Dear Stackerinos,
It is my honor to introduce you to yet another follow along!
Feel free to heckle me, call me out on my bullshit, give me tips, thoughts, ideas, ask questions, whatever. This follow along will hopefully last for quite some time.
Intro:
Been doing numbers between $500-1000$ revenue/day for quite some time. Full time aff marketer for 3+ years. I have always been stuck inside my little comfortzone of traffic. Traffic that i know my way around. I know what works there but it has its limits. It is time to change that. To force myself to not be sloppy and unstructured in my affiliate marketing i decided to create this follow along in order to achieve a few goals;
a) Step out of my comfortzone.
b) Really follow through on a campaign cutting unprofitable things (almost) only based on significant statistics. To this day i have always gone purely on my hunch as to what will be profitable. This has resulted in MANY potentially profitable campaigns that has had -10%ROI after initial testing, then to be given up on due to my lack of Donald Trump-ness.
c) By keeping this diary of my actions i hope to come closer to a clear process i can replicate to make any new campaign profitable. I want to develop a step-by-step guide for myself that i can look at and follow when i launch my next campaign.
Setup: STM Tracker on dedicated server, GIMP Image Editor, Cheap programmer from oDesk and my sass.
Offer: I will not be split-testing offers for this follow along. I know this offer converts well for others and i will stick with it for now unless it goes down. It is a UK Pin Submit Anti Virus offer. UK is pretty strict on their advertising laws afaik and the offer owners are very strict with keeping landing pages compliant. They will review every single lander i want to run. This hurts me a bit as it will take ages to get new stuff approved. Hopefully i will quickly get a hunch on where the line goes on how aggresive i can be in my texts and then just change LPs on my own. The offer only accepts a few UK carriers which is good as it slims down my targeting a bit from the get go. Also it's Android traffic only. Sadly no Symbian 
Ad networks: Buzzcity & Millenial Media. Why? I feel like these 2 networks are on each side of the lenience-spectrum. Buzzcity approves almost anything whilst Millenial Media is super strict (from my experience anyway). It will be interesting to see the difference in traffic quality & conversion rates across these 2 networks. I'm a bit worried about getting stuff approved on MMedia. If any of you have experience running antivirus type stuff on MMedia, feel free to give me some advice.
Budget: I will see how quickly it progresses and how close i am to being in the green. The closer i feel the more i can spend. But i'd be willing to tank 3-500$ per day after some promosing initial numbers.
Banners & LPs: Will make 5x different style landers to start testing with. 2 of which i know convert well from previous anti-virus campaigns and 3 completely new ones. Some features on my LPs that i am going to try boost my CVR with are: Testimonials (not seen that done on Anti Virus before), Countdown timers, Date Call Out, GEO Call Out, Phone Model Call Out. The banners will be a mix of old successful banners of mine and some more tailored towards Samsung system messages.
I have decided to start my campaign targeting only Samsung phones. Why? Samsung phones account for ~30% of UK smartphones according to some graph i saw somewhere. Thus, Samsung has the most volume by far. Choosing only 1 brand to begin to conquer also gives me some other perks. I can tailor ads and landing pages to imitate samsung designs, i could carefully sneak in some samsung logos in my ads etc. I also know from previous campaigns running other anti-virus type stuff that some Samsung models are very strong converters.
The plan:
The goals for this campaign before i can consider it a success is that i would like a 1000$ profit per day with at least a 160% ROI (Where 100%ROI=Break even). How i imagine this situation to look like is that i will be running the campaign on about 2-4 ad networks, preferably on a fairly high CPM targeted to my proven winning LP/Model/Carrier-Combos. It would be ideal if i could have 1-3 big campaigns on these networks instead of having to micromanage 50 camps on each network with very slim targeting.
How do i get there?
In order to achieve my goals i have set up 2 strategies that i would like your input on which makes more sense.
Strategy One;
I start by running RON campaigns targeted to the right carriers + Samsung phones on both networks and start split-testing my creatives hard. I will keep on split testing landing pages and banners until i have achieved a certain predecided (negative or positive) ROI. In your opinion, what ROI would make sense in this case? What would be a reasonable goal for ROI on straight RON traffic without having cut any pubs or phone models and carriers? 110%ROI? 90%ROI? 70%ROI? 50%? (100% being breakeven). After finding the optimal banner & lander i start cutting models/carriers/publishers and start split-testing minor things on the LP like colors/images/whatever. Then start spreading to other networks and buy myself a pure gold Maserati.
Perks to this strat; Will propably grant the more volume. The better the LP/Banners are from the beginning, the more models and carriers i can theoretically convert.
Strategy Two;
I take the most successful banners/landers from my initial test, try to make some improvements on them and then just straight run them for a while gathering stats (unless ROI is absolutely TERRIBLE). After a few hundred dollars spend and some conversions i should be able to start cutting carriers/models pretty quickly and once a winning combo is found i start testing landers more heavily.
Perks; This strategy (given that at least 1 LP does decently OK) i think will allow me to find +ROI quicker, but sacrificing volume (by cutting models/carriers early). If i find 2-5 Samsung models that way are better than all the others, i could then scale that model+carrier combo on to more networks and hopefully have instant +ROI there. The issue is that the offer has quite a high payout, thus making the farming of significant stats more costly. It also sucks a little bit that you can not see OS versions in the STM tracker.
For some of you who has had success running similar campaigns, what would you say is the breaking point for success on these types of camps?
Is it finding that sweet combo of OS version, model, carrier that suddenly granted you god given ROI?
Or is it that AB tested landing page that bumped your ROI from the red to the green?
Mix of both? If so, which should i strive to achieve first?
And, if you have any input on the specific traffic sources i chose. I am all ears to try different sources (INB4 Decisive). Do you think that Antivirus Pin Submits will be a no go on MMedia for some reason? Airpush?
I might have a little prematurely posted this as i am still designing landing pages and banners and i have plenty of other stuff that im working on simultaniously. So it will probably be a few days if not a full week before traffic goes live. Granted i need to get all my LPs approved by both the offer guys and the ad networks. Also my Matrixguy needs to finish sprucing up my LPs. I'm in this one for the long haul anyway so fuck it! Hopefully there'll be some discussion before and i can avoid some early roadbumps.
Much Love
The Peoples Champ
Airpush rejects antivirus banners, landers as misleading, since March 30th - Thank google for that one 
@iamattila,
Hm, bad news, i had imagined Airpush could work well for this campaign otherwise. Rather give you a thanks instead of google mate.
Right now making the last 2 landers. As per usual i get REALLY excited as i design them. IF PEOPLE DONT CONVERT ON THIS LP I DONT KNOW WHAT WOULD WORK is in my head at the moment. 2 of my LPs will replicate the offer LP design which should also boost CVR's a bit.
Good start mate, looking forward to seeing your progress!
Good luck with buzzcity uk traffic.. never got that to convert..
the south american traffic on the other hand, can be gold..

Looks extremely good! Looking forward to seeing how this one progresses.
One very minor question - why GIMP? I'd expect you could squeeze a lot more productivity out of Photoshop, although of course it depends how much you've scripted in GIMP.
Thanks for the encouragement guys.
Now finished my landing pages and made loads of new banner templates all sent in for approval... Hope to hear back soon.
@Caurmen
No real reason to use GIMP. Just started using it years ago and never really took the time to learn photoshop properly. Though i have never felt that GIMP was holding me back in banner creating. There's a lot of scripts and stuff like you mention that you can add if you want some specific features like batch processes etc.
Getting an AV Pin Submit through on MMedia is going to be very tough. At best they'd let you direct to an App Install but never a Pin Submit. Also they've been getting stricter with creatives as time has gone on but you can get creative. Good luck man!
I can save you a lot of time and money...
Pick another offer or pick 2 other traffic sources. Like the honorable Dr. Lavish pointed out Millenial don't allow anything relates to battery or virus and Buzzcity is known for having shit traffic. By all means, go back to scale on Buzzcity but its not a great place to start.
^ definitely Lavish & Deondup are mobile masters
People are still pushing AV hard, just with different approaches from the new policies out there. Have to keep searching and testing bro
Anti-virus is being pushed on Airpush with cloaked setups like this:
BANNER: "Android slow? Clean up now! Install >>"
PAGE: "Scan completed! You have the delete_contacts virus, and [12] other viruses! Click to add the #1 anti-virus protection for Google Nexus 4 before your contact list is wiped! Install and Clean >>"
The difference is that you can't say "you have a virus" on the BANNER anymore. You just need to sufficiently scare them enough on the LP (and show a different one to the reviewers/Airpush people), and entice the same sort of click with the banner, perhaps using interesting icon imagery like I saw on this banner (it was like an evil pacman eating thing image, suggesting virus, would have gotten a good CTR).