AB Testing / Positive vs negative colour usage
Will more users sign up if the action button is a positive colour as opposed to a negative colour? (i.e. green vs red)
My A/B Test totalled with 12 total testers and in the end the hypothesis was confirmed, although the results may have varied in different conditions and with another category of testers. Being friends with the testers made them treat it in a more relaxed way, and may have affected the test in a way or another.
Six users were more inclined to click the green button, and even if two or three users proceeded on clicking the red one (at least once), they admitted that the colours in the first version (control) were more “inviting”. Reasons for not clicking the red button were oriented towards the colour of the button (“Red was too dark”) and towards the fact that “It didn’t feel right in that context”.
Four users were inclined to click the red button in the second version (variation), although all of them clicked through the first version (control) as well.
One of them said: “even if I like both versions, clicking the green one seems like the logical choice here.”
While one user decided not to click any of the two buttons with the reasoning was that they did not want to subscribe, another one decided to click all of them, stating that “it makes no difference to him”.
After researching if more users sign up if the “sign up” button is a positive colour (green) as opposed to a negative colour (red), my A/B Test resulted in my hypothesis proving out to be true, with 6 users clicking the green buttons versus 4 users clicking the red ones. Two users decided to do things that I couldn’t count as actual results (one did nothing, one clicked everything and had no reason behind it), therefore those two were not added to the total.
In the 6th week I had the opportunity to have the A/B Test in an environment filled with classmates, which meant I had a lot of test-subjects I could use to confirm my hypothesis.
Doing the test with acquaintances (friends) meant most of them would not treat the test seriously or would joke around while testing, which was definitely a variable that influenced the results of the test.
Another variable that sure had a part in the results of this test was the fact that all of us knew about the A/B Test — therefore, everyone knew there was a “catch”, so the results were mixed and the reasoning behind user actions was sometimes unclear, or people simply picked something because they wanted to joke and the test not to go right.