How many KPIs should an early-stage startup track

Searching for the right KPI is normal, and it doesn't stop once you're experienced. Shaul Olmert has started several companies, and the same struggle repeats each time, as he told Etgar Shpivak on Founders' Marketing Compass: "I'm still struggling to find the right KPI, spending a lot of time measuring a lot of different things that don't really lead anywhere." Knowing what to measure is a sign the company has matured.

Etgar Shpivak's answer is to stop trying to pick a KPI and start searching for it. The one that matters is the single number that narrates the business, which Shaul Olmert framed as choosing "the one KPI that really tells the story of your business." You get there by trying things, watching what the results reward, and cutting what leads nowhere. It's exploration before it's a decision.

The right KPI changes as the product and the market change, so the one that fits today rarely lasts. Shaul Olmert optimized PlayBuzz around how often its content got shared, then picked a different number for KwaKwa: whether creators return to build a second course. He is direct that the current metric only fits the stage he is at now. Treat any KPI as right for now, and plan to replace it.

Before product-market fit, the point where a product finally sticks with real users, the honest measure is whether anyone comes back on their own. Etgar Shpivak, who works with seed and Series A founders, would watch one signal at that stage: do the first users return without being prompted. A startup with no regular use has no real KPI yet, and naming one only rewards activity. Wait for repeat behavior before you crown a number.

For an early team, the trait that matters is range, or a real pull toward your problem. Shaul Olmert would rather hire someone who has done a dating site, an employment agency, and a B2B marketplace than five dating sites in a row, because the range shows they can learn a new business. Early on, he says, you need someone to help invent the story. See what a first marketing hire should own.

Marketing, in Shaul Olmert's account, is the work of turning what the company does into a story people grasp, and it has to be owned across the company instead of handed to one hire. He defines it as reducing a complex reality into digestible portions. That is also why a blunt one-sentence value proposition tends to beat a careful paragraph.

Share rate, the share of people who saw a piece of PlayBuzz content and passed it on. Once people started discovering content through Facebook, the team optimized for that number and little else, and PlayBuzz became the most-shared website on Facebook. It took years to find, after a long stretch of measuring things that led nowhere.

Small teams tend to chase the number easiest to move, like a raw signup count, over the one hardest to fake. On a team of eight, nobody catches the swap, and everyone works hard to move a metric that doesn't really prove value. The pick should be the most honest number, not the most convenient one.

A founder with one number arrives at the review holding a question instead of a crowded spreadsheet. The meeting stops being a tour of a dozen charts and turns into an argument about the one thing the company is trying to move. The other numbers still get watched by someone; they just stop setting the mood in the room.

The other metrics don't disappear, someone still watches them, but they stop setting the direction. Once there's one number the company lives on, the rest usually turn out to be habit: metrics added every time the board asked a hard question, and never removed. They become background information, not what drives decisions.

Choosing the right KPI is mostly a founder's job, but an outside operator helps you see which single number your stage rewards. Etgar Shpivak, a marketing consultant who works with seed and Series A founders, does this directly with the founder, naming the one metric the business should live on and cutting the rest. Etgar Shpivak works at exactly the point where that number is still being found. You can see how he works on his bio page.

First name the single number that, if it moved, would prove you're creating value at the size you are today. Then translate it into each person's weekly work, so every task ties back to that number. Without that link, the KPI stays a board slide instead of something that steers the day-to-day.