Difference in stats between ad networks and google analytics
 

So I tried buying mobile redirects and popunders from reporo // juicyads //adultmoda and was wondering if you guys have the same discrepancies between the stats provided by the ad networks and stats on google analytics.

I'm was buying only iOS / Android traffic so analytics should pick up all of them...

Ratios I'm roughly seeing are:
Juicyads - 20K redirects // Analytics showing 1,500 visits
Reporo - 33K redirects and popunders // Analytics showing 800 visits
Adultmoda - 15K redirects // Analytics showing 500 - 800 visits

I can understand some difference between stats (people closing browser windows and other) but these differences seem too high.

Any of you can share their stats? (stats are for mobile traffic only)

10x

While I can´t comment on your specific traffic, I´m seeing huge discrepancies between different stat tools everywhere these days.

start watching your own stats...then compare ip addresses...mac addrsess etc

just sayin...we bought some traffic from some places....once

Google analytics often doesn't recognise the java script code from ad brokers.. So sometimes the stats don't show the traffic accurate.

It's a bitch.. :321GFY

AFAIK
GA uses javascript and if browser doesnt support it , you cant count its hit.

but these differences seem too high

Since this traffic was for smartphones only (iOS / Android) - they all (or at least most of them) should be JS enabled (Don't think too many people go to the trouble of disabling JS on their mobile devices)

Did you confront the brokers with this?

You should setup your own server side script to track these things. Third party scripts that are located remotely = no way ever will be accurate....

Well that is beacuse GA generaly have a problem with redirected traffic,for example i once sold german desktop traffic redirect and other side said how they see only 10 hits while actual number was between 1500-2500,and they were using GA.So in shortly when buying traffic use your own counter based on server.

yes, they just said they don't "trust" third party scripts for tracking... was wondering if anyone else is experiencing the same discrepancies for mobile traffic

I'm using a parameter for each traffic source so I imagine it should be the same as any other traffic coming to the site? I'm tracking o GA via landing page hits and not by referrer so I believe it shouldn't matter how the traffic came to my site.

Anyone can recommend a server side script with a decent GUI for this?

Well i spent a lot of hours researching counters and concluded following -none of them is 100% reliable.
And there is two types of server based counters - the one which uses their own tracker(by javascript,php include,whatever)and they need to be manually inserted into every page which you want to track.
Advantage of such script is how they have a lot of features.
Script like this are : piwik,crawltracker,crazystats,etc...
Other type is where script simply read logs of your httpd server(apache,nginx,whatever you use)but they are kind poor when it comes to features,and also while GA having problem with registering correctly hits,these one can show more hits then it's actual number.
Script like this are:webalizer,awstats

Normally i would recommend to use piwik,which is in counting world same what is wordpress in blogging world,but problem with it how is resource hug,so you need to be careful if you have a lot of sites - add site by site,see how piwik is behave,and then do further optimization,like settings optimization or even installing it on separate vps/server which wont be used for anything else.So it can work fine for you if you properly set it.

And when it comes to alternative to piwik,i used for some time crawltrack which is great because it doesn't store entire useragent of visitor into database,it store synonym instead,so for example if user browser is something like firefox 24 blabalbalglfdlgfdlgfdgldf gr9ut93ugf 9gdgrfd it will simply store as firefox and as result that will much less usage of database.But that script last release was in 2011 and i kind got impression how it have some bugs which is why i at the moment hesitate to use it.

So what i will do,probably take either piwik or crawltracker and modify them till they will work properly.
Or maybe even modify a trade script,since trade script already log same info as counters and that way i would avoid double writing of same data.

Thanks for the detailed reply though re piwik - not sure how it will be different than using GA since it uses JS it will probably give me the same results.
As for server based programs - I have awstats, webalizer and analog installed on the server but getting the data I need from them is virtually impossible

Interesting enough, I added a couple of events for that redirect traffic yesterday and now when looking at the GA report for yesterday, traffic has went up by 7 times with most of it reported as "(not set)" landing page so I can't really get any info from it. Very frustrating

there are so many clueless people buying bulk traffic or people who don't know how to track it..

+- 10% compared to GA is normal.. if more it's 100% rip-off

Well, after some code tweaking I now see more visitors coming from ad networks though still about 100% difference than what is reported on the ad networks.

Another thing I notice now is the crazy bounce rate (around 90%).
Are you seeing similar bounce for mobile redirects / popunders? Worried it might hurt the site's SEO efforts