The follow along THEY don't want you to know about (ok its just mobile follow along) (5)
The follow along THEY don't want you to know about (ok its just mobile follow along)
Hopefully I wont be banned for using click-bait 
But its what we do ,right ?
Little Introduction (---ugh not that little )
Skip to follow along section if its too much to take
Hey everyone , I'm new here , full time affiliate for 3 years , started PPV , then had my best months on native ( you could have guessed it from the title) made xxx,xxx overall on native. then I lost it all , not all but I lost money , focus and it took me a while to get back on my feet , few months actually, terrible I know but I put myself in this situation and no help on the way anytime soon so its up for me to fix what I ruined. my budget isnt huge now , but it isnt small either , $10k.. the good news is I'm in a 3rd world country so monthly cost is not an issue , am still full time. but I have to get back on track soon, am giving myself 6 to 8 months tops otherwise I will be in trouble.
I have to be upfront and say I really hate follow alongs , why? because most of the time they just die , they start with excitement and then people just disappear. that was my experience from another forum and I think I iamattila mentioned something about it somewhere.
The Follow-Along
So am one month into mobile , I started with some waaaay toooo low payouts just to have the feeling for the numbers since its significantly different from my native campaigns with so many variables its annoying sometimes , why the hell did I leave native , anyway!
Currently I'm breaking even on a campaign its app install
Spent: $529.82
Revenue: $480
Payout:$0.5
I think the real numbers are closer to breaking even because I spent around $40 one day without realizing I changed the folder name the other day so the traffic went to 404 , so its breaking even for two weeks give or take.
I wanted to give it up actually but my AM told me I should stick with it since many publishers tried this one and many didn't convert at all so I must be doing something right , I "emotionally" wanted to quit this one but I know for sure its better to stick with it , breaking even is not a bad place , just need a nudge on the right direction.
There is a good post on Malan darras blog about the same topic http://www.malandarras.com/greenred
So the problem is am still looking for patterns with this one , I cant find anything thats worth cutting , I use all sort of statically significant calculators and tools , no device worth cutting , no isp or carrier worth cutting. So I think its time to look for combos , which am not used to. I also made a big mistake with my testing , the budget is not spread all over the day , which is causing me two problems now , I cant look into day parting data on
Here are some numbers for June , not sure if its a decent start or not
I have a question hope someone can help with:
Looking at some numbers lets say 8th of june I had $106 cost and $104 revenue
Traffic source charged me for :32,810 impressions
Affiliate network recorded: 12,828 impressions ONLY 7,226 are unique.
The CPM for that day was $3.24
I'm using a vps on singapore and my geo is in SE Asia , am also using cloudfront. the server is 4GB and 2 cores , I tested it on 1000clicks per minute using this sweet little tool https://loader.io/ and had no spikes or any problem in fact I felt the server is too much for the traffic am getting right now.
So I feel something is wrong with the numbers here , the 5.2k loss from traffic is fine and acceptable , but the voluum recorded clicks and affiliate ( not unique) clicks there is another 4k impressions lost there , needless to say that half of traffic is not unique so does this add up? am I missing something here? I feel like something is wrong between voluum and network there is a big loss and also traffic source is giving me way too many non unique traffic ( I was using 1/24 cap on this day). any insight or tips on this would be helpful guys. If you read all the way to the end then you are my hero. I've been banging on keyboard for a while I know
Breaking even is def not a bad place man. I hope to be where youre at in a month myself. And i dont have much to feedback for you but i remember vortex providing some resources on the topic of page speed in the big AM FAQ thread. The links are under the
"Why is my lander CTR so low?" part:
http://stmforum.com/forum/showthread...AM-FAQ-Thread!
Hopefully there of any use to you. Good luck and hope to see ur progress!
Originally Posted by abysss
So just when I was about to launch the campaign on a new traffic , I found this
I took it for granted that cloudfront is handing files well while I was leaking there. I will make a small launch tomorrow on same campaign without CDN but on a nearby server only and see if it affects anything , its not a real split test obviously but I will check how many traffic loss is there and see if the CDN is the main reason , still cant explain the duplicate traffic but I can see why you should always blame yourself before blaming the traffic source. on 8th of june only I lost 18% of my traffic from cloudfront , I'm not sure if my setup would be better on S3 direct to cloudfront and take linode out of this.
Update:
After I had this problem with popads , I contacted them I gave them screenshots of everything , they said its not their fault , and its the affiliate network problem and how they calculate raw and unique traffic , I launched the campaign again and went from breaking even to -60% not sure if its a coincidence but feels like they punishing me for complaining by s sending me bot traffic now.
I opened up dropbox that day and had a look on my native campaigns I launched , did some spying and launched 3 campaigns , 2 failed and I kept last one running
I still don't want to do native but it feels good to have some green days again , hopefully I wont quit pops I want to master it soon.
What I will do now is keep the native campaign on background and do minimal optimization to keep it green and focus on pops again , my next goal is xx/day on pops , I will try same campaign on a different source. popads are devious for doing this , the CR changed from 1.8% to 0.6% overnight , I ran it yesterday and I ran it today and same result. not going to run with them again
On technical level I was doing a mistake on cloudfront , I was cashing the whole page instead of different elements , its on their terms that if you get more than 3000 clicks per minute you might have runtime error, or at least thats what I saw, I left cloudfront and went for maxcdn , started to cache elements only not the whole page , their rep was confusing however he said there is another setup to have the whole page cached and he doesnt recommend I cache the whole page , but when I tested it it was working fine , he said he cant guarantee how my setup will perform under heavy traffic but he wasn't clear enough what the other setup would do. anyway I stopped caching the whole pages now.
Home > Mobile > Follow-along Campaigns