<b>3 weeks. Zero lift. Best test I ran all month.</b>

<b>3 weeks. Zero lift. Best test I ran all month.</b>

Tested a $40k redesign idea on a fraction of traffic first. Beautiful new pricing page. Animations. The works.

Result: flat. Dead flat. No lift, no drop.

Sounds like a loss. It saved the company $40k. We were about to rebuild the whole page on a hunch. The test said the hunch was worthless.

Null results aren't failures. They're cheap insurance against expensive opinions.

— Test the cheap version of the big idea first
— A flat result that kills a bad roadmap item is a win
— Log your nulls, they stop future arguments

Go find the biggest "we should redesign X" opinion on your team and test it small. Report back.