Writing
Essays on growth, retention, expansion, and the things operators learn the hard way. Updated when something is worth saying.
OCTOBER 2026
What a restaurant kitchen's one-word protocol teaches about overlapping launches, contaminated experiments, and why heavier process is not the fix.
Read essay →OCTOBER 2026
Borrowing the method from a blind AI model test: why our experiment readouts, creative reviews and channel calls are graded with the names on, and four cheap ways to take them off.
Read essay →SEPTEMBER 2026
How a run of small, unglamorous UX fixes grew GTV without a single extra user, and why growth leaders keep reaching for the acquisition budget instead.
Read essay →SEPTEMBER 2026
What a novel about a stranded science teacher gets right about experimentation, and why growth teams keep postponing decisions behind infrastructure.
Read essay →AUGUST 2026
The real cost of slow data isn't the answers you waited for. It's the questions you quietly decided weren't worth asking — and why no dashboard has ever fixed that.
Read essay →AUGUST 2026
The worst drop-off in a lending app came after users had cleared every hard step. Why the fix was pointing, not persuading, and how to tell which problem you actually have.
Read essay →AUGUST 2026
Why we added a payment method that cost more on every transaction, and what it taught me about the costs that never reach the payments line.
Read essay →AUGUST 2026
Growth teams are pointing AI at copy variants and ad creative — work that was never the bottleneck. The expensive work is deciding which of fifteen competing hypotheses earns engineering time.
Read essay →JUNE 2026
Running Gojek's regional expansion across Singapore, Thailand, and Vietnam taught me that distribution isn't a channel plan — it's a theory of why someone adopts your product in a specific context.
Read essay →JUNE 2026
AI makes the right growth motion faster. It also makes the wrong motion faster. The $65B is a bet on the shovel — most teams still haven't figured out where to dig.
Read essay →JUNE 2026
Benedict Evans called AI a 1997 internet moment. He's right. But adding AI to your existing workflow isn't the shift — rethinking which parts of your acquisition logic break when AI is everywhere is.
Read essay →MAY 2026
Activation feels like the lever because it moves. Retention is the ceiling that defines unit economics — and almost nobody names the behavior driving it.
Read essay →APRIL 2026
What changed when we stopped treating vouchers as margin leakage and started designing them as the activation loop itself.
Read essay →APRIL 2026
The cleanest attribution dashboard I ever built told me the wrong story for nine months. The lessons cost real money.
Read essay →MARCH 2026
Most playbooks I see for SEA market entry are European playbooks with the names changed. Why the unit economics break and what to do instead.
Read essay →MARCH 2026
How I cut a $10M+ marketing budget by a third at GOJEK while growing new user acquisition YoY.
Read essay →Stay in the Loop
I post twice a week. Tuesday is tactical, Thursday is reflective. Follow there if you want them as they land.
Tell me what’s broken before we talk. A few questions about your business and the problem in front of you, so our first call starts on your situation, not introductions.
Tell me about your problem →10 questions · About 4 minutes · Reply within two business days
Rather write? Email me directly.