Before getting started I've been doing alot of reading and my sole focus is going to be facebook from here on out. I've been rapping my head around how many angles I can possibly take and it seems endless.
Things I need to get familiar with while working in this campaign is proper ways to test. So before I get started I'm trying to figure out a proper way to test images. I have no clue what to grab other than just random stuff.
The angle I'm going to start with is date team fans. Take a date to the game etc...
I know this may not be the best way to start, but at the least i can get some experience and see what works.
When looking for pictures of say girls or guys with team jerseys on. Do I select different hair colors. Select a blonde, burnette, red head, off color streaks (blue, gree, red) etc.. Find 5 images and see what has the higest CTR?
Then say it's the blondes getting the most clicks. Do I find 4 blonde images and 1 red head and keep testing that way?
I know there really isn't a wrong way to do it, but this will help me get started and moving foward.
If it was me I would:
1) Target single guys in whatever age range, target a specific team and all things related to that team only.
2) Use images of some hotties wearing that team's gear.
3) Don't get caught up on hair colours, just find some chicks who make you think "daaayum I'd like to sit beside that".
4) Use say 3-5 images at first, make some duplicates of each ad so you are running the same image multiple times - two identical ads will stick differently depending on their initial CTR and this can be the difference between -50 and +50% ROI on an advert.
5) Delete/pause ads that aren't cutting it, rotate in one new image at a time to replace those you killed off. Keep your ad groups TIGHT inside a campaign, i.e. don't give Facebook many ads to choose from inside your campaign, otherwise the ad delivery falls to shit and 99% of the impressions go to an established ad.
6) Make new campaigns doing the exact same thing but targeting a different team.
Once you get a process going just have a bunch of images queued up ready to go into rotation once you've tested the ones currently running. IF you start seeing a trend with blondes giving consistently higher CTRs for a specific campaign then sure - find more blondes and test them. Just don't assume anything to begin with. You never know - you may find blondes giving higher CTRs every single time and then one ginger with a pink stripe comes in and dominates.
Once you have some traffic flowing and conversions coming in you should get a general idea of your EPCs network-side. Knowing that, you can then think "ok, my advert needs to get below that threshold cost per click to be likely to make me $$$". Then you will get an idea of what CTR you need to get, so can make decisions about killing off adverts. Just make sure you look at the EPCs for a specific advert (use subids!) before killing it off - it may have a low CTR but for whatever reason it's conversion rate could be higher giving EPCs 3x higher than your average.
You're getting help from Mr Green and Smaxor right? Hit them up if and when you have dilemmas in testing!
I answered your questions from your PM here.
What and how should I pick these images? The obvious answer would be girls in chiefs jerseys. Correct.
What would be considered a good test? Girls with different hair color. Then if the winning ad is blonde. Choose all blondes or what? Every image is going to give you different results. Just because a girl is blonde it doesn't mean she has the same facial expression, composition, or lighting as every other blonde. You need to test all kinds of images.
I'm also looking around on places like hotornot.com, but I heard not to take images from google or bing. What would you recommend for finding pictures of girls with jerseys or even just small niche focuses like this for the future? There is a boat load of references to dating images on these forums. You can find them by searching for dating images in the search bar.
Do I start with 5 images and 1 ad copy. So 5 ads in total per campaign? Follow Zeno's #4 point.
----------------
Bottom line all these answers are assumptions based on my previous experiences. You need to stop asking questions, stop trying to be perfect, and just get your hands dirty. You learn so much faster this way. How much money have you spent testing so far?