So I posted I would do a case study a couple days ago. It’s been running through quite a few impressions on PoF and I decided better to cut my losses on an experimental campaign before I start blowing more than the hundred or so I spent on it.
Let me list the targeting options I selected on Plenty of Fish ads, again.
Overall, performance on the CPA side of things were bad. That is what you get when you target broad like this on Plenty of Fish. Dating offers on Plenty of Fish work, but only if you put in the effort to dig down deep and micro-target, build landing pages specific to your targets, and split test different offers on each of those targets. On this campaign I spent $110 or so and made maybe 30 – 40 back… so I know there is potential in the demographic, but I was just direct linking and that’s pretty much a no-no these days.
Here is the image I used:
An old, old image I used to use on Facebook. This image probably made me 10 – 20k on its own a couple years ago.
Anyways, how did each color perform? Here are the results ordered from highest CTR to lowest:
Some of the results here are pretty interesting for me. While it isn’t all substantial evidence, the discrepency between green and red is pretty shocking. I was thinking red would rank much higher, but instead its all the way at the bottom.
Green, however, ranking 1, didn’t come to me as a surprise. Time and time again, every time I post an ad with a girl wearing the color green, it has usually been a winner for me. Use green, folks. It has worked for me time and time again, and is somewhat proven by this small little case study experiment.
Again, none of this evidence is completely substantial. It’d probably been better to run each ad through 100 clicks, but knowing that PoF dating ads run out of steam fast (the CTR can drop like a brick in the course of a week, in some instances, days), I figured that might cost me upwards of $500 in the end. One thing to remember is that these colors might not work for everyone either. Females may prefer red over green. Seniors might like purple more than black. And changing the image might even affect color CTRs. I may end up experimenting with all of these to bring you a more conclusive data set in the future.
Copyright © 2006-2011 JustinDupre.com
I had a feeling that red would perform worse than green and blue but it's crazy how different they actually are. Thanks for the info!
I wonder if this has a psychological link to the traffic light system. It seems coincidental that green/go is at the top and red/stop is at the bottom. Interestingly we've found Orange buttons split test better than other colours.
I think it is that red means warning or stop in our subconscious.
this is incredible thanks for sharing this valuable information here :)
Excellent post! I am really surprised about the color red though and that it's at the bottom of the list.
Thanks!
I would have guessed blue, because I have always heard that means trust and green means greed. Good study though, awesome to see the results to use them for future campaigns.
Cool case study, I would have guessed that red would have performed a lot better…
I'm really surprised by blue, since POF is blue I thought that would have performed much worse than it did.
This is not statistically significant, which means it's a worthless case study.
Like Derp said, this is no where near statistically significant…. this data is completely useless.
I do think, yes, I could have done a bit more testing to see which color would top the list, however there is a very clear conclusion around the color red when you compare it to green or blue.
There are a ton of factors that could kick red into the top spot, and many variables are completely out of an affiliates control (if you've been an affiliate marketer for any decent amount of time, this is basic knowledge… it is not a stable business!), but for this particular test, demographic, image, and platform, red is not a winner.
Derp and Joe don't know their statistical significances from their.. elbows.
The Green and Red tests alone comprise almost 100,000 impression, and Red shows a clear 70 percent lead in performance.
If that's not "statistically significant" I don't know what is, maybe a magical hand writing instuctions on a wall?
Oops, I mean the Green outperforms the Red, obviously.
I think green makes plenty of sense… after all, green m&m's are supposed to be for romance (and rumor has it, they make you "frisky") … I think green actually has an additional undocumented effect on our minds. ;)
I would think that woman would click red more because of a red heart + love. It seems on your test it had the least amount of impressions and least clicks. SO I guess I agree with you Justin, red should of been top, but i guess not.