Google has uncovered a bit on how they manage ad safety with a new blog post yesterday afternoon.
Google announced they have shut down about 800,000 advertisers and blocked over 130 million ads from showing up on their network. They have come a long way in this regard and it is a constant battle between Google and rogue advertisers.
Google has documented some of the general steps they take to detect bad ads and advertisers.
(1) They monitor ad and landing page content to detect scams, malware and such.
(2) Manual reviews of ads after being flagged by an algorithm
(3) User generated spam reports
Here is the full video:
For those who can't watch it, here is the transcript:
0:05 Hi, I'm David Baker. I'm the head of ads policy engineering at Google.
0:10 We have a vested interest in creating the least intrusive and best ad experience for
0:15 our users. That means showing you the right ads when you want them, and doing our best
0:20 to never show you ads that can be harmful. That's why our ads engineering efforts are
0:24 so focussed on protecting the user.
0:27 This is a very difficult challenge given the number of ads online, and there are a lot
0:32 of malicious players. We've dedicated tens of millions of dollars and hundreds of engineers
0:37 and policy experts to building systems and tools to detect violations of our policies.
0:43 And when we find a scam, we don't just ban one ad, we ban the advertiser from ever working
0:49 with Google again. In 2011 out of the billions of ads that were submitted to us, we shut
0:55 down about 800,000 advertisers and more than 130 million ads for violating our policies.
1:01 But how do we find them?
1:03 Our first defense is automated systems that proactively catch scams by monitoring the
1:08 content of each ad and scanning the advertiser's website before the ad is ever shown. When
1:13 we detect a threat, our systems automatically seeks out and find similar websites and scams,
1:19 immediately broadening our reach.
1:21 And quite often, our experts personally review complex or suspicious ads that our software
1:27 has flagged, but which needed a real person to make a decision.
1:31 Advertisers, users and other authorities can report scams directly to our policy department,
1:36 and we'll respond as quickly as possible.
1:39 We're committed to developing the best experience for our users and remain committed to earning
1:44 your invaluable trust.
Forum discussion on Google+.