Paid ads, landing pages, creative and testing, in plain English.
Set the sample size before you launch, run whole weeks, and don't stop the moment a result looks significant. Here's how to time an ecommerce test, and what to do when traffic is thin.
Small tweaks need huge samples to prove anything. With a few thousand visitors, test big, meaningful differences first. Here’s the order to work in.
Every test metric answers a different question, and some crown the wrong winner. Here's what CTR, conversion rate, CPA and ROAS each miss, and how to pick one primary metric plus guardrails.
Change the ad and the page at once and you can't tell which one moved the result. Here's how to structure ad tests, page tests and angle tests so each answers one clear question.
The variation won, you rolled it out, and the lift faded. Here's why that happens, from regression to the mean to shifting traffic, and how holdouts, re-tests and better records protect your next winner.
Random tests produce random learning. Here's how to write hypotheses, prioritize with a simple score, run big swings before details, and keep a test log that feeds the next brief.