Uncovering an advertising fraud scheme (NSFW) (2011)

Although an absolutely great article, this is 2 years old. Previous discussion here: <a href="https://news.ycombinator.com/item?id=2333824" rel="nofollow">https://news.ycombinator.com/item?id=2333824</a>

This is a major ongoing topic in the ad world. For those interested in the topic, here are two more articles from today on publisher fraud:<p><a href="http://www.adexchanger.com/online-advertising/to-catch-a-botnet/" rel="nofollow">http://www.adexchanger.com/online-advertising/to-catch-a-bot...</a> <a href="http://www.adweek.com/news/technology/meet-most-suspect-publishers-web-148032" rel="nofollow">http://www.adweek.com/news/technology/meet-most-suspect-publ...</a>

So whats the major disincentive to people who read this and immediately think "Kaching!"? (me)<p>There must be a down side, apart from the seedy feeling someone might get from doing this.

I tried running some PPC ads once, for a month or so. I spent a lot of cash and got nowhere. Curiously I showed a healthy conversion rate on some ads, which encouraged me.<p>I spent an entire day tracking down a lot of these things. Turned out that in every single case, there was no sale, and I couldn't believe in a million years that traffic from these sites were really following my ads.<p>I stopped spending money with Google. A year later they sent me a voucher and begged me to try again. So I tweaked some settings, cut the list right down and spent their $150. Same results.<p>My overriding impression is that display ads, PPC ads and the like are an absolute cesspool of fraud and it's a waste of time trying to untangle the mess.

&#62;For the technically curious: reading the address of the top frame is a challenging problem. For security reasons, browsers do not allow cross-domain scripting. So, it is not possible to just call the "top" object and read its properties. We have a proprietary solution for this.<p>Are they exploiting a security bug?

I can recognize a lot of stuff I used to do when I was 16 and owned that kind of website. A lot of linking to send traffic to other websites who would send traffic back. A lot of fake traffic through iframes, and often through &#60;img src="url" width="0" height="0"&#62; because iframes fake traffic would be protected against. A lot of exchange with that kind of websites.<p>There was a huge bubble.<p>Also those websites gives you what you want to see. We all started doing porn, but the most clever guys understood that you had to post more. Gore, Fights, Shock...

Wow, searching for "buy traffic" or "real human traffic" (sans quotes) is an eye-opener. They may be real humans and not bots (though I doubt it) but it's still fraud.<p>What a cesspool.

Ironically, this is only possible because advertise resellers created all sort of markets for scamming the publishers of actually getting paid for the ad they show (advertises have to pay for 3rd party reporting, brand protection, etc, etc, etc. all that dillutes the $3~4 cpc the brands pay to the 3c the publishers receivers in the end)<p>if it worked like it works in TV or radio, this wouldn't be possible to begin with. or wouldn't be profitable.

I think I'm missing the connection to the "legitimate" publisher. Is the publisher paying the scammer for traffic? How are they doing that? unless they're in on the scam as well.

Close... but the business model for most of these operations is traffic generation. Need 1M uniques this month to hit your numbers? No worries!! Enter your credit card here!

FYI: The NSFW content is a handful of tiny pornographic images.

one flaw on the estimates: it assume 100% of the uses leave the popunder running.<p>anyone have any data on the effectiveness of popunders? this is one area i never have to deal with (thankfully)

That was a very insightful read, thanks.