Latest writing
Account for the behavior your measurement creates
A practical measurement footprint review for separating product behavior from behavior created by the act of studying it.
Read essayRecent notes
Recalibrate activation before the signal drifts
A practical activation calibration card for checking whether an early growth signal still predicts the outcome the team thinks it does.
Remove the push before you call the behavior durable
A growth PM field note on withdrawal tests, durable behavior, and learning what remains after a prompt, credit, or streak protection goes away.
Explore the work
Keep a constraint ledger beside the growth hypothesis
A practical entry point into hypothesis writing, structured thinking, and better growth conversations.
In defense of side quests
Notes from the middle on side quests, growth judgment, and the kind of work worth keeping.
More writing
Make a segment earn the right to change the product
A growth PM field note on separating meaningful user differences from noisy slices with a practical segment warrant and learning plan.
Estimate the rebound before you credit the rescue
A natural-recovery baseline for separating regression toward the mean from the incremental effect of a product rescue.
When the labels move, the trend may move with them
A growth PM field note on separating real behavior change from changes caused by renamed choices, with a practical continuity bridge and split-sample test.
Look on both sides of the cutoff
A growth Product field note on policy cliffs, near-cutoff comparisons, and deciding when a hard threshold is worth keeping.
Make protective friction earn its place
A growth Product field note on protective friction, harm budgets, and testing when a slower path produces a healthier outcome.
Choose the unit before you optimize the flow
A growth PM field note on finding the unit of work users actually carry, with a practical unit boundary sketch and a test for choosing where completion belongs.
Design the reconciliation before the records disagree
A growth Product field note on reconciliation briefs, shared truth, and testing how teams recover when records diverge.
A saved choice is not a permanent intention
A growth Product field note on intent freshness, revalidation, and knowing when a past choice is safe to reuse.
The same steps in a different order are a different product
A growth Product field note on path dependence, sequence audits, and testing whether the order of a journey changes its outcome.
Hold the journey still long enough to learn
A growth PM field note on creating a learning window around consequential experiments without turning a fast team into a frozen one.
Keep the assignment in the experiment
A growth PM field note on preserving random assignment, diagnosing treatment reach, and separating the effect of a product policy from the effect among users who received it.
Make the proof no broader than the claim
A practical field note on reference classes, proof distance, and making customer evidence useful without sanding away the conditions that produced it.
Measure what the new path displaced
A growth PM field note on substitution, incrementality, and tracing where a new path gets its apparent wins.
Do not mistake a stack of tests for a portfolio
A growth PM field note on finding shared assumptions across experiments and building a portfolio that can teach you more than one lesson.
A green check does not grant authority
A growth PM field note on the gap between product permissions and real decision rights, with an authority gap map for finding false completion.
Follow the work past the first output
A practical last-mile trace for testing whether a product output reaches the real situation where it can create value.
Test the dose, not only the feature
A growth PM field note on dose response, product intensity, and testing how much of a useful intervention users can actually absorb.
Measure whether demand had a chance to appear
A growth PM field note on imperfect detection, hidden product demand, and a practical demand visibility audit for making better roadmap and experiment decisions.