Fraud detection in spooky ad networks

BlendTraffic

Newbie
Joined
Jan 24, 2017
Messages
3
Reaction score
1
Hey Guys!

Im working on a fraud detection platform for advertisers.
My current task is to understand the volume of frauds in difference Ad networks.
Please help me gather a list of ad networks you have experience weird none-human traffic in.

I'll list the first one I know -

ExoClick (Porn & File Sharing network)
I bought 39,000 popunders all over the network (US Only) - 4700 of them were automatic bots
(They were using selenium framework)

If you need help measuring your bots, pm me.
 
I would like to know about these:

Adcash
Zeropark
Popcash
Propeller
Revenuehits
Adsterra
Looksmart
Popads
Tonic
gunggo
redirect.com

Just to mention some of them...
 
This sounds like a fun project. Will you publish your findings? Also, how do you determine if a hit is a bot or not?

Here's my contribution: admitad.com It's an affiliate network of affiliate networks, so you should find everything there.
 
OP if you can help with rtxplatform.com
 
I once worked with FraudBuster and 7serch. Nearly all bots according to fb
 
I once worked with FraudBuster and 7serch. Nearly all bots according to fb

Are you saying that according to FraudBuster all traffic from 7Search is bot and fraudulent?
 
Try the Native Ads networks: Outbrain, Revcontent, Taboola, Content.Ad, MGID.

They have bot traffic and fraudulent publishers.
 
Are you saying that according to FraudBuster all traffic from 7Search is bot and fraudulent?

Well it showed a lot of proxy traffic so i consider it bot as it didn't convert even 1%
 
I would probably add AdWord fraud test tomorrow

Its hard (and expensive) for me to launch a bot-detection campaign in each one of the networks.
I need volunteers who already tried those networks, to share with me their Google Analytics (for historical results)
or implement my Javascript in their landing pages (for future results)

Skype/Mail me: [email protected]
 
I would probably add AdWord fraud test tomorrow

Its hard (and expensive) for me to launch a bot-detection campaign in each one of the networks.
I need volunteers who already tried those networks, to share with me their Google Analytics (for historical results)
or implement my Javascript in their landing pages (for future results)

Skype/Mail me: [email protected]

Hi, can you explain more about the javascript you are using for detecting fraud and if you get some kind of report to proof the bot traffic as most of the networks look the other way when you ask them for refund because of this issue. Also will this method of yours can also send the bot traffic to another landing when detected?

I would like to help on this as I have bad experiences with bot traffic!

Thanks...
 
Hi, can you explain more about the javascript you are using for detecting fraud and if you get some kind of report to proof the bot traffic as most of the networks look the other way when you ask them for refund because of this issue. Also will this method of yours can also send the bot traffic to another landing when detected?

I would like to help on this as I have bad experiences with bot traffic!

Thanks...

The bot detection algorithm is based on many things, two easy examples would be -
Automation frameworks (PhantomJS, Selenium, etc..) leaves a signature on the page that I can detect.
Bots usually use Cloud / Servers hosting environments, this also could be detected.

We can generate a proper report for the ad networks, that explains the reason we suspect each one of the users of being a bot.
What the ad network would do with this kind of report? I've no idea.
Maybe we can generate some kind of a blacklist, so the advertiser could avoid some fraudy publishers over the network.

Bot traffic could be easily transferred to other page.

You are welcome to contact me directly
 
Last edited:
Fraud detection is one of the most challenging tasks for most people. But, now a days, their are so many products available in the market that help to identify fake clicks and help to save your budget from fake clicks that are done by bots.
 
Back
Top