SaaS Metrics for Founders: MRR, Churn, Runway and What to Track First

The first year of a subscription business produces a startling amount of numbers. Your billing tool has a dashboard, your analytics tool has another, and every founder forum has a list of thirty metrics you apparently must track. Most of them are noise at your stage. With a few dozen customers, a handful of figures tell you almost everything: how much recurring revenue you have, whether it is growing, whether customers stay, what it costs to win them, and how long your cash lasts.

This guide covers those numbers in the order worth tracking them, with plain definitions, the mistakes that quietly distort them, and a setup that needs nothing more than a billing export and a spreadsheet. It is written for first-time founders selling subscriptions to small businesses, where volumes are low, plans are cheap and every customer matters.

Key takeaways

Why fewer metrics beat more metrics

A metric earns a place on your sheet only if a change in it would change what you do next week. Revenue growth might lead you to hire or hold. A rising cancellation rate might lead you to call customers. A count of page views leads to nothing.

1. MRR and ARR: the foundation

Monthly recurring revenue (MRR) is the predictable subscription revenue you would expect to bill in a normal month, with every customer converted to a monthly amount. Annual recurring revenue (ARR) is simply MRR multiplied by twelve. Use whichever your customers think in; many small-business products live in MRR, and ARR becomes more natural once annual contracts dominate.

A single total hides the interesting part, which is how it moved. Break the change in MRR each month into four pieces:

Net new MRR is new plus expansion, minus contraction and churn. Seeing the four parts separately stops you celebrating a good sales month that was quietly cancelled out by cancellations, and it shows whether growth comes from winning customers or from growing the ones you have.

2. Customer churn and revenue churn

Customer churn is the share of customers at the start of a period who cancelled during it. If you began the month with 40 customers and 2 cancelled, customer churn for the month is 2 divided by 40, or 5 percent. Revenue churn is the same idea for money: churned MRR (and often contraction too) divided by MRR at the start of the month.

They diverge when your customers are unequal. Losing three of your smallest plans might be 7 percent customer churn but only 2 percent revenue churn. Losing one large account might be 2 percent customer churn and 10 percent revenue churn. You need both to know whether you have a broad retention problem or a concentration problem.

• Count voluntary and involuntary churn separately. Some customers choose to leave; others are lost because a card failed. The second kind is often recoverable with retries and reminders, and it needs a different fix.
• Define the moment of churn once and keep it. The date the customer asked to cancel and the date the subscription ended are different. Pick one and apply it every month.
• Do not quote a “good” churn rate as a target. Acceptable churn depends heavily on who you sell to, the price and the contract length. Businesses selling to very small companies usually see more churn than those selling to larger ones. Track your own trend and ask whether it is improving.

3. Net revenue retention, simply

Net revenue retention (NRR) answers one question: if you only looked at the customers you had a year ago, how much recurring revenue do they pay you now, as a percentage of what they paid then? It includes upgrades, downgrades and cancellations, and it excludes anyone who joined since.

A simple version: take the MRR from customers who existed twelve months ago, add their expansion, subtract their contraction and churn, and divide by the starting figure. Above 100 percent means your existing customers are collectively paying you more than before, so the base grows even with no new sales. Below 100 percent means the base is shrinking and new sales have to fill a leak first.

4. ACV and average revenue per account

Average revenue per account (ARPA) is MRR divided by the number of paying customers. Annual contract value (ACV) is the annualised value of a typical contract, which for a monthly plan is roughly ARPA times twelve. These tell you how big your typical customer is, and that drives nearly every decision about how you sell.

5. Gross margin, and why it matters more for AI products

Gross margin is the share of revenue left after the direct cost of delivering the service: hosting, payment fees, third-party services you pay per use, and the support time directly tied to serving customers. Revenue minus those costs, divided by revenue.

For classic software this number is often comfortable and stable, so founders skip it. AI products are different. Every generation, analysis or automated action can carry a real per-use cost from a model provider, and that cost scales with how much customers use the product. A heavy user on a flat plan can cost more to serve than they pay. That is why margin deserves an early place on your sheet if your product leans on AI: it is the check on whether growth is making the business healthier or just busier.

6. CAC and payback period

Customer acquisition cost (CAC) is total sales and marketing spend in a period divided by the number of new customers won in that period. Include everything: ad spend, tools, contractor fees, and a fair share of salaries if people spend their time selling. Founders often leave their own time out, which flatters the number; at least note it.

The payback period is how many months of gross profit from a new customer it takes to recover the CAC. Take CAC, divide by the monthly gross profit per customer (ARPA times gross margin), and you have it in months. It is more useful than most ratios at this stage because it uses only things you can observe now: what you spent, what customers pay and what delivery costs.

7. LTV: useful idea, shaky number

Customer lifetime value (LTV) estimates the total gross profit a customer will generate before leaving. The usual shortcut divides monthly gross profit per customer by monthly churn, which assumes churn stays constant forever. That assumption is the weak point.

With few customers and a short history, the churn figure behind LTV is barely measured. One extra cancellation can halve your estimated lifetime. A pricing change, a new segment or a better onboarding flow will shift retention in ways your early data cannot predict. So an LTV figure in year one is a hypothesis, and an LTV-to-CAC ratio built on it is a hypothesis divided by a noisy number.

8. Burn, net burn and runway

Gross burn is the cash you spend each month in total. Net burn is cash spent minus cash received from customers: the amount your bank balance actually falls. Runway is cash in the bank divided by net burn, expressed in months.

Annual prepayments inflate cash without being revenue you have earned yet, so track cash and MRR separately.

9. Leading indicators: activation and trial-to-paid

MRR and churn are lagging: they report what already happened. Two earlier signals tell you what is coming.

Activation is the share of new sign-ups who reach the first moment where the product delivers value. You define that moment, and it should be a concrete action: sent a first campaign, imported contacts, created a first deal. Customers who never activate rarely stay.

Trial-to-paid conversion is the share of trial users who become paying customers. Measure it per sign-up cohort, after the trial period has fully elapsed for everyone in the group, otherwise recent cohorts look worse than they are. Keep trials separate from paying customers in every other metric; a trial is a prospect, not a customer.

10. The cohort view, in plain words

A cohort is a group of customers who started in the same month. A cohort view lists those groups down the side and months since sign-up across the top, showing how many from each group are still paying after one month, two, three and so on.

Overall churn blends new customers, who leave most often, with older ones, who rarely do, so cohorts show whether changes you made actually worked.

Vanity metrics to ignore

These are not worthless, but they are poor decision tools on their own:

• Total sign-ups. Includes people who never activated and people who left. Track activated sign-ups instead.
• Page views and visitors. Useful for diagnosing a funnel, misleading as a scoreboard. Traffic that never converts is a cost.
• Total users ever. Counts everyone who ever joined, including churned accounts. It can only go up, so it tells you nothing.

A simple monthly metrics sheet

One row per month, filled in on the same day each month, is enough. Suggested columns:

The notes column matters more than it looks. Six months later, a spike you cannot explain is useless, and one you can explain is a lesson.

Tracking this without expensive tools

You do not need a metrics platform with a handful of customers. Most billing systems can export subscriptions, invoices and payments as a spreadsheet file. A reliable routine looks like this:

• Export subscriptions and invoices at month end into a fresh tab.
• Convert every plan to a monthly amount. Divide annual plans by twelve and quarterly plans by three.
• Mark each customer as new, expanded, contracted, churned or unchanged compared with the previous month.
• Sum those groups into the sheet above using simple formulas.
• Subtract refunds and credits so the figures reflect money you actually kept.

A weekly 15-minute founder review

Monthly numbers tell you what happened; a short weekly check keeps you from being surprised. Block the same fifteen minutes each week and go through these in order:

• Money (3 minutes). Cash balance, any failed payments to chase, runway if anything big changed.
• Customers (3 minutes). New paying customers and cancellations this week, with the reason for each cancellation.
• Funnel (4 minutes). Sign-ups, activations and trials ending this week; who is close to converting and needs a nudge.
• Acquisition (3 minutes). Where this week’s new leads came from, and whether the channel spend still makes sense.
• One decision (2 minutes). Write down the single thing you will do differently next week because of what you saw.

Common mistakes that distort the numbers

• Counting one-off fees in MRR. A setup charge or a single project inflates the month and then vanishes, making next month look like a collapse.
• Mishandling annual plans. An annual payment is not that month’s MRR. Divide it by twelve. Counting the whole payment in one month makes one month look huge and the other eleven look like decline.
• Mixing trials with paying customers. Trial users in your customer count lower your ARPA, distort churn and make CAC look better than it is.
• Ignoring refunds and credits. Leaving them out overstates revenue and hides dissatisfaction.

How Scalry handles this

Scalry is a sales and marketing platform, and it is worth being clear about what that means here: it is not a billing-analytics tool. Subscription metrics such as MRR, churn and net revenue retention should be computed from your billing data, in a spreadsheet or an analytics tool, using the approach above. Scalry does not replace that.

What it helps with is the front half of the story: where leads come from and how they move towards becoming customers. The CRM tracks deals with a value, an expected close date and a stage, and records the source of each lead. Funnels capture UTM data along with each lead, so you can see which campaigns and channels the people who eventually pay arrived from. That feeds directly into the acquisition side of your sheet: new customers by channel, and the pipeline behind next month’s numbers.

Questions to ask before you commit

Whichever way you are leaning, these are the questions that change the answer. Work through them against your own numbers rather than anyone's feature matrix — including this one.

• Can I state my MRR, and does it exclude one-off fees? If annual plans are in there at full value, it is wrong.
• Do I know my net burn and runway this month? Not last quarter.
• What is my CAC over the last three months, including my own time?
• Where did my last ten paying customers come from? If you cannot say, source tracking is your first gap.

If you cannot answer the first three, you are not ready to choose between Scalry and your current metrics yet — you are still deciding what you need, which is a cheaper problem to solve first.

A plan for your first month of tracking

In week one, export subscriptions, invoices and refunds and write down your definitions of MRR, churn and a paying customer. In week two, build the monthly sheet and backfill what your data allows. In week three, choose an activation event and compute trial-to-paid for finished cohorts. In week four, hold your first fifteen-minute review and write down one decision. After a few months you will have a trend, and trends are what small numbers are good for.

Frequently asked questions

What is the difference between MRR and ARR?

MRR is monthly recurring revenue: the subscription revenue you expect in a normal month. ARR is annual recurring revenue, which is MRR multiplied by twelve. They describe the same revenue on different time scales. Use MRR while most customers pay monthly and ARR once annual contracts are common.

What is a good churn rate for a SaaS startup?

There is no single number that applies to everyone. Benchmarks vary by customer segment, price point and contract length, and products sold to very small businesses usually lose customers faster than those sold to larger companies. Track your own churn month by month, look at the raw counts as well as the percentage, and focus on whether the trend is improving.

How do I calculate runway?

Divide the cash you have in the bank by your net monthly burn, which is monthly spending minus the cash customers pay you. Recalculate it every month, because both the balance and the burn change. Treat annual prepayments carefully, since they raise cash without being revenue you have fully earned.

How do I put annual plans into MRR?

Divide the annual price by twelve and count that amount as MRR in each month the plan is active. Do not count the whole payment in the month it is received. Quarterly plans are divided by three in the same way.

Can I trust LTV and the LTV to CAC ratio early on?

Not much. LTV depends on churn, and with few customers and a short history your churn figure is barely measured, so one cancellation can change the result a lot. Treat early LTV as a rough guess. Payback period, retention by sign-up month and whether your oldest customers are still paying are more dependable signals.

What is the difference between customer churn and revenue churn?

Customer churn is the share of customers who cancelled. Revenue churn is the share of recurring revenue that was lost. If the customers who leave are your smallest, revenue churn will be lower than customer churn; if you lose a large account, it can be much higher. Tracking both shows which problem you have.

Do I need special software to track SaaS metrics?

No. A billing export and a spreadsheet are enough in the first year. Convert every plan to a monthly amount, classify each customer as new, expanded, contracted or churned, and sum the groups. Consider a dedicated analytics tool once the manual work becomes slow or error-prone.