Photo by Jordan Madrid on Unsplash
The North Star metric framework, popularised by Sean Ellis, is a powerful focus mechanism. Ellis's original criteria are still the cleanest: a NSM should be actionable, not a ratio, capable of scaling over time, and correlated with revenue without being revenue itself (Ellis on GrowthHackers, 2019; Lenny's interview with Ellis, 2024). The trap is treating the metric as static.
It is not. Lenny Rachitsky's survey of 40+ growth-stage companies found that roughly 25% had changed their North Star recently or planned to (Lenny's Newsletter, 2021). The case list reads like a who-is-who of product-led companies: Dropbox shifted from engagement to paid growth; Figma and Uber moved from revenue to market share; Spotify pivoted from engagement to consumption after podcasts changed the product surface. None of these reset because the company failed; they reset because the strategic ground moved.
Brian Balfour and Casey Winters' Reforge essay argues the harder point: blindly buying in to a single metric can kill you. Their case study is Pinterest's WARC (Weekly Active Repinner or Clicker) metric, which optimised for engagement at the cost of the broader ecosystem (Reforge, attributed Balfour). Amplitude's product-leadership guidance reinforces it: a NSM should be paired with three to five complementary inputs and reviewed every 6 to 12 months for early-stage companies (Amplitude, 2024).
A heuristic that survives: pick the single number that, if it moved 30% next quarter, would unambiguously mean the business is healthier. Wire 3 to 5 input metrics underneath to keep the NSM honest and make the cadence quarterly rather than annual. When the input drifts faster than the NSM, that is the early signal that the metric is becoming a lagging indicator. Move it before it stops paying.