Open any analytics dashboard and you get forty numbers. Sessions. Signups. Bounce rate. Daily actives. Email open rate. Feature adoption. For a software business doing less than a few thousand a month, almost all of it is noise. Five numbers are not noise. Learn these five, check them once a month, and you can form a reasonable view of whether the business survives the next year.
What makes these five different is that they connect to each other. Pageviews do not cause revenue. Follower count does not cause revenue. But churn feeds lifetime value, lifetime value divided by acquisition cost tells you whether growth is profitable, and everything rolls up into monthly recurring revenue. It is a closed system. Change one input and you can trace exactly where the effect lands.
MRR is the heartbeat
Monthly Recurring Revenue is active paying customers multiplied by average revenue per customer. If you have 40 customers paying an average of $28, your MRR is $1,120. That is the whole formula.
Two things trip people up. Annual plans should be normalised: divide the annual price by twelve and count a twelfth each month, rather than booking the full amount in the month it was paid. And one-time or lifetime purchases do not belong in MRR at all. If you sell a lifetime deal, count the cash in your bank balance, not in recurring revenue, or you will convince yourself you are growing when you are actually selling your future revenue at a discount.
The absolute number matters less than its composition. Net new MRR is new customer revenue, plus expansion from upgrades, minus contraction from downgrades, minus churned revenue. Two businesses can both add $400 of MRR in a month. One added $400 in new customers and lost nothing. The other added $900 and lost $500. Those are completely different businesses, and only the composition tells you which one you are running.
Churn is the leak in the bucket
Churn is the percentage of customers you lose in a period. Ten customers lost out of 100 at the start of the month is 10% monthly churn. The arithmetic is unforgiving. At 5% monthly churn, about 54% of a cohort is still around after twelve months, which means you replace your entire customer base in a bit under two years just to stand still. At 10%, roughly a third survives the year.
Track two versions. Logo churn counts customers. Revenue churn counts dollars. They diverge when your small accounts leave and your large ones stay, or the reverse, and that divergence is informative. If logo churn is high but revenue churn is low, you are losing the customers who were never a good fit and that is survivable. If revenue churn is higher, your best customers are leaving and that is an emergency.
A churn number on its own is a symptom, not a diagnosis. Segment it by how long people stayed. Cancellations inside the first thirty days usually mean an onboarding or positioning problem: they bought something other than what you built. Cancellations at month six or later usually mean a value problem: the product worked, but not enough to keep paying for. Those two failures have completely different fixes, and an average churn rate hides which one you have.
LTV is what a customer is worth before they leave
Lifetime Value is average revenue per customer, multiplied by your gross margin, divided by your monthly churn rate. A $20 per month product with a 90% margin and 5% monthly churn gives you 20 × 0.9 ÷ 0.05, which is $360 of lifetime value per customer.
Gross margin is the part people skip. For software it is usually high but never 100%: payment processing takes a few percent, and hosting, storage and third-party API costs come off the top. If you publish media to a dozen networks and store video, your storage and egress bills are a real cost of goods. Ignore them and your LTV is inflated by exactly the amount you ignored.
Early on, LTV is a projection built on very little data. If your product is eight months old you do not have evidence of a two-year customer lifetime, no matter what the formula says. A useful discipline is to cap the assumed lifespan at 24 months. It keeps you honest, and it stops you justifying acquisition spend on the back of a lifetime you have never observed.
CAC is what you paid to get them
Customer Acquisition Cost is everything you spent on getting customers in a period, divided by the number of new customers in that period. Ad spend, sponsorship fees, the contractor who edits your videos, the tools that exist purely for marketing. Add it up, divide, done.
The honest version includes your own time. Founders love to describe organic content as free, and it is the most expensive thing they do. If you spend fifteen hours a week writing posts and you would charge $60 an hour for that work, that is $3,600 a month of acquisition cost that never shows up in the spreadsheet. You do not have to pay yourself, but you should count it, or you will conclude that organic is infinitely efficient and paid is a rip-off when the truth is more nuanced.
Split blended CAC from paid CAC. Blended divides all acquisition spend by all new customers, including the ones who found you through a friend. Paid CAC divides only paid spend by only paid-sourced customers. Blended tells you what the business costs to grow. Paid tells you whether a specific channel is working. Confusing the two is how people conclude their ads are cheap when in fact word of mouth was carrying them.
The LTV to CAC ratio decides whether growth is profitable
Divide lifetime value by acquisition cost. Below 1, you lose money on every customer you acquire and growth actively destroys the business. Around 3 is the widely cited healthy target: you earn back three times what you spent, leaving room for the costs the formula ignores. Much above 5 usually means you are underinvesting in growth and could afford to buy more customers than you are buying.
Pair the ratio with CAC payback period, which is CAC divided by monthly gross profit per customer. It tells you how many months until a customer has repaid what you spent to acquire them. For a bootstrapped business this is often the more important number, because a 4:1 ratio with an eighteen-month payback will still run you out of cash. Ratios describe eventual profitability. Payback describes whether you survive long enough to see it.
- MRR — active customers × average revenue per customer, with annual plans divided by 12.
- Churn — customers lost ÷ customers at the start of the period. Track logo and revenue churn separately.
- LTV — average revenue × gross margin ÷ monthly churn rate. Cap the assumed lifespan at 24 months.
- CAC — all acquisition spend ÷ new customers, including a fair value for your own time.
- LTV:CAC — aim for 3 or higher, and check CAC payback period alongside it.
How the five feed each other
Here is why churn deserves more attention than it usually gets. At 6% monthly churn, the average customer lifespan is about 17 months. Cut churn to 4% and the average lifespan becomes 25 months. You just increased lifetime value by roughly half without raising your price, without spending a cent more on acquisition, and without adding a single new customer. That is the highest-leverage fix available to most small software businesses, and it is almost always cheaper than buying more traffic.
The same logic runs the other way. If your LTV to CAC ratio is bad, you have three levers and only three: charge more, keep customers longer, or acquire them more cheaply. Every growth tactic you read about is an instance of one of those three. Knowing which lever your numbers are asking you to pull saves you from spending a quarter on the wrong one.
Calculating them without a data team
Set aside an hour on the first of each month. Stripe gives you MRR, churn and average revenue directly, and its billing reports will get you most of the way to LTV. Your bank statement and calendar give you CAC. Write the five numbers in a spreadsheet, one row per month. Twelve rows in, the trend will tell you more than any dashboard.
The hard half is CAC by channel, because you have to know which channel produced which customer. This is where most solo founders give up and use a blended number forever. It is also the specific gap seenpaid was built to close: it publishes to your social accounts and reads your Stripe, so a payment can be traced back to the post and platform that produced it. Once revenue is attributed per channel, per-channel CAC stops being a guess.
What these five will not tell you
All five are lagging indicators. They describe what already happened. They will tell you churn rose last month, but not why. They will tell you CAC doubled, but not that the reason is a competitor bidding on your brand terms. The numbers set the agenda; conversations with customers supply the answers.
So pair them. Check the five monthly, and when one of them moves in a direction you did not expect, go and talk to five people it happened to. The metric is the alarm. The conversation is the diagnosis. Founders who only do the first drown in dashboards; founders who only do the second get very good anecdotes and no idea whether the business is working.
Start with MRR and churn this month. Add the other three next month. You do not need a data warehouse, an analytics hire, or a business intelligence tool to run a small software business well. You need five numbers and the discipline to look at them when they are ugly.