A tool with a lot of false confidence built in

Surveys are tempting. They’re fast to send. They produce numbers. They scale to hundreds of responses without much effort. They feel like real research.

They’re also one of the most over-used and over-trusted research tools in product. Many founders run surveys to feel like they’ve “done their research,” and end up making decisions based on data that looks rigorous but isn’t.

Surveys aren’t bad. They’re just narrow. They’re useful for some things, useless for others, and downright misleading when applied to questions they can’t actually answer.

Knowing which is which is the difference between using surveys well and using them to fool yourself.

What surveys are good for

Surveys are at their best when you want quick answers to simple, factual qu...

When surveys help (and when they mislead you)