First metrics to track for a new app or SaaS

From the SoleOS answers series — written about our own product space; grounded in published definitions and documented behavior, never invented numbers.

The first metrics for a new app are the ones that answer four yes/no questions: is anyone arriving, are they signing up, are they reaching the "aha" moment, and will anyone pay. That's visitors, signups, activation, and first revenue — four numbers, tracked as a funnel, not a dashboard full of charts. Everything else is noise until you have those four in place and a real decision to make with them.

This guide is published by SoleOS, which builds portfolio dashboards for founders running multiple products — but a single new app doesn't need one yet, and this guide covers when it does.

The four numbers that matter on day one

Visitors. Not pageviews, not sessions — unique people who showed up. If this is zero or near-zero, nothing downstream matters: that's a distribution problem, not a product problem, and no onboarding polish fixes it.

Signups. Of the people who arrived, how many created an account or started a trial. This is your first real conversion event, and it tells you whether your pitch — the landing page, the store listing, the tweet that sent them there — matches what you're actually offering. A low signup rate off decent traffic usually means a positioning problem, not a traffic problem.

Activation. Of the people who signed up, how many reached the point where the product actually did something for them — imported a file, sent a first message, connected an account, whatever "got the value" looks like here. This is the metric most new founders skip, and it explains almost everything else: a product can have healthy signups and still fail if nobody activates.

First revenue. Of the people who activated, how many converted to a paying customer, or, pre-monetization, how many said yes when you asked for a price. This is the only number that proves willingness to pay — worth tracking from user one, even as a manual log of who said yes to what.

If you can name today's value for each of these four without opening a tool, you're in good shape. If you can't, that's the gap to close before adding anything else.

Why this is a funnel, not four separate charts

The four numbers above aren't independent — they're stages, and the ratio between adjacent stages is more informative than any single count. Visitors-to-signups tells you about your pitch. Signups-to-activation tells you about your onboarding. Activation-to-paid tells you about your pricing and value delivery. Looking at any one number alone invites the wrong conclusion: "traffic is up" means nothing if the signup rate cratered at the same time, and "signups are up" means nothing if none of them activate.

Practically, this means every number needs a denominator. Not "43 signups" but "43 from 600 visitors, a 7% rate." Not "12 activated" but "12 of 43 signups, 28%." The rate is the metric; the raw count is just an input. This funnel logic is covered in more depth in how to track analytics as a solo founder, including wiring these stages up in whatever tool you already have.

Vanity metrics feel good and change nothing

Total registered users, cumulative pageviews, and social follower counts share a property: they only go up, they feel like progress, and they rarely change what you do tomorrow. A total-users counter that reads 4,000 doesn't tell you if the app is dying — most of those people could have signed up once and never come back. Pageviews reward content nobody acts on. Follower counts move independently of whether anyone downloads, signs up, or pays.

The tell is simple: vanity metrics are cumulative totals with no denominator and no time window attached. Real metrics are rates, over a defined period, tied to a funnel stage — this week's visitor-to-signup rate, this month's activation rate. If a number can only go up and never forces a "we need to fix this" conversation, it's decoration, not a metric.

The rule: if it wouldn't change a decision, don't track it

Before adding any metric, ask what you'd actually do differently if the number moved. If a swing in either direction wouldn't change what you build next week, don't bother instrumenting it yet. This single filter eliminates most of what analytics tools tempt you to track on day one — session duration, scroll depth, feature-by-feature breakdowns, device splits, referrer detail. All of that becomes useful later, once you have enough volume that the coarse four-number view stops being informative. At the start it's just surface area to maintain and a reason to feel busy without deciding anything.

This is also the test for a fifth metric. If your activation rate raises an obvious next question — "from which acquisition channel?" — that's decision-driving, add the breakdown. If you're adding a metric because a blog post said to, it fails the test.

A spreadsheet is enough at first

You do not need an analytics platform, a BI tool, or a portfolio dashboard to track four numbers. A spreadsheet with one row per week and four columns — visitors, signups, activated, paid — covers this stage completely, and it forces you to look at the numbers by hand rather than glance at a chart and move on. If you're already using an analytics tool for something else, its default dashboard is almost certainly sufficient too; you're looking up four counts, not building custom reports.

That changes once you're running more than one product and need to compare stages across products without checking several tools each week. That's a portfolio problem, and it's the specific gap SoleOS is built for — but it's a second-app problem, not a day-one one. A single new pre-revenue app should reach for a spreadsheet, not a portfolio dashboard; browse SoleOS's guide library for a broader reference as it grows.

Instrument only the activation path

The only work worth doing up front is instrumenting the two or three events that sit on the path to activation — the specific actions a user takes between signing up and getting value. If your "aha" is uploading a file and seeing a result, that's two events: file uploaded, result viewed. Connecting an account and seeing a first synced data point is the same shape. Resist instrumenting every button and page at launch — you don't know yet which will matter, and most of what you tag now gets deleted in three months anyway.

Visitors and signups are usually already captured by your hosting platform, auth provider, or default analytics install, no extra work required. First revenue is captured by your payment processor. The only genuinely new instrumentation is those 2-3 activation events, and they're worth naming precisely — "activated" as one event tied to the exact action that constitutes value beats a pile of granular feature-usage events you'll never look at individually. SoleOS's metrics reference has definitions for these and the later-stage metrics below, if you want a common vocabulary to work from.

How the metric set grows up

The four-number funnel fits the first weeks after launch, when the question is "does this work at all." Once you have a few months of paying customers, the questions change, and so should the metrics:

• Churn becomes trackable once customers have had a chance to cancel — logo and revenue churn tell you if what you built holds onto people.
• Cohort retention replaces single-point activation once you can watch a signup cohort over multiple weeks — are week-1 activators still active in week 4, week 8.
• LTV becomes meaningful once churn is stable enough to model, telling you how much you can spend to acquire a customer.

None of these are day-one metrics. Adding them before you have the volume or history to make them meaningful just adds noise you can't act on — the "if it wouldn't change a decision" rule applies here too. Get the funnel working first. For turning early activation and revenue signals into paying customers, see how to get to your first $100 in MRR; if you're still validating whether to build at all, how to validate a SaaS idea covers the stage before any of this.

Frequently asked questions

What's the very first metric to set up before anything else?

Visitors. If nobody is arriving, every other number in the funnel reads zero or near-zero, and you'll waste time debugging onboarding or pricing when the actual problem is that no one has seen the product yet.

How do I define "activation" for my specific app?

Activation is the single action that separates someone who tried your product from someone who got value from it — not "logged in," but the moment it did the job it promised. Write that moment down in one sentence before building any instrumentation.

Should I track revenue before I have a paywall?

Yes, informally. If you're pre-monetization, log every time you ask someone to pay and what they said, even as a manual note. That's your first revenue signal, and it's more honest than any proxy you could substitute for it.

Do I need a dashboard tool for four metrics?

No. A spreadsheet or the default dashboard in whatever analytics tool you already use is enough. Reach for a dedicated dashboard once you have more than one product to track side by side.

When should I add churn and retention?

Once you have paying customers with enough history that cancellations and repeat usage are actually happening — typically a few months in, not week one. Adding them earlier just gives you noisy numbers with no pattern to read.