What metrics should an early-stage startup actually track?

Early-stage startups should track the fewest metrics that prove their core hypothesis is true—primarily growth rate, user retention, and direct engagement quality. Tracking too many numbers too early creates a false sense of progress and obscures the one signal that matters most: are real people getting genuine value from what you built? Focus on metrics that change your decisions, not ones that comfort you.

Growth Rate Is the North Star Metric

For an early-stage startup, week-over-week growth rate in active users or revenue is the single most revealing number. It compounds. A startup growing 7% week-over-week doubles roughly every 10 weeks; one growing 1% barely moves. The absolute number doesn't matter much in the first months—what matters is whether that number is moving in the right direction consistently. This is why growth rate, not total user count, is what serious investors look at when evaluating early traction.

Paul Graham's argument that rapid growth is what defines a startup—not simply having received outside funding or working on technology—has a direct implication for which metrics belong on your dashboard. If growth is the defining characteristic, then the metric you track most obsessively should directly measure growth. Everything else is a diagnostic tool to explain why that number is or isn't moving.

One practical trap: founders often substitute 'activity' metrics for genuine growth metrics. Signups are not the same as active users. Downloads are not the same as engaged users. Page views are not the same as retained users. Define 'active' tightly—what does a user have to do to count as deriving real value from your product this week?—and track that number weekly without exception.

Retention Reveals Whether Your Product Actually Works

Growth metrics tell you whether people are coming in. Retention tells you whether they stay. An early-stage startup with strong retention and slow growth has a distribution problem—solvable. An early-stage startup with weak retention and fast growth has a product problem—far harder to fix, and ultimately fatal once the paid acquisition tap runs dry.

The most useful retention view at the early stage is a cohort retention curve: take everyone who started using your product in a given week, and track what percentage is still active 1 week, 4 weeks, and 8 weeks later. The shape of that curve tells you more than any single number. A curve that flattens out—even at a relatively low percentage—indicates a segment of users who genuinely need your product. A curve that keeps declining toward zero means you haven't found product-market fit yet, regardless of what your growth rate looks like.

For consumer products, strong retention often looks like 20-30%+ still active after 8 weeks. For SaaS or B2B tools used daily at work, expectations are higher. But rather than benchmarking against industry averages, compare your retention curve against your own previous cohorts. Improving retention cohort-over-cohort is the directional signal that counts most.

Qualitative Engagement Is a Metric Too

Paul Graham's observation about Airbnb—that roughly 30 days of direct, in-person user engagement made the difference between the company surviving or dying—points to something most metric dashboards ignore entirely: the quality of individual user interactions is data. At an early stage, talking to your users weekly and tracking what you learn is as important as any quantitative dashboard.

The specific things worth tracking qualitatively: How did this user find you? What made them try the product? What almost made them quit? What would they tell a friend the product is 'for'? Aggregating answers to these questions over 20-30 users reveals positioning gaps, feature priorities, and the exact language your early adopters use to describe the problem—which is invaluable for both marketing copy and investor pitches.

One lightweight system: keep a simple document where you log one insight per user conversation, tagged by theme. After 15-20 conversations, patterns become obvious. This is not a substitute for quantitative metrics, but it is the fastest way to understand *why* your quantitative metrics look the way they do—and what to change.

Revenue and Burn: The Two Numbers That Determine Your Timeline

Even if you're pre-revenue, you should be tracking monthly burn rate and calculating your runway weekly. Paul Graham's point that spending a lot of money is dangerous for early-stage startups—not just because it depletes resources, but because it makes the company rigid and harder to pivot—argues for extreme spending discipline before product-market fit is confirmed. Your burn rate is a direct measure of how long you have to find the thing that works.

If you have any revenue, track monthly recurring revenue (MRR) and the week-over-week change in MRR as separate figures. The absolute MRR is less important than whether it's growing and at what rate. Also track MRR expansion versus new MRR: if existing customers are spending more over time, that's a signal of real value delivery. If almost all your MRR growth comes from constantly acquiring new customers, you need to investigate whether retention is the culprit.

At the very early stage, also track average revenue per user (ARPU) and whether it's going up or down. Founders often inadvertently train users to expect low prices by discounting heavily to get the first customers. Knowing your ARPU and tracking it gives you visibility into that dynamic before it becomes a structural pricing problem.

What Not to Track (Yet)

The temptation to build elaborate dashboards with 20+ metrics is real, especially when analytics tools make it easy. Resist it. Every metric you track has an implicit cost: it takes attention, it invites rationalization ('our DAU is down but our MAU looks fine'), and it can create the illusion of insight without producing action. At the early stage, a metric only belongs on your weekly review if a change in that number would cause you to do something different immediately.

Vanity metrics to deprioritize early: total registered users (not the same as active), social media followers, press mentions, and app store ratings with small sample sizes. These feel good and look good in pitch decks, but they don't tell you whether you're building something people need. Investors who understand early-stage companies will not be impressed by these numbers if your core retention or growth metrics are weak.

Perhaps most importantly: don't track competitor metrics obsessively. The energy founders spend monitoring competitor feature releases, pricing changes, and funding announcements is almost always better spent on their own users. As Graham notes in the context of YC evaluations, competitors rarely kill startups—poor execution does. The metrics that tell you about your own execution quality are the ones worth your limited attention.

“Rapid growth is what makes a company a startup.”

— Paul Graham, source

The one thing to do

This week, define exactly what 'active user' means for your product, calculate your week-over-week growth rate for the last four weeks, and plot a simple cohort retention curve—those three numbers will tell you more than any other dashboard you could build.

Frequently asked questions

How many metrics should an early-stage startup track?

Three to five at most: weekly growth rate, cohort retention, burn rate, and (if applicable) MRR. Add a metric only when you're certain a change in that number would directly cause a decision. More metrics than that creates noise and dilutes focus.

When should a startup start tracking revenue metrics?

As soon as anyone is paying you anything, even a small amount. Revenue is the cleanest signal that someone values your product enough to exchange money for it—far cleaner than signups or engagement. Start tracking MRR the moment you have your first paying customer.

Is it okay if my early numbers are very small?

Yes—what matters is the direction and rate of change, not the absolute size. A startup with 50 users growing 10% week-over-week is in a better position than one with 5,000 users growing 0%. Small numbers are expected at the seed stage; flat or declining numbers are the warning sign.

How often should early-stage founders review their metrics?

Growth rate and retention should be reviewed weekly, burn and runway monthly. Daily checking of metrics often leads to overreaction to noise; weekly review forces you to look for real trends and gives you enough distance to interpret what you're seeing.

Sources

More playbook answers · Growth Prophet home