The Best Practices Trap
Best practices are survivorship bias with a PowerPoint deck. They tell you what worked for someone else in a context you'll never fully understand. Here's what to do instead.
"Best practices" have a hidden assumption: that what worked somewhere else will work here. This is survivorship bias dressed in business casual.
Phil Rosenzweig's research in The Halo Effect showed that most "best practices" are reverse-engineered from successful companies after the fact. We look at winners, identify what they did, and call it a formula. But we never study the companies that did the same things and failed.
Google's 20% time "best practice" worked at Google. Yahoo copied it and nothing happened. The practice wasn't the variable. The context was.
The deeper problem: best practices discourage experimentation. Why test something new when you can import a proven playbook? The result is organizations that optimize for conformity and call it strategy.
The reframe: Replace "what's the best practice?" with "what's the smallest experiment we can run this week?" Best practices give you someone else's answer. Experiments give you your own data. One of those compounds over time. The other doesn't.
