What metrics should a solo SaaS track?

A small number that reflect whether the product actually works: activation (do new users reach the core outcome), retention (do they come back and keep using it), and revenue (are people paying). Those three tell you whether the product delivers value, whether that value lasts, and whether it's a business. Everything else — pageviews, signups, vanity totals — is mostly noise unless it feeds one of those. The test for any metric is simple: would this number, if it moved, change what you do? If not, don't track it.

What is a vanity metric?

A number that feels good and looks like progress but doesn't inform a decision — total signups, pageviews, followers, cumulative counts that only ever go up. They're seductive because they grow and feel like success, but they don't tell you whether the product is actually working or what to do next. The antidote is to ask whether a metric would change a decision: real metrics (activation, retention, revenue) prompt action when they move; vanity metrics just get reported.

Should solo builders rely on analytics or talking to users?

Both, but early on qualitative wins. A handful of early users don't generate statistically meaningful numbers, so quantitative analytics are thin and easy to over-read. Talking to users directly — watching where they get stuck, hearing why they stay or leave — teaches you far more at that stage than a dashboard. As the user base grows, quantitative metrics become more reliable and take on more weight. Use the few metrics that matter, but don't let a dashboard replace actually talking to the people using the product.

How I Use Analytics as a Solo Builder

"Be data-driven," everyone says, and the tools now make it trivial to track every click, funnel, and cohort you can name. So most dashboards end up measuring everything and telling you nothing. I run the opposite way: a handful of metrics that would actually change a decision — activation, retention, revenue — and I ignore the numbers that feel good and inform nothing. The test I put to any metric is one question: if this moved, would I do something different? If the honest answer is no, it doesn't get tracked. For a solo builder the scarce resource is attention, not data, and a wall of charts is an efficient way to spend all of it and learn nothing.

The metrics trap

The trap is that collecting is free and reading is not. Faced with a hundred numbers a tool will happily draw for you, it's easy to build an elaborate dashboard, watch the lines climb, and feel informed while learning nothing that changes what you'd do next. More metrics isn't more insight; past a point it's more noise piled on the few signals that matter. The skill was never collecting the numbers. It's throwing almost all of them away.

The few that matter

Three things tell me whether the product actually works:

Activation — do new users reach the core outcome? If they sign up and never get value, nothing downstream matters.
Retention — do they come back and keep using it? Value that doesn't last wasn't value.
Revenue — are people paying? The one signal that says the problem is real enough to solve for money.

Value reached, value that lasts, value paid for. Almost every other number earns its place only by feeding one of those three.

The would-it-change-a-decision test

The test for any metric: if it moved, would I do something different? A real metric prompts an action. A vanity metric just gets reported into a slide. Keep the ones that would change what you do; delete the rest from your attention.

That one question kills most of what people track. Cumulative signups only ever climbs, and prompts nothing. Pageviews feel like traffic and almost never change a decision. But activation dropping tells me onboarding broke this week and I should go find where. Retention sliding tells me the value isn't landing after all. Track the numbers that would make you act; stop watching the ones that only make you feel productive.

Qualitative beats quantitative early

Here's the part the dashboards won't tell you: early on you don't have enough users for the numbers to mean anything. A handful of people generate thin, noisy data that's easy to over-read into a trend that isn't there. At that stage talking to users — watching where they stall, hearing why they stay or leave — teaches me more in an afternoon than a month of charts. The numbers earn their weight as the base grows; until then the highest-signal analytics I have is a conversation.

Respect privacy while you're at it

One more line I hold: the measurement shouldn't cost the user their privacy or the site its speed. This site fires no analytics until you've actually consented to it — and I lose nothing by that, because the signals I care about are few and lightweight to collect. The heavy trackers buy you data you were never going to act on, at the price of a slower page and a worse deal for the visitor. Track less and respect more; the insight was only ever in the handful of metrics anyway.

What usually goes wrong

Vanity metrics. Watching numbers that feel like progress (signups, pageviews) but prompt no decision.
Tracking everything. Drowning the few real signals in a hundred irrelevant ones.
Over-reading thin data. Treating a handful of early users' numbers as statistically meaningful.
Dashboard instead of conversation. Letting analytics replace actually talking to users, especially early.

Track the three that matter — activation, retention, revenue — run the would-it-change-a-decision test on everything else, and lean on conversations while the numbers are still thin. Good analytics for a solo builder isn't a bigger dashboard; it's a smaller one, watched for the few things that make you act. Measure what changes a decision, and ignore what only changes your mood.

Part of Building AI Products. See also my kill criteria for product experiments and how I onboard users to a solo SaaS. The newsletter sends one practical build lesson every two weeks.