📄 Article
Free
Intermediate
Why we picked it This is the piece that made cohort-based LTV the default way serious investors think, and it is exactly the right medicine when you have three months of data and no settled curve. Hsu argues you should read empirically realized cohort LTV instead of plugging numbers into a formula, and he shows why young cohorts should stay short lines (you genuinely do not know their LTV yet) rather than being extrapolated into a confident-looking figure. Treat it as a way to bound and shape your estimate, not to manufacture one.
Diligence at Social Capital, Part 3: Cohorts and (revenue) LTV
From Social Capital / The Startup (Medium) by Jonathan Hsu About 20 minute read
- Prefer empirically realized cohort LTV over a formula: with thin history, a single LTV number is a guess dressed up as math.
- Young cohorts are short lines on purpose, do not extrapolate them; watch the shape (flat, sub-linear, linear, super-linear) instead of chasing one figure.
- What you are really looking for early is evidence of linear or super-linear LTV in at least some cohorts, which tells you retention is holding, not just decaying.