How We Test
Every experiment follows the same seven steps. No shortcuts, no cherry-picking.
We're not scientists and we're not pretending to be. But every piece on this site follows the same honest structure, because a good experiment needs rules or it's just an anecdote wearing a headline.
1. State the hypothesis first
Before we start, we write down what we think will happen. That line goes in the piece, unedited, even when we're completely wrong. Especially when we're completely wrong.
2. Set real constraints
A budget, a deadline, a rule we're not allowed to break. Without constraints, "we tried it" means nothing. Constraints are what make the result mean something.
3. Track the real numbers
Every experiment reports actual cost and actual time, not estimates dressed up as facts. If we don't know a number, we say we don't know it. We never invent a statistic to make a piece sound more authoritative than it is.
4. Write down what went wrong
Every single experiment has a section for it, even the ones that went great. If nothing went wrong, we didn't push hard enough.
5. Give a direct verdict
Would we do it again? Who should try this, and who should skip it? We answer both, plainly, no hedging.
6. Leave one thing worth stealing
Every piece ends with a concrete takeaway you can use even if you never run the experiment yourself. If we can't find one, the piece isn't finished.
7. Publish it as it happened
We don't hold pieces back to "see how it performs" or wait for a better news cycle. When it's done, it goes live. See About for who's actually doing the testing, or submit an idea if you want to see something specific put through this process.