Split testing your affiliate campaigns is always important. It is amazing the different things you can split test and the different results you will see with minor tweaks. With Plenty of Fish one of the primary things I like to split test (once I have a banner with a decent clickthrough rate) is the login count. The amount of times someone has logged into their POF account says a lot about the type of user they are. Earlier this month I started a POF campaign testing different login counts promoting the lesbian dating site PinkWink. The results are pretty interesting and I thought I would share them!
In the example below you will see the banner I used, the traffic I received from Plenty of Fish based on the targeting for each campaign, and my results from Prosper202.
Unique POF Banner
Having a unique banner is really important with Plenty of Fish if you are going to get a high enough CTR (clickthrough rate) to be profitable. Most banners have the text on the right and the image on the left. I decided to make my 310×110 banner with one big image across the top and a small strip of text at the bottom to make it stand out. I am direct linking the campaign so I knew if I used the women from the landing page that would help make the connection from POF to the offer (helping conversions).
Here is the banner I created:
It is easy to see the connection right? The close up on the women’s face should help grab the attention of the POF user and the pink bottom and border will make it stand out even more. My CTR has fluctuated, but based on the 60,000+ impressions I have gotten it’s off to a good start.
Split Testing Login Counts
My main reason for running this POF campaign was testing different login counts. Plenty of Fish lets you target in 50 login increments minimum so that is what I did. I setup 5 different campaigns (shown above):
- PinkWink lc50 targeting users who have logged in 50 times or less
- PinkWink lc100 targeting users who have logged in more than 50 times, but less than 100 times.
- PinkWink lc150 targeting users who have logged in more than 100 times, but less than 150 times.
- etc…
So, the users I’m targeting in my PinkWink lc50 campaign are obviously the newer users to Plenty of Fish. The PinkWink lc250 people have logged in 250+ times so they’ve been around awhile. Before I started the campaign I assumed that the best converters would be the PinkWink lc50 people and the most volume would come from PinkWink lc250. Well, one of those was right.
Lesbian Leads
I do not have a ton of data for leads. I have only gotten 44 clicks total, but the data I received has been interesting. I have gotten the most clicks from the users who have logged in 250+ times and they are also the only group of people who have converted. The second biggest group were the people who had logged in between 51-100 times. Although I’ve gotten 8 clicks from those users, not one of them have converted. Based on the volume and conversion data I have, I should optimize for the 250+ people.
Now, I know I do not have a ton of data here, but the data I do have makes me think I can keep this profitable. I have gotten another lead since I originally took these screenshots and it was another user with 250+ logins. I would assume that these women have already all seen PinkWink, but apparently it is not saturated within that market so I may have to try to ramp it up. With the PinkWink lc250 campaign I’ve made around $7 profit, which is nothing to write home about, but if I can find a way to draw in more traffic I may be able to scale it.
I hope this case study has got the gears turning in your head and you able to use my information to setup your own profitable POF campaign. Please let me know if you have any questions. Let’s make some money!
Yet another great case study on POF Luke, very inspirational.
I have not tested POF before now. I will try it out . Thanks for the post
So this is what your life has come to? Trying to get lesbians to click on your ad and sign up to a lesbian dating site? Your parents must be proud.
Lol. I hope you’re serious.
I’m dead serious. What do you tell people you do for a living? Well, I, umm, i sit around all day in front of my computer trying to figure out how to make money off the lesbian dating scene.
Google Adwords thinks your blog is about lesbian dating as well. I guess it is right.
I was looking for the “like” button on this comment and realized this wasn’t Facebook. So anyones, my favorite comment on this entire site. :-)
Joseph, are you mad because we make more money while we’re snoring than you do when you’re busting ass? You pist Bro? What are you doing on this blog any way?
Great stats Luke! Shows how important testing is! Guess I need to start some campaigns on POF. Thanks for sharing :)
I asked this question in your last post,but did not get any reply…
How did you make the banner at the very first of this post?
I created the banner using photoshop.
why the fuck you are promoting lesbian dating
that’s fucking gay
Great case study Luke. I would have assumed the new users would convert the best as well. The results are quite surprising.
Hi Luke,
Nice case study. I just want to point here that majority of affiliates always missing a very important key when it comes to advertising on POF and that is the variable POF providing for us to use in our ad copies and I am sure that if affiliates will start to use them they will say a jump in their conversions rates and here I will name some of them so people can start to use them:
{gender:default} –> Replaced by Male or Female
{age:default} –> Replaced by exact age of the logged on user
{state:default} –> Replaced by name of the state or province of the residence of user.
An example about this we can write
Title: Are you {gender:default} and {age:default}?
We have perfect match for you {state:default}. Just click here to get started.
Regards
Mahmoud Selman
Was wondering, you didn’t talked about your bids for this case study. Also, did you bid the same thing for all login counts (or added .05$?0.10$?) the closer you were to LC50?
Hmmm…you’re right. My bid was $0.35 CPM and it was consistent across all logins.
Don’t you think you’d have gotten better results by raising your bids the newer the users were? Also…following your campaign and LC setup, PinkWink lc250 would actually be lc200-250, not 250+…So did you just skip the 200-250 and made it a 250+ lc?
It may have had different results had I staggered the bids, but I was happy with the results I was seeing and didn’t change them. I wouldn’t suggest this for anyone looking to really scale/optimize.
Yes, it looks like I skipped 200-250. Didn’t even notice that. lol :)
Hi Luke, which email editor would you recommend for this? something simple, not Photoshop or Gimp….
You might want to try Pixlr.
Hi Luke, I tested POF over a two-three week period, not in the same target market and I tested starting where you were price wise and rising by nearly 50% and found it hard to monetize and also frustrating in the number of ads that were either rejected or had to be tweeked and resubmitted
Obviously it must work for certain people because it funds the operating side of a site that’s run this way for years but I found it difficult to turn a profit on it.
I found it very difficult to convert on POF….No matter what the offer was!
hey Luke, i Just wondered would you be willing to share what other targetting you used on this case study? I ask because I have a pinkwink campaign running on PoF right now and am getting hardly any impressions even though my bid is 51 cents CPM