Product Metrics: The Numbers That Show Whether Your Product Works

Most Founders paying for a build get shown a dashboard full of good news. Signups are up, downloads are up, and the user count only ever climbs. It's very hard to tell from product metrics like these whether the product is working, or whether you're paying $10,000 a month to watch a number that can't go down.

I recently wrote Business Finance 101 for The Wandering Pro, a community built around helping people grow their careers and businesses. It covers margins, cash flow and the one number that matters for a services business. Writing it made a gap obvious: the money side of a business has a standard set of numbers every owner learns, and product metrics mostly get guesswork. This guide is the product side, written from the Founder's chair, for someone paying for a product and trying to tell a real signal from a flattering one.

Paying Customers Come Before Product Metrics

A fair warning before the numbers. Product metrics matter for any product business, and this guide is all about them. But in our own consulting work, we regularly meet Founders who are deep in dashboards, funnels and analytics tools before a single customer has paid them. It feels like progress. It's motion, and it crowds out the only work that counts at that stage: getting someone to pay.

If that's where you are, take this guide with a big bag of salt, the kind we've seen clients eat in our day to day work. Until you have paying customers, your one number is how many you have, and your best analytics tool is a conversation with each of them. Ask why they bought, what they use it for, and what would make them leave.

The metrics below start earning their place once a single customer can't swing them. With 10 customers, one cancellation is 10% churn, and the number tells you more about that one account than about the product. As a rule of thumb, somewhere around 30 to 50 paying customers, with at least 3 months of history, is where churn, LTV and the rest start describing the business rather than a few individual customers. Before that, track them lightly and talk to your customers a lot.

The Example Behind Every Number

To keep the math consistent, every calculation below uses the same example. It's a made-up B2B tool, with numbers chosen to be realistic and nothing more:

• 120 paying customers at an average of $250/month, so $30,000/month in recurring revenue
• An 80% gross margin, so $200 of gross profit per customer each month
• 3% of customers cancel each month
• $9,000/month spent on sales and marketing, winning about 4 or 5 new customers a month

We'll come back to it in every section, so by the end you can see how the numbers connect, and what a buyer would pay for it.

EBITDA Looks Different in a Product Company

EBITDA is earnings before interest, taxes, depreciation and amortization. It strips out how the business is financed, what it pays in tax and the paper cost of writing assets down, so what's left is what the business itself earns.

In the services example in Business Finance 101, EBITDA came in at 21.5% of revenue against a pre-tax Net Profit of 20%. The advice there was to recognize EBITDA and not obsess over it until someone is trying to buy you. For a services business that's right. There's little to depreciate, so the 2 numbers sit close together.

A product company is different. Building software is an investment that gets written down over years, so depreciation and amortization are real and large, and interest shows up more often too. A year of our example product looks like this:

• Revenue: $360,000
• Cost of Revenue: $72,000 (hosting, support, payment fees)
• Gross Profit: $288,000 (80%)
• Sales and Marketing: $108,000
• Product and Team: $96,000 (including the Founder's $60,000 salary)
• General and Admin: $18,000
• EBITDA: $66,000 (18%)
• Depreciation and Amortization: $24,000 (development written down)
• Interest: $6,000
• Tax: $7,560
• Net Profit: $28,440 (8%)

The gap between EBITDA and net profit went from about 1.5 points of revenue in the services business to about 10 points here. That gap is why EBITDA matters for products: it shows what the product engine earns before financing and accounting choices, and it's the number a buyer or investor compares across companies.

One caveat. EBITDA isn't a standardized accounting measure, and it ignores what you spend on the platform and equipment that keep the revenue coming, which is why Warren Buffett has criticized it for decades. Use it to compare. Never make it your only number.

The Product Metrics That Tell You Whether It Works

EBITDA tells you what the business earns. It can't tell you why, or whether next year looks better. For that you need a small set of product metrics, and they follow the path a customer takes: you pay to acquire them, they get value from the product or they don't, they stay or they leave, and the best of them spend more over time.

Most of these apply to any product. 2 blocks further down depend on how you sell, one for subscriptions and one for one-off sales, so skip the one that isn't yours.

3 habits make any of them more useful. Read the trend over at least 3 months, because a single month is mostly noise. Split the number by cohort or by channel before you trust an average, since an average is where problems go to hide. And decide in advance what number would worry you, so a bad month gets treated as a signal instead of explained away.

What a Customer Costs: CAC and Payback

Customer Acquisition Cost (CAC): everything you spend to win customers in a period, divided by the customers you won. Count all of it: ads, sales salaries, tools, outside help, the content budget. Our example spends $9,000 a month and wins 4.5 customers on average, so its CAC is $2,000.

On its own, that number means little. What matters is how long it takes to earn back.

CAC Payback Period: CAC divided by the gross profit one customer brings in each month. Our example earns $200 of gross profit per customer per month, so it takes 10 months to earn back the $2,000. Under 12 months is healthy for a small B2B product. Past 24 months, you're financing your customers, and every new sale makes the cash position worse before it makes it better.

When a team reports acquisition, ask for CAC broken out by channel. A blended CAC can hide one channel that works and 2 that burn money. Ask too whether the Founder's own time on sales calls is in there. It usually isn't, and for an early product it's often the biggest acquisition cost of all.

Conversion Rates: the steps underneath CAC, from visitor to signup, signup to trial, and trial to paid. You don't need every one on a Founder's dashboard. You need to know which step loses the most people, because that's where the next dollar of improvement is cheapest.

Does It Last: Retention, Churn and Lifetime Value

Acquisition gets the attention. Retention decides whether any of it was worth paying for.

Churn Rate: the share of customers who cancel in a period. Our example loses 3% a month, about 3 or 4 of its 120 customers. That sounds small until it compounds. At 3% a month, only about 69% of today's customers are still around in a year. Products sold to small businesses commonly run at 3% to 7% a month, and even at the low end it's a leak worth fixing.

Retention Rate: the other side of churn, and the more useful way to watch it. Track it by cohort, meaning the customers who joined in the same month, followed over time. A healthy cohort curve drops early and then flattens out. A curve that never flattens means the product hasn't yet found the people who really need it.

Take 2 of our example's cohorts, measured as the share of customers still paying after they joined:

• January Cohort: 95% still paying after 1 month, 88% after 3, 81% after 6 and 72% after 12
• April Cohort: 92% after 1 month, 82% after 3 and 74% after 6, still running

The April cohort is dropping faster at every point. Something changed between January and April, a new channel, a pricing change or a rough onboarding release, and the blended churn number would take months to show it. The cohort view shows it in weeks.

Customer Lifetime Value (LTV): the gross profit one customer brings in before they leave. The simple version is monthly gross profit per customer divided by monthly churn. For our example, that's $200 ÷ 3%, or about $6,700.

LTV:CAC Ratio: the number that ties acquisition to retention. Our example earns about $6,700 from a customer it paid $2,000 to win, a ratio of roughly 3.3 to 1. The usual rule of thumb is 3 to 1. Below it, growth costs more than it returns. Far above it, you're probably underspending on growth.

Now watch what retention does to the whole picture. If our example cut churn from 3% to 2% a month, the average customer would stay about 50 months instead of 33. LTV rises to $10,000, and the ratio climbs to 5 to 1, without touching the price or the ad budget. For most products, retention is the cheapest growth there is.

Average Revenue per User (ARPU): total revenue divided by customers, $250 for our example. Watch the trend more than the number. A falling ARPU alongside a rising customer count often means discounts are doing the selling.

Is It Actually Used: Activation and Stickiness

A customer can pay you and still barely use the product. Usage is where churn shows up first, usually months before the cancellation does.

Activation: the share of new customers who reach the first moment of real value. You have to define that moment for your own product, and defining it is half the work. Our example is a reporting tool, so activation means sending a first report to a client within 7 days of signing up. 60% of new accounts do it. The other 40% are most of the churn you'll see 3 months from now.

If you don't know your activation event yet, work backwards from the customers who stayed. Take everyone still paying after a year and look at what almost all of them did in their first week that the customers who left didn't. That action is your activation event. It's rarely the signup or the first login. It's the first time the product did the job someone bought it for.

Time to First Value: how long a new customer takes to get there. Shortening it is some of the highest return product work there is, because every day a new account spends confused is a day it can talk itself out of paying.

Stickiness (DAU/MAU): daily active users divided by monthly active users, which reads roughly as how often a typical user comes back. Our example sits at 15%, about once a week, which suits a tool people use for weekly reporting. Always read it against how often the job actually happens. A payroll tool used twice a month is doing fine at 10%, and a chat tool at 10% is in trouble.

Usage is also where a report can quietly flatter. "Active users" only means something once you know what counts as active, and logging in isn't using. Launching without a defined usage signal is one of the MVP launch mistakes that hides a failing product for months.

Unit Economics: Contribution Margin

Contribution Margin: what's left from each sale after the costs that grow with it, like hosting, payment fees, support time and onboarding. It's the per-unit version of the gross profit lesson in Business Finance 101. Our example keeps $200 of every $250, and gross margins above 75% generally read as healthy for software. Watch for products with a service hidden inside them. If every new customer needs 10 hours of setup from your team, that time is a cost of the sale, and it belongs in the margin.

Product Metrics for Subscription Products: MRR, ARR and NRR

Skip this block if you don't sell subscriptions.

MRR and ARR: monthly recurring revenue and its annual version. Our example runs at $30,000 MRR, which is $360,000 ARR. Count only recurring revenue. One-off setup fees and annual contracts paid up front get spread across their months or left out, otherwise one good month looks like a trend.

MRR Growth Rate: the month over month change, best read as a bridge. New MRR from new customers, plus expansion MRR from upgrades, minus contraction from downgrades and churned MRR from cancellations. The total can look fine while churn quietly gets worse underneath it.

Net Revenue Retention (NRR): the most important health number in a subscription business. Take the customers you had a year ago and compare what they pay today, including upgrades and after cancellations and downgrades. Above 100% means the business grows even if it never signs another customer.

Our example started the year with $25,000 a month from its existing customers. Over 12 months, cancellations took $7,500 of that, downgrades took $500, and upgrades added $3,000, which leaves $20,000. That's an NRR of 80%. Every year, a fifth of the revenue base has to be replaced before the business grows at all, and that's where the 3% monthly churn finally shows up in dollars.

Gross Revenue Retention (GRR): the same calculation without the upgrades, 68% for our example. It shows how much of the base is truly holding on its own.

Expansion Revenue: the upgrades by themselves. In many subscription products, the cheapest new revenue comes from customers who already trust you.

Product Metrics for One-Off Sales: Order Value and Repeat Rate

Skip this block if you sell subscriptions. When customers buy one purchase at a time, there's no recurring revenue to lean on, so lifetime value has to be built from how they behave.

Average Order Value (AOV): revenue divided by the number of orders.

Purchase Frequency: how many orders a customer places in a year.

Repeat Purchase Rate: the share of customers who buy more than once. It's the one-off world's version of retention, and it's usually the number that separates a business from a string of launches.

LTV for One-Off Sales: AOV × purchase frequency × gross margin × the years a customer stays. Suppose our example sold report packs instead of a subscription, at $300 a pack, 3 packs a year, an 80% margin and a 2 year customer life. That's $300 × 3 × 0.8 × 2, or $1,440 of lifetime gross profit, against the same $2,000 CAC. A ratio under 1 to 1 is a clear signal that this product shouldn't be sold this way.

Putting Your Product Metrics on One Page

Our example's monthly scorecard fits on one page, with the usual healthy range next to each number:

• CAC: $2,000 (the healthy level depends on price, so judge it by payback)
• CAC Payback: 10 months (healthy: under 12 months)
• Monthly Churn: 3% (healthy: under 2% for B2B)
• LTV:CAC: 3.3 to 1 (healthy: 3 to 1 or better)
• Activation (First Report in 7 Days): 60% (healthy: rising month on month)
• DAU/MAU: 15% (healthy: matches how often the job happens)
• Gross Margin: 80% (healthy: above 75% for software)
• NRR: 80% (healthy: 100% or more)
• North Star (Reports Sent Each Week): 410 (healthy: rising month on month)

Read down the list and the story writes itself. The example acquires customers efficiently and keeps healthy margins, but it leaks them: churn sits above the range, NRR sits well below it, and the LTV:CAC ratio only clears 3 to 1 because the margin carries it. The next dollar belongs in onboarding and activation, not in more ads. That's the kind of call a single page of the right numbers lets a Founder make in a few minutes.

Vanity Metrics vs the Product Metrics Worth Asking For

A vanity metric is one that can only go up: total registered users, total downloads, raw signups, page views, followers. None of them is false. They just can't tell you when something is going badly, which is exactly why they end up on the slide.

Each one has a real counterpart, one that's capable of disappointing you:

• Total signups → activation rate
• Total downloads → retention by cohort
• Daily active users → stickiness, measured against a defined action
• Revenue this month → the MRR bridge and NRR
• Customer count → LTV:CAC
• "Users love it" → churn, and the reasons people give when they leave

If the report you get every month only has the left column, ask for the right one. A team that can't produce it isn't necessarily hiding anything. It does mean nobody is measuring the things that would show the product failing, which is worth knowing, and it's a large part of what we look at in a Product Team Audit.

Pick One Number: Your North Star

Every metric above answers a narrow question. A North Star Metric is the one number that best measures the value customers actually get, chosen so that when it grows, the business grows with it.

A good North Star passes 3 tests:

• It measures value delivered to the customer, rather than activity
• It moves before revenue does, so it works as an early signal
• The whole team can influence it

For our example, revenue is the obvious pick and the wrong one, because it lags. A better North Star is reports sent to clients each week. A customer sending reports is getting value, and customers getting value renew. Our example sends about 410 a week across its 120 accounts, and when that number dips, churn tends to follow a month or 2 later.

A North Star doesn't replace everything else. It's the number at the top of the page, with CAC, retention and margin underneath it explaining why it moved.

How Buyers Read Your Product Metrics

Business Finance 101 said not to obsess over EBITDA until someone is trying to buy you. This is the part where someone is. Every number in this guide is something a buyer will look at, through one question: how much of this revenue will still be here after they own it?

The Multiple: small products sell for a multiple of profit, or sometimes of revenue. Acquire.com's January 2026 report put the median SaaS profit multiple at 3.9x in both 2024 and 2025, and their guidance for Founder led SaaS is 3x to 5x annual profit, or 1x to 3x annual revenue. The double digit multiples in headlines belong to fast growing, venture backed companies. Even public SaaS fell from around 18x revenue at the 2021 peak to about 5.5x by the end of 2025.

What the Multiple Applies To: for a small, owner run product, buyers usually work from Seller's Discretionary Earnings (SDE), which is profit with the owner's salary and genuine one-off costs added back. EBITDA takes over once a management team runs the business. Revenue multiples come in when a growing product reinvests its profit, so today's earnings understate it.

Our example has an EBITDA of $66,000, and the Founder's $60,000 salary adds back, so its SDE is about $126,000. At 3x to 5x, that prices the business at roughly $378,000 to $630,000. On revenue, 1x to 3x its $360,000 ARR gives $360,000 to $1,080,000. Where it lands inside those ranges comes down to the metrics in this guide.

What Moves the Multiple:

• Retention and NRR: buyers ask for the month by month revenue bridge and retention by cohort. Our example's 80% NRR pulls it toward the bottom of the range, and fixing churn is the single biggest lever on its price.
• Customer Concentration: any one customer above 15% to 20% of revenue is a risk, because if they leave after the sale, the buyer paid for revenue that walked out the door.
• Margin: gross margins above 75% read as healthy, and profitable small SaaS products typically run at 50% or more after all costs.
• Founder Dependency: if the Founder sells, codes and handles support alone, the buyer prices in what it'll cost to replace them.
• Verifiable Numbers: marketplaces like TrustMRR verify revenue straight from the payment provider. Buyers pay for numbers they can check, so clean books and payment data count for more than a strong pitch.
• Clean Ownership: a registered company, IP assigned to it (including anything contractors built) and contracts without change of control surprises. Without those, there's nothing to buy.

Handling the Revenue Question: if a buyer asks for revenue early, while the number is still small, that number becomes the anchor for the whole conversation. Lead with the sales cycle, the product's age and its potential, then give the customer count and the average ticket, and let the buyer run the upside math. Know those numbers cold before any call.

2 Kinds of Equity: if a conversation turns to equity, ownership equity and profit-share equity are different controls. Know which one is on the table before agreeing to any percentage.

Getting acquired deserves a full guide of its own. If this section draws enough interest, that's the next one we'll write.

What We Left Out on Purpose

A full due diligence pack runs to dozens of analyses. We cut the ones that serve a consultant or an investor more than a Founder: revenue forecasting accuracy, revenue backlog, deferred revenue beyond a passing mention, discount rate impact, revenue per employee, partner contribution, sales cycle length, sales efficiency ratios and support resolution metrics. They're real measures. They just won't tell you whether the product you're paying for is working.

A Dashboard Should Be Able to Disappoint You

The test for any set of product metrics is simple: could this report ever show you bad news? If every number on it only goes up, it's a scoreboard for the team that made it.

Start small. Get CAC and payback, churn by cohort and activation, and if you sell subscriptions, NRR. Put a North Star on top. With those, you can read your own product the way a buyer would, years before one ever calls.

A Quick Glossary

• Activation: the share of new customers who reach the first moment of real value
• ARPU: average revenue per user, total revenue divided by customers
• ARR and MRR: annual and monthly recurring revenue
• CAC: customer acquisition cost, total acquisition spend divided by customers won
• CAC Payback: the months of gross profit needed to earn back CAC
• Churn: the share of customers (or revenue) lost in a period
• Cohort: customers who joined in the same period, tracked together
• Contribution Margin: what's left from each sale after the costs that grow with it
• DAU/MAU: daily over monthly active users, a measure of stickiness
• EBITDA: earnings before interest, taxes, depreciation and amortization
• GRR: gross revenue retention, revenue kept from existing customers without upgrades
• LTV: lifetime value, the gross profit from one customer over their lifetime
• LTV:CAC: lifetime value against acquisition cost, with 3 to 1 the usual benchmark
• North Star Metric: the one number that best measures the value customers get
• NRR: net revenue retention, revenue kept from existing customers including upgrades
• SDE: seller's discretionary earnings, profit with the owner's salary and one-off costs added back