Most teams track too many numbers and act on none of them. A North Star Metric fixes that: one number whose improvement most reliably reflects the long-term health of the business.
It must capture value delivered to users, not vanity. Page views are vanity. For an early product, user retention is often the honest North Star. For a streaming service like Spotify, the share of active listeners who convert to paid captures both product value and monetisation in one figure. For retail, same-store sales growth plays the same role.
The real power move is decomposing the North Star into levers. Take a food delivery business. Total monthly orders can be written as the sum of existing, new, and resurrected users, multiplied by average monthly orders per user. Suddenly one number becomes four levers, and each lever can have an owner, a budget, and its own experiments.
Because it changes what you choose to design. If frequency is the weak lever, you design habit features. If resurrection is weak, you design the win-back journey. The equation tells you where your craft will compound and where it will be wasted effort.
Early-stage companies rarely have clean data. Build the model anyway with educated assumptions, then refine as evidence arrives. An imperfect model the whole team can argue with beats waiting for perfect numbers. Model the worst case too: knowing the retention floor your business can survive forces honesty about which lever needs you first. Start, build, hit an error, refine, rebuild. That loop is the actual skill, and it is one we make every Nofolios resident practise on real briefs.
