The Metrics That Matter Before Product-Market Fit

Half the metrics in your next pitch deck don’t mean anything. Not because they’re calculated wrong — because they measure the wrong thing for the stage you’re actually in. Before you’ve found product-market fit, your job isn’t to grow. It’s to find out whether there’s anything worth growing. And that gets measured differently.

What you’ll learn

• Why total users and downloads lie to you at the early stage
• How to pick your North Star based on your business model, not someone else’s
• Why cohort retention is the one judge that can’t be bribed
• How to calculate CAC, LTV, and contribution margin without opening a forty-tab spreadsheet
• Why runway is the metric that decides whether the others matter
• How to see it all in one place without hiring a data analyst

Vanity vs. truth: the table that saves you a bad board meeting

A vanity metric goes up no matter what. A truth metric only goes up when the business actually gets better. Confusing the two is why sharp founders spend six months “growing” toward nowhere.

Vanity metric	Why it misleads	Truth metric that replaces it
Total signups	Rises with marketing spend, not with product value	Active users at week 4 (retention)
App downloads	Correlates with ad budget, not with value delivered	% completing the key action (activation)
Website visits	Measures traffic, not intent	Visit-to-trial conversion rate
Social followers	Doesn’t pay the bills	Customers who renew without being asked
Cumulative all-time revenue	Ignores whether the customer churns next month	Net recurring revenue after churn (net MRR)

The left column is what goes in a deck because it always trends up and to the right. The right column is what tells you whether you’ll still have a company in eighteen months.

Your North Star isn’t your neighbor’s North Star

North Star is the one metric that, when it rises, means you’re delivering more real value. You don’t pick it because it’s trendy. You derive it from your model:

• Two-sided marketplace: completed transactions, not signups counted separately on each side.
• Recurring-use SaaS: key actions completed per active user (not logins — logins mean nothing if users aren’t doing anything inside).
• Consumer product with per-unit margin: units sold at positive margin, not gross revenue.
• B2B tool replacing manual work: hours or tasks the customer no longer has to do by hand because of you.

The litmus test: if your North Star goes up one month and your business is objectively no better off, you picked the wrong one. Change it without worrying about “losing historical continuity.” The historical series of the wrong metric isn’t worth anything.

Retention: the only real judge of product-market fit

Here’s the metric most founders avoid looking at directly, because it hurts.

Take a group of users who started using your product the same week (a cohort). Track them each following week: how many are still active? Plot the curve.

There are two possible shapes:

1. The curve drops and then flattens at some point above zero. That flattening is the most reliable signal of product-market fit there is. It means there’s a core of people for whom your product is indispensable, not a novelty.
2. The curve drops toward zero without leveling off. It doesn’t matter how many new users you pour in at the top — you’re filling a bucket with a hole in it.

Most founders watch today’s active-user count instead of the shape of the cohort curve. It’s the most expensive mistake at the pre-PMF stage, because a strong acquisition funnel can mask broken retention for months — until CAC rises and there’s nothing left to hide behind.

Unit economics without a 40-tab spreadsheet

Three numbers, in order of importance at the early stage:

1. Contribution margin per customer — the price the customer pays minus the direct variable cost of serving them (this excludes fixed costs and salaries; it’s only the cost that scales with each new customer).

Example (illustrative figures): you charge $49/month for your SaaS. Server, support, and payment-processing cost per customer runs $8/month. Contribution margin: $41/month per customer. If that number is negative, you don’t have a growth problem — you have a product or pricing problem, and growing just makes it bigger.

2. LTV (customer lifetime value) — monthly contribution margin multiplied by the average time a customer sticks around before churning.

Example: $41/month margin, average customer stays 14 months (based on your actual retention curve, not a guess) → LTV ≈ $574.

3. CAC (customer acquisition cost) — everything you spent on marketing and sales divided by the number of new customers acquired in that period.

Example: you spent $2,000 on ads and landed 20 customers → CAC = $100.

The LTV/CAC ratio commonly cited as healthy (3

or better) is useful, but with fewer than fifty customers it’s more noise than signal — one large account or one unusual campaign skews it completely. What you can trust with a small sample is the per-customer contribution margin: if you’re losing money customer by customer, you don’t need a bigger sample to know something’s off.

Runway and burn: the survival metric

None of the above matters if you run out of cash before you get an answer.

Runway (available cash divided by net monthly burn) isn’t just one more item on the list — it’s the one that puts a clock on all the others. It defines how many retention and pricing experiments you can run before the conversation stops being “do we have PMF?” and becomes “are we shutting down?”

Two simple disciplines almost no one keeps rigorously:

• Review net burn every month, not every quarter. A quarter is plenty of time for a small problem to become a structural one before you see it coming.
• Separate launch burn from recurring burn. A one-time expense (a tool, an event) shouldn’t inflate your runway projection the same way a cost that repeats every month does.

How to track this without a data team

None of this requires a data team or a thirty-widget dashboard. It requires that revenue, expenses, and cash live in one place and update themselves without you having to ask anyone.

At Frihet, that’s exactly what the real-time financial dashboard does: invoices issued, expenses captured via OCR, and cash balance all reflect instantly, with no manual month-end close needed to know where you stand today. It works the same whether you’re a solo founder invoicing clients or a twenty-person team — management shouldn’t have a size ceiling.

And if you already work with an AI agent for your business workflows, Frihet’s MCP server gives it direct access today to that same data — invoices, expenses, customers — so you can ask it for revenue or MRR cohorts by customer signup date, or a margin summary, without exporting anything by hand. Usage retention is a different animal: seeing that curve requires your own product-activity data, which isn’t something Frihet tracks. That’s what you can already do today. The direction this is heading — an agent that doesn’t just report on your business but acts on it autonomously — is still being built, not something we’re promising as a feature today.

Common mistakes measuring pre-PMF

• Chasing a borrowed North Star. Copying the metric some other founder tweets about without asking whether it fits your model.
• Watching the total instead of the cohort. A number climbing at the top can hide a leak at the bottom.
• Optimizing CAC too early. With few customers, it’s noise; it pushes you to make decisions on a sample too small to support the weight.
• Confusing revenue with margin. Billing a lot while losing money per customer is worse than billing less with healthy margin — the first one scales the loss.
• Checking runway only when it’s scary. By then you’ve already lost options you had two months earlier.

After PMF, the game changes

The day your retention curve consistently flattens, the weight shifts. Retention stops being the verdict and becomes maintenance. Growth-efficiency metrics enter the picture — CAC payback, account expansion, LTV/CAC with a sample you can finally trust — because now it’s worth investing in scaling something you know works.

But that’s a different article, with different priorities. Confusing the two stages is like flooring the accelerator before you know the engine starts.

The early-stage discipline is simple to state and hard to sustain: measure little, measure what matters, and don’t lie to yourself about the number that’s easy to make go up.