points to keep straight:

• NRR counts only revenue from customers who existed at the start. New customers are not included. Several pages blur this.
• NRR above 100% means the existing base grew in revenue overall. It does not mean every group of customers expanded, and one page overstates it.

Gross revenue retention (GRR) is the same but leaves out expansion: (3,000 − 30 − 120) ÷ 3,000 = 2,850 ÷ 3,000 = 95%. It tells you how much you keep before upgrades help.

With 100 customers, one large account can change NRR a lot. Treat small-number swings carefully.

Acquisition metrics

CAC

CAC (customer acquisition cost) = sales and marketing spend ÷ new customers in the period. Suppose you spent $1,500 in the month on everything used to win customers, including paid tools, ads and the cost of your own time if you want a true figure. With 10 new customers: 1,500 ÷ 10 = $150.

Decide once whether to include your own time. Leaving it out flatters CAC, and it is the most common omission in early-stage numbers. Also note that sales cycles can lag: spend this month may win customers next month. When volume is low, use a three-month total.

CAC payback

Payback is how many months of gross profit it takes to earn back the cost of winning a customer:

Payback (months) = CAC ÷ (monthly revenue per customer × gross margin)

Assume, hypothetically, an 80% gross margin (revenue minus the direct cost of serving the customer, such as hosting, payment fees and support). Monthly gross profit per customer = $30 × 0.80 = $24. Payback = 150 ÷ 24 = 6.25 months.

The unit matters. Use monthly revenue per customer to get months. Dividing by annual contract value gives years, not months. If you have annual plans, divide annual value by 12 first.

Lifetime value and the LTV to CAC ratio

A simple estimate of lifetime value (LTV):

LTV = monthly revenue per customer × gross margin ÷ monthly churn

Here: 30 × 0.80 ÷ 0.04 = $600. Dividing by churn assumes churn stays constant, which is a strong assumption, and it is least reliable early on. Skipping gross margin overstates LTV; in this example LTV would be $750 instead of $600, an error of 25%.

The ratio is LTV ÷ CAC = 600 ÷ 150 = 4. Many sources say 3 or more is good. There is no solid source for that threshold, so treat it as a rule of thumb. The ratio is also unstable with few customers, because small changes in churn move LTV a lot. Payback is usually easier to act on: it tells you how long your cash is tied up.

Cash metrics

If you are not yet profitable, two more numbers matter more than any ratio:

• Burn = cash going out minus cash coming in, per month.
• Runway = cash in the bank ÷ monthly burn. With $30,000 and a $3,000 monthly burn, runway is 10 months.

Cash metrics are often mentioned only in passing. For a young company, running out of cash ends the experiment before the ratios matter.

What to track at each stage

With a handful of customers, do not compute ratios. Keep a list and write one line for each person who leaves. That tells you more than a percentage of five.

How to set it up cheaply

A spreadsheet is enough. One tab per month, with the starting, new, expansion, contraction and churned figures, then the formulas above. Add a column for where each customer came from.

If your billing tool exports subscriptions, use that as the source of truth for MRR and churn, and compare it against your sheet once a month. For product usage and traffic, see the analytics options in SaaS tools for startups. Whatever you use, write down your definitions on one page: what counts as an active customer, whether trials are included, how annual plans are treated, and whether CAC includes your time. Then keep them fixed so months stay comparable.

Annual plans, trials and free tiers

Three situations break naive formulas:

• Annual plans. Count one-twelfth of the annual price as MRR each month. Churn only shows at renewal, so a young annual business may look like it has no churn for a year. That is timing, not safety.
• Free trials. Decide whether a trial user is a customer. The usual answer is no: count them as customers when they first pay, and track trial-to-paid separately.
• Free plans. Free users are not revenue. Track free-to-paid conversion on its own, and do not let free users lower your churn average.

For how these choices connect to what you charge, see how to price a SaaS product.

Using the numbers to decide

Numbers only matter if they change what you do next. With the example month above, a founder might reason like this (still hypothetical):

• Churn of 4% a month is the biggest leak. Losing about 39% of customers a year means acquisition has to run to stand still. Before spending more on acquisition, talk to the four who left and find the common reason.
• Payback of about six months is tolerable if cash allows. The question is whether you can fund six months of waiting for each new customer. Check it against your runway.
• An NRR of 97% says the base shrinks slightly even with upgrades. If upgrades are working, find out why, and offer them to more customers.
• The ratio of 4 looks healthy but rests on churn staying at 4%. If churn gets worse, it falls quickly.

When you are deciding where to put the next hour, this is how saas metrics for startups earn their place: each number points to a conversation or an experiment. Write down one hypothesis per month ("if we fix onboarding step two, churn drops"), change one thing, and compare the next month's numbers. Do not change five things at once, because you will not know which one worked.

A final note on honesty with yourself and others. If you share metrics with advisers, customers, buyers or investors, give the definitions with the numbers. "MRR of $3,210, excluding trials and one-off fees, annual plans spread over twelve months" is a statement someone can check. A bare figure is not. And do not round up: a small, accurate number is more useful than a large vague one.

Why I give no benchmarks

Advice on SaaS metrics is full of targets: churn bands, retention levels, payback limits, margins. Some cite named reports, but the numbers differ from source to source. Even where a source is real, benchmarks depend on company size, price point, customer type and survey sample, and a young product with 40 customers is a poor fit for averages from larger firms.

A better use of your time:

• Compare yourself to last month.
• Compare groups of your own customers with each other.
• Treat a benchmark as a question to ask, not a score to reach.

If you want external numbers later, find the original report, read how it was sampled, and note its date.

Mistakes to avoid

Wrong formulas. Check each against your own example, as above. Several guides contain errors.

Mixing MRR and net new MRR.

Counting trials or free users as customers.

Leaving gross margin out of LTV and payback.

Ignoring your own time in CAC.

Reading a percentage from a tiny base. Four customers out of 100 is 4%; one out of 10 is 10%.

Changing definitions midstream. You lose the ability to compare months.

Tracking too much. Pick a handful and look at them on a fixed day each week.

Using ratios to avoid talking to customers. The numbers tell you what; only conversations tell you why. See how to get your first SaaS customers.

Where CircleSaaS fits

If you list your product in a directory, treat it like any channel: tag the link and measure what comes back. CircleSaaS is a directory shaped as a circle of 10,000 tiles, and I do not have referral data to show you yet. Add a UTM tag to your link, look at visits and sign-ups after a few weeks, and put the result in your CAC sheet. If a channel costs money and returns nothing measurable, drop it. The launch checklist has a step for setting up that tracking before you announce anything.

What may change in 2027

These are things to watch, not predictions.

• Usage-based and credit-based pricing make "MRR" less predictable. If you adopt them, decide how to treat variable revenue.
• Billing tools keep adding built-in metrics. Check how they define churn and MRR before you trust them.
• AI features change cost to serve. If each use has a model cost, gross margin may move month to month, so recompute payback regularly.
• Investors and buyers ask for definitions. Written definitions help whenever someone asks how you calculate a number.

Frequently asked questions

Which SaaS metrics should I track first?

Revenue (MRR), churn and acquisition cost. Add NRR and payback once you have a few dozen paying customers.

What is a good churn rate?

There is no reliable benchmark, and the right level depends on price, customer type and stage. Track your own, find out why each customer left, and aim to improve month on month.

Do I need LTV?

It helps, but it is a rough estimate that depends on stable churn. Payback is easier to act on early.

How do I calculate churn with annual plans?

Count churn at renewal, and look at the share of customers who renewed. Do not assume annual plans mean zero churn.

Should CAC include my time?

For a true picture, yes. At least record both versions so you know the difference.

Is a spreadsheet enough?

Yes, until the volume makes it painful. Move to a billing tool's reports when you have enough customers to need them, and check their definitions first.

Summary

Keep SaaS metrics for startups few and consistently defined: MRR, churn, net revenue retention, CAC payback and the LTV to CAC ratio, each calculated in a spreadsheet with your own written definitions. Check formulas against a small worked example, skip borrowed benchmarks, and use the numbers to decide where to spend your next hour.