Carry everywhere · One page, no scrolling through a course
Every framework I actually reach for on the job, compressed to the point of use. Not a study guide — a working reference for the moment I'm staring at a dashboard, writing a brief, or explaining a number to a client.
Different questions get asked at different altitudes. Answering a quarterly question with a daily glance — or vice versa — is how strategists chase noise.
Read two numbers together, never one alone. The combination tells you which layer is broken — hook, body, offer, or landing page — so you fix the actual thing instead of the whole ad.
Before judging any ad: confirm spend has crossed ~3× target CPA or 3× AOV. Below that threshold you're reading noise, not signal — low spend deceit. And always actually watch the ad, not just the dashboard — numbers say something's wrong, watching tells you what.
Diagnosis states what's actually true right now. Hypothesis is the bet on what changes it. Never skip straight to hypothesis — a test with no diagnosis behind it can't be falsified, which means it can't compound your learning.
Run both, every week, on purpose — not as a leftover after "real" work. All-exploit decays as winners fatigue; all-explore never compounds into scale.
Harry's built for men, sold to women buying gifts. The persona in the deck is a hypothesis until purchase data says otherwise — check before every brief.
Primary KPI is the one number you're optimizing against — usually ROAS, CPA, or MER. Secondary metrics exist to explain why the primary moved. Confusing the two means chasing a number that isn't the actual goal.
This is the frame for the conversation above the ad account — the one you have with a founder about where to actually invest. Paid acquisition is deliberately excluded: "media buying has gotten to be such a vanilla, table-stakes thing."
Consistent publishing volume compounds trust before a dollar of paid spend touches it.
Post-purchase survey, repeat-rate by SKU, engaged-but-never-bought cohort recovery.
In-person and unboxing moments — where a large share of real purchases actually happen.
The pillar everything else is downstream of. No amount of creative fixes a weak product.
Not a pillar: paid acquisition