The One Metric I Watch as a Solo Founder

There's a genre of startup advice that says pick your north star metric and optimize everything toward it. It's not bad advice, but for a solo founder it can turn into a dashboard you stare at instead of a product you build. I went through a phase of tracking everything (signups, activation, churn, MRR, visits, trial conversion) and the honest result was that more numbers made me less clear, not more.

So I stripped it down. These days there's one metric I actually watch, and it's not the one I expected.

The metrics that lied to me

First, what I stopped trusting.

Signups looked great and meant little. A good week of traffic produced a spike of signups, and most of them poked around once and never came back. Optimizing for signups optimized for curiosity, not value, and curiosity doesn't pay. I'd feel good on a high-signup week and the good feeling was unearned.

MRR is the real scoreboard, but it's a lagging, lumpy one for a small company. A single customer landing or leaving swings it enough that week-to-week it's mostly noise. It tells you where you've been, not whether what you did this week worked. I couldn't steer by it, because by the time it moved, the cause was a month old.

Visits and rankings are the vanity end. Useful context, but they measure attention, and attention is upstream of everything and proof of nothing. Plenty of attention converts to nothing if the product doesn't deliver the moment someone arrives.

The one I actually watch

The metric I settled on is the number of people who asked their database a real question this week and came back to ask another one.

It's clunky to say and that's sort of the point — it's not a clean dashboard number, it's a behavior. Did someone connect a real database, ask a question that mattered to them, get a useful answer, and then come back later with another question. That return, that second question, is the whole product working. The first question can be curiosity. The second question is value, because nobody comes back to a tool that wasted their time the first round.

Everything I care about is downstream of that behavior. If people keep coming back to ask more questions, they activated, they're getting value, they'll probably pay, and they might tell someone. If they ask once and vanish, no amount of signups or traffic or ranking will save me, because the thing isn't delivering. The returning question is the earliest honest signal that the product does what I claim.

Why this one and not retention or activation

You might say that's just retention, or activation, with extra words. Close, but the specificity matters to me, and the specificity is why it works as a steering metric instead of a reporting one.

Generic "retention" would count someone who logged in and left. Generic "activation" might count someone who connected a database and stopped. The metric I watch requires the actual valuable action, a real question and then another, which filters out the motions that look like progress but aren't. It's narrow on purpose. A narrow metric you can act on beats a broad one you can only report.

It's also a metric I can do something about this week. When it dips, the question is concrete and answerable: why did people who asked one question not ask a second? Usually it's a specific, fixable thing — the answer to their first question was wrong, or slow, or the setup to ask it was too painful and they bailed. That's a metric that tells me what to build next, which is the only kind worth a solo founder's attention.

The discipline underneath it

The real lesson wasn't which metric. It was that as a solo founder, the cost of watching the wrong metrics isn't just wasted time. It's misdirected work. Every metric you track pulls your effort toward moving it, and if you track signups, you'll build for signups, and you'll get a spike of people who never come back and a good feeling that lies to you.

So I'd rather watch one number that, when it goes up, means the product genuinely worked for a real person, and when it goes down, points me at exactly what to fix. For me that's people coming back to ask their database another question. For your product it'll be a different behavior, but the test is the same: find the single action that, if people keep doing it, means you've won, and if they stop, means nothing else matters. Watch that. Ignore most of the rest. The dashboard with twelve numbers was making me feel informed while keeping me from seeing the one that counted.

If you're curious what people keep coming back to ask, that's the thing I'm building at QueryBear. But the metric discipline is free, and it's the most useful thing I've changed about how I work alone.