Posted by Mitch Mitchell on Apr 6, 2009
This is another guest post by Diego Norte; you can view his contact information at the end of this post:
As many of you know, I did a decent sized study on headline color of other people’s sales pages.
Both profitable and unprofitable pages were included in that study. In the end, I reported that we found red headlines to be the most unprofitable color. Green, blue and black were all found to be closely tied for the most profitable headline colors.
This was in stark contrast to the findings of many famous copywriters who insisted that they consistently saw red headlines outperform other colors in split tests.
I still don’t know, but I did throw in “red” as a test case for most of my sales letters.
In most cases, my split tests agreed with the study I had performed of other people’s sites.
However, one multivariate result runs completely counter to that result.
The sales page where red is winning by a long shot is Glyphius.com.
Here are the numbers:
Visitor Value: $0.59
Bright Red (#ff0000):
Visitor Value: $1.76
Wow! That’s a huge difference. The amount of visitors and sales to that page is definitely significant.
Why is the color “red” winning on this particular site, but blue, green and black win on a vast majority of sites?
What is causing such a large difference in the results?
We’re talking about almost 3 times as many sales with the red headline? How can that be?
Why would headline color matter that much in any case (unless it was an unreadable color like yellow on white or pink on purple)?
Unfortunately, these aren’t things I can answer with the aid of statistics. Those are just the facts. On that one site, red is king. I don’t know why. I don’t know how to figure out why. It just is.
These kinds of mysteries don’t need to be solved to profit from the results though. You simply have to make sure they are under test on your sales page. The easiest way to do that is with TestiVar. You simply set it up once and it takes care of everything after that.
It changes the frequency that each version of each variable under test is shown every time you make a sale. Eventually, it will notice that you have a statistically significant result and it will change the frequency that it shows the losing versions of that variable to 0% and show only the statistically significant winner.
You start making a higher percentage of sales long before a statistically significant amount of traffic and sales though. TestiVar is optimized to utilize statistically insignificant data to influence the frequency that likely winners will be shown.
In the above case, the red headline was shown three times more often than the blue, green and black headlines after only three sales. That resulted in more sales more often because red would eventually be found to be the statistically significant winner.