Lean Analytics
by Alistair Croll & Benjamin Yoskovitz
The rare metrics book that tells you which number to ignore — and when to move the goalposts.
- Status
- Read · February 2018
- Bought
- February 22, 2018
- For
- Founders and early product leads · Engineers who own a metric they didn't choose · Anyone drowning in a dashboard nobody reads
The one-paragraph verdict
The core argument is sharp: at any moment there is exactly one metric you should obsess over, and it changes as you move through stages (empathy, stickiness, virality, revenue, scale). The discipline of picking one number and a line you must cross before you’re allowed to care about the next one is genuinely useful, and the 30-odd case studies give you real benchmarks instead of hand-waving. The recurring criticism is fair, though — it’s padded, the same ideas circle back two or three times, and some of the SaaS/e-commerce benchmarks already read as dated. Read it for the framework, not the page count.
Who should read it
Founders, early PMs, and engineers who suddenly own a “north star” they didn’t pick should read it cover to cover once. If you already run experiments against a single funnel metric and know your stage, you’ll find most of it familiar — borrow the chapter on your current stage and skip the rest.
Where it earned its place
The OMTM idea is exactly how I think about measurement: one load-bearing number per phase, with an explicit threshold before the next one matters — not a dashboard of forty equally-loud charts that no one acts on.
Skip it if…
You want statistical rigor or causal-inference depth — this is a product-judgment book, not a stats book. Skip it too if you’ve already internalized lean-startup mechanics; the marginal idea-per-page here is low.