I ran the button-color test so you don't have to. Nothing.

I ran the button-color test so you don't have to. Nothing.

Gut call said red beats blue. Gut call was wrong. So was the test, basically.

Setup. CTA button. Control = our blue. Variant = aggressive red. 7,800 sessions, big sample, 12 days.

Result. Red 3.21%, blue 3.18%. Inside the noise. Flat. The famous 'red button = +21%' case studies you've seen? Context-specific or cherry-picked.

Then I tested button COPY on the same blue and got +17% same week. Verb beat color, every time.

Stop testing colors. It's the cheap dopamine of CRO and it almost never moves.

Go skip the color test. Test the words on the button instead. Report back.