Over the last few months, we’ve talked with many startups looking for tactics they can use to try out new ideas quickly. These aren’t strictly Minimum Viable Products, but rather features they think might work—because of a hunch, anecdotal user feedback, a pattern in their data, or a competitor’s feature set.

But what’s the right way to confirm or deny this hypothesis? Here are a few tactics you can try out to test whether a feature has legs, without investing a lot of time and effort into development or gold-plating.