It's a pretty long story (read it here). But here's the short version for adsense publishers: An enemy of mine clickbombed the ads on my site between April 12 and 14. On April 18, I got this dreadful "Adsense Account Is Disabled" email from Adsense Team. It said there were invalid clicks or impressions on my adsense account; therefore, to protect advertisers, Google had to ban my account. It's fortunate for me. I noticed abnormally high CTR on April 13 and immediately informed Adsense Support about it. I stated categorically that somebody else was clickbombing the ads in order to make my adsense account banned. I asked them to ban the IPs. After a several hours, I got a response with ticket number from adsense staff that he would monitor the situation closely. Ok, on the 18th i didn't believe that my account was banned. I had reported the suspicious clicks on the ads. It's apparent that Google teams work loosely with no communication to the others at all. I appealed through this Invalid-Click Appeal Form. The most important thing I mentioned in the form was the support ticket number in which I reported suspicious clicks and suspicious IPs in server log. Adsense team reviewed my case and reinstated my account today! The bottom line: when you see abnormally high CTR, report to adsense team immediately. If your account was to be banned, you would have good chance to get reinstated.
I digged up 700MB server log to find the suspcious IPs used to generate invalid clicks and how it was done. "It's impossible!", I thought. "This huge 700MB server log doesn't record adsense clicks." Lucky for me the server log recorded adsense searches, as I use custom results page on my site. CTR was 50% on adsense search; therefore, the offender had to search repeatedly. I coded my own Log Analyzer from scratch to find who repeatedly searched. Dang! Jackpot. Some of offending IPs showed up in the small list. After further linking related IPs, I came up with a list of suspicious IPs and saw how the bad guy click on the ads. Apparently he visited some pages and clickbombed, searched and clickbombed from 4 networks and 12 IPs in total. So my whole point is you can get suspicious IPs from server log. It helps to include the IPs in your appeal form. The day you're banned, download a copy of your server log from hosting control panel. If you delay, the server might prune the log and you will lose all the evidences. My log analyzer in action.
Can't say "be careful next time", because it's not totally in my control. I can ban the IPs but that guy will use proxies and his other networks to clickbomb. Now I'd like to shout out to Adsense Engineers reading this. You should implement a warning system before a serious decision is made. I know I've agreed that you guys can ban me without prior notice without any reason, but your punishment system has disrupted my business and insulted me, as an honest, long-standing publisher. And you may say "well, you're just one out of millions of publishers. You don't make any difference." I say i'm one of thousands of honest publishers who got insulted. And the day you treat more of us as being indifference, you'll see how easy it is for us to just switch. Because Adsense doesn't provide us the protection against click-fraud, I suggest you install AdLogger to limit the number of clicks per visitor or send warning email to you when a visitor repeatedly click the ads. I wonder why a multi-billionaire company like Google can't provide any protection for publishers. It's a real shame, when you consider that a 20-year-old student with little money managed to create AdLogger to protect adsense publishers, and Google with billion dollars doesn't care. It's a shame.
I couldn't agree with you more on this statement, Google does NOTHING to protect us and it's a sad shame. What log analyzer was that you showed? I'm interested in getting it for my sites.
it's too easy to screw with adsense publishers anybody can steal your adsense code and put it on a spammy site and get you banned
Yeah correct and Google has done NOTHING to protect us. Why don't Google team add a textbox that we can enter our websites, so that the ads won't display on any other site besides ours. I don't see any implication arise from doing this. And as a web programmer, i don't see any slight difficulty in coding this functionality. It could be done in no more than two days or no more than a week for testing.
But, that wouldn't really help prevent click fraud. G needs a system that can detect a clickbomb (or some other attack) and turn ads off for a set amount of time for that IP. Then if it kept reoccuring (using proxies or whatever) disable ads for X amount of time to that site. We should not be responsible because they can't code a system that protects their ads, then blame US when we get click bombed!
Yes, it's not clickfraud prevention. It however prevents your adsense ads from showing on offending pages that are not under your control. We know it's against the adsense TOS if your ads appear on pages showing adult or violent contents. It's another way to get your account banned. The bad guy would put your adsense code (with your pub-id) in offending pages. It's a silent killer. You don't know it since adsense doesn't report it to you. Even you know it, it's not under your control. That's the point. If you could restrict your adsense to your sites only. It's totally under your control. For clickfraud prevention, I had suggested that Google implement a warning system. Instead of banning publishers on sight, Google would first give them two weeks or so to solve or appeal the issue of suspicious activities at back-end, using state-of-the-art ticket system. The ads should be running normally but the payment is forced on-hold until the issue is solved. Doing this ensures that our adsense revenue is unaffected during the resolution period. Ok, if any publishers can't resolve the issue on time, Google can ban them. It wouldn't be late to refund the advertisers, because the payment has been forced on-hold. I believe this will protect good publishers from ever getting banned. It may turn bad publishers into good publishers. And it will keep click fraudsters away from the system for good. This warning system may come with implications. Firstly, the real clickfraudsters will have multiple chances in designing a fool-proof clickfraud system. With the chances, they will eventually have found a sophisticated way to fool adsense clickfraud algorithm. Ok, that is when Google has to improve the smartpricing algorithm. Make smartpricing works on IP level. What makes invalid clicks different from valid clicks is that the invalids have 0% conversion rate. So as soon as an IP converts too close to 0%, just make its click worth $0. If needed, Google should deduct revenue generated from the blacklisted IP and refund to advertisers accordingly.
Wow you sure sound like you know what your talking about IMHO - I don't think any of the blame for invalid clicks should come down to the publisher - I *believe* G should just wise up and make a system that couldn't be so easily outdone with clickbombing attacked, etc. Any invalid clicks - Ban the IP from displaying ads on that site - problem solved. If they are using proxies or another way - block ads on the page or website. It should NEVER come down to a full account ban, unless G can't and won't take the notion that YES theirs immature people on the web looking to get some people banned. (of course there is exceptions - pubs clicking on their own ads 100 times, and the usual things...) I just don't see why they haven't - and I don't think they will anytime in the future... which just SUCKS...