Dealing Fraudulent Traffic

[T J Tutor]

I read an interesting article in Webmaster Magazine about detecting non human traffic, which we are all hampered by in our traffic buys.

comScore acquired MdotLabs in early Aug. 2014 to augment its non-human traffic (NHT) detection methods used in the Media Metrix and validated Campaign Essentials (vCE) product suites.

In order to identify malicious and phony traffic from bots, click farms, pay per view networks, and alike, comScores NHT detection systems combined with MdotLabs allow for a mathematical equation that results in identifying true human digital audiences for our ads and promotions.

Developing the highly sophisticated techniques for identifying and eliminating non-human traffic, which often runs in excess of 50% of my traffic sometimes on a campaign, isn't being embraced as deeply as one would expect because many traffic sources are relying on the extra traffic from these fraudulent sources to increase revenues.

How have you dealt with your traffic sources when it comes to this fraudulent traffic. I can identify some of it and block it, but I haven't ever been able to eliminate all of it.

 

[Dispply]

I guess for now it's impossible to eliminate fraud traffic as too many campaigns are interested in selling it. Making mix of pure "white" traffic and trashy one and selling them as normal quality.
However things are changing, I've seen that last year was first time, when amount of fraud traffic was less then amount of human one (like 47/53).

As being network, we deal with it everyday, so we have some metrics showing us the fraud traffic. We prefer to block source/sub_source. Many my colleagues are using MaxMind.

 

[T J Tutor]

I guess for now it's impossible to eliminate fraud traffic as too many campaigns are interested in selling it. Making mix of pure "white" traffic and trashy one and selling them as normal quality.
However things are changing, I've seen that last year was first time, when amount of fraud traffic was less then amount of human one (like 47/53).

As being network, we deal with it everyday, so we have some metrics showing us the fraud traffic. We prefer to block source/sub_source. Many my colleagues are using MaxMind.

Yeah, I've read a number of white pages making the claim in fraud traffic reduction, but I have not experienced this myself from the various traffic sources I use.

Max Mind looks good!

 

[DanielFIRE]

Networks today can track each and every click to the level of a specific computer so they know how many clicks/installs came from one place even if he is using different ip's.