Search for SaaS metrics and you will find dashboards full of CAC payback periods, net dollar retention, magic numbers, and LTV:CAC ratios. Almost all of it is written for companies with hundreds of customers and a sales team, and applying it to a business with forty users produces numbers that are technically calculable and completely meaningless.
The problem with premature metrics is not that they waste time. It is that they create false confidence. A lifetime value figure derived from three months of data and eleven customers will be wildly wrong, and decisions made on it will be wrong in the same direction.
Before roughly $10k MRR, you are not optimising a machine. You are trying to establish whether the machine works at all. That calls for a much smaller set of numbers.
1. Monthly Recurring Revenue, Split by Movement
MRR is the obvious one, but the total is the least interesting part. What matters is the movement underneath it.
Break the change each month into four components. New revenue from customers who did not exist last month. Expansion from existing customers paying more. Contraction from existing customers paying less. Churned revenue from customers who left entirely.
Two businesses can both show $6,000 MRR and be in completely different health. One added $1,200 of new revenue and lost $200. The other added $2,000 and lost $1,000. The second is working much harder to stand still, and the blended total conceals that entirely.
This split is also the earliest reliable warning system you have. Churned revenue creeping up while new revenue holds steady is the signature of a retention problem arriving before it shows in the headline number.
2. Activation Rate
Activation is the percentage of new signups who reach the moment where your product delivers its core value — importing their data, sending their first campaign, inviting a teammate, whatever the equivalent is for you.
This is the most under-tracked early metric and frequently the most valuable, because it sits upstream of everything else. Trial conversion, retention, and word of mouth are all downstream of whether people ever experienced the thing you built.
It also isolates the problem. If activation is thirty percent and conversion among activated users is healthy, you do not have a pricing problem or a product problem — you have an onboarding problem, and that is comparatively cheap to fix. Without this number, that same situation looks like a general "conversion is bad" fog.
Defining your activation event precisely is a prerequisite, and it takes real thought. It is worth the afternoon.
3. Cohort Retention
A single churn percentage averages across customers who joined at wildly different times under wildly different versions of your product. It tells you very little.
Group customers by the month they joined and track what fraction remain active in each subsequent month. Read it as a grid: each row a cohort, each column a month of age.
The question this answers is the one that matters most in the early days: is the product getting better? If your June cohort retains better at month three than your March cohort did at month three, the changes you shipped in between are working. If every cohort decays identically, you are shipping without improving retention, no matter how busy the changelog looks.
With small numbers the percentages will be noisy, so read the shape rather than the decimals. A curve that flattens — where retention stops declining and holds — indicates a genuine core of people for whom the product has become part of how they work. That flattening is the earliest credible evidence of product-market fit, and it is visible long before revenue makes it obvious.
4. Time to First Value
How long does it take, in minutes or days, between signup and the activation moment?
This is a diagnostic partner to activation rate. If activation is low, time to first value usually explains why. Long paths lose people at every step, and each additional required action compounds the loss.
Measure it as a median rather than an average — a handful of users who activated three weeks later will distort the mean badly. Then attack it directly: which step in the path takes longest, and can it be removed, deferred, pre-filled, or replaced with a template?
Reducing this number tends to move activation, which moves conversion and retention. It is one of the few metrics where improvement propagates through the whole funnel.
5. Qualified Signup Volume
You need to know whether enough of the right people are arriving, and raw signup counts do not tell you that — they mix genuine prospects with curious browsers and bots.
Define what qualified means for your product. Often it is as simple as a work email address, or a self-reported role that matches your target, or completion of the first meaningful step. Then track that number and where it comes from.
Two things make this useful. First, it is the only early metric that tells you whether your marketing is reaching the right people rather than merely reaching people. Second, tracking it by source lets you see which channel produces customers rather than traffic — a distinction that frequently reverses your assumptions about which channel is working.
6. Conversations Per Week
This is not a conventional metric and it is arguably the most important one before $10k MRR.
How many real conversations did you have with users or prospects this week? Not support tickets closed — actual conversations where you learned something you did not previously know.
Every early-stage company that finds its footing does so through an accumulation of these. The founders who plateau are almost always the ones who stopped talking to people and started reading dashboards, because dashboards tell you what is happening and only conversations tell you why.
Five a week is a good target. Track it like a metric because otherwise it silently drops to zero during busy periods, which are exactly the periods when you most need the signal.
What to Ignore for Now
Several widely-cited metrics are actively misleading at this stage.
Lifetime value requires a stable churn rate over a meaningful period. With a young product and a small sample, any LTV figure is a fiction built on a guess, and it will be too optimistic.
Customer acquisition cost, when most of your customers arrive through founder effort, communities, and word of mouth, mostly measures how you allocated your own unpaid time. It becomes meaningful once you spend real money on acquisition.
Net dollar retention needs enough customers and enough time for expansion and contraction patterns to be real rather than noise.
Vanity totals — page views, total signups ever, social followers — feel good and correlate with nothing.
None of these are bad metrics. They are simply premature, and premature metrics do not just waste attention; they manufacture confidence in numbers that cannot support it.
Keep It on One Page
Six numbers fit on a single page, and a single page is something you will actually look at. Review it weekly, at a fixed time, and write one sentence about what changed and why you think it changed.
That last part is where the value compounds. The sentence forces you to form a hypothesis, and next week you find out whether you were right. Over a few months that habit produces something no dashboard provides: an accurate mental model of how your particular business actually behaves.
Get the six right, keep talking to people, and add the sophisticated metrics when you have enough data to make them honest.