Flashback to 3-years ago – Click Fraud and IP exclusion Still Important

As one of my many new year’s resolutions I was doing some house cleaning on my sites and stumbled upon this post I wrote back in 2007 about Google’s IP Exclusion tool.

I think it’s a topic that needs to be revisited. I’ve seen consistent increase in competition and CPC’s across my client base for the last several years and in order to be sure costs don’t increase as well, you need to be vigilant in stopping fraudulent clicks and spending on inefficient keywords.

I can’t tell you how to exactly do that because each account is unique but making sure you are on top of your analytics (bounce rate, conversion rate, CPC, cost per conversion, et al) along with proper geo-targetting and IP exclusion will go a long way staying ahead of the game.

check out my thoughts on IP exclusion

Adwords Click Fraud Fighting Tool – IP Exclusion

Google has just added a nifty tool to its Adwords tools portfolio, the ability to block IPs from seeing (and clicking) your ads. In your Adwords tools section you will see a link for ‘IP Exclusion’ which will allow you to put in IP addresses that you want to block from seeing your ads.

You should see something like this:

This is no doubt a response for the long time complaint from advertisers that “my competition is clicking on my ads to spend my money” and a welcome response in my opinion. This is not the answer many are seeking sophisticated click-fraud, but I suspect it will help make advertisers feel better and also eliminate a lot of the petty click-fraud that no doubt happens by less than ethical competitors.

There are two main reason why you should do this:

  • 1. To prevent even the temptation of clicking on your ads by your competition.
  • 2. To stop ad copy-cats.

The first one is pretty obvious, if you can help fight potential click fraud in anyway possible, Why wouldn’t you? The second reason may not be so obvious but you’d be amazed how often the competition will monitor your ads and see if you are up to anything new and copy what they like. I’ve personally seen this on ads I have written for competitive industries and its pretty annoying to say the least. Anything that I can do to keep them from seeing (and clicking) on my ads, even if only temporarily, I am more than happy to do.

The big question many people will face will be: How do I find my competition’s IP address? Off the top of my head I can think of three ways to do this. First, the most ethical way would be to analyze your log files and find IPs and hostnames that are associated with your competition and use a combination of trace routing and IP location finder applications. Here’s a free trace routing web app for you to use [link to IP tracing app.]

Secondly, use your analytics package to block a range of IPs that have not delivered quality traffic to your site. A slightly more risky way, as you may block future customers or people in a research phase of a buying cycle. I wouldn’t recommend this for most people

The third and maybe less ethical way, would be to use a little “social engineering” mixed with some web savvy-ness. You could use a blank unrelated domain and send an email and link to a few secretaries or employees of the competition, and nab their IPs, with log files, analytics software, or a small piece of code to output their IPs to a destination of your choosing. This method would probably be more accurate than the other twos but I’m sure people will have ethical questions about this type of competitive spying. I’d bet you could probably call someone in the company and ask them for their IP, which is not an uncommon way for hackers to compromise corporate networks.

However you decide to find out this information, I highly advise the use of this tool that Adwords has given us.

Read more in the Adwords Help here.

Google Releases Clickbot Clickfraud Study

According to Google

Clickbot.A is the name of a botnet that Google’s Click Quality and Security Teams investigated last year. Using our findings, we published “The Anatomy of Clickbot.A” – a detailed case study on botnet-based click fraud for the benefit of the technical research community.

Clickbot.A is an example of a botnet operator attempting a click fraud attack against syndicated search engines. Google was able to identify clicks on our advertisers’ ads that exhibited Clickbot.A-like patterns and flagged them as invalid. While Clickbot.A is a specific example of a botnet application that conducted click fraud, botnets can also be used for keylogging, distributed denial of service (DDoS), and other types of attacks.

Due to the potential for misuse and the inherent loss of control that can result from having a machine participate in such a botnet, we hope “The Anatomy of Clickbot.A” will help facilitate further collaboration between search engines, Internet Service Providers (ISPs), anti-virus vendors, and other parties on the Internet in managing botnets and similar threats.

You can download the white-paper study from the Google Adwords Blog

Google to Offer more Click Fraud protection

CNET.com is reporting that next month Google will allow advertisers to block specific IPs.

Beginning next month, Google plans to give advertisers the ability to prevent their pay-per-click ads from being shown to competitors suspected of repeatedly clicking on the ads to drive up their cost.

This will be a great tool for those who suspect their competition is clicking on their ads and costing them advertising revenue and opportunity to gain new customers. The official announcement is set to be posted on the Adwords blog later today.

Google also has an elaborate explanation of what they consider ‘Invalid Clicks’ on the adwordsa Blog.

As part of our effort to provide you with more information on invalid clicks, we wanted to give you some additional background on how our systems, processes, and teams work together to manage click fraud for our advertisers, while also sharing what the overall landscape of invalid click detection at Google looks like.

While I love that Google is taking a proactive stance on click fraud but its funny they keep implementing new click fraud preventions when they say that click fraud is only 2%.


Is Click Fraud Really Less Than 2%?

On Monday Andy Beal had a good piece about his talks with a Googler on the topic of click fraud and how they say the actual click fraud that an advertiser sees is %1 or lower.

Now in theory this is great but is it really an accurate statement? I don’t think so and here is why. I am not 100% sure that Google’s invalid click detector works flawlessly. It can’t. And won’t until Google’s sophisticated computer systems can read people’s mind. This less than %1 is the amount of clicks that pass through Google’s system inadvertently and get charge to advertisers but what about the click fraud that their system doesn’t pick-up on? That is the number that everyone wants and the same number I still believe is 10-20%. If I click on an ad with no intent on buying something from the site or performing their preferred action, how does Google know? How does Google know about the guy who runs some mesothelioma sites and war-drives for a few hours each weekend clicking on his $10-20 ads a few hundred times? Sure this is small scale but to think it doesn’t happen at the enterprise level is silly. There are a lot of unethical businesses and business people out there that will do anything to make a buck and put a hurting on their competitors.

So while Google says that click fraud is less than 2% that are misrepresenting what click fraud is to John Q. Customer.

*update* – they have made a statement correction and says now that it is the number of “invalid clicks” rather than pure “click fraud”.

BusinessWeek: The Vanishing Click-Fraud Case

BusinessWeek.com has a nice follow up to the forgotten click-fraud case that pitted Google versus a click-fraud extortionist.

To make a long story short, back in 2004, a guy named Michael Anthony Bradley rolled up to Google, showcased his click-fraud program dubbed “Google Clique.” Showing off his program that could generate millions of false clicks without being detected by Google’s anti click-fraud system (if there is such a thing). Bradley attempted to extort $150,000 from Google but was quickly arrested by officers in the next room.

Two years later, no charges have been filed and for all we know Bradley’s system could be running wild on our accounts. Why have no charges been filed you ask? Simply put, if Google pressed charges it would have to divulge secrets about its advertising to prove anything to the courts. Secrets Google is not willing to share, even to the government.

Its worth a few minutes to read the article for yourself

Why You Should Use 2nd Tier Search Engine PPC

A post over on SearchEngineJournal.com today was talking about his experience with click fraud and the second tier engines, in this case Miva. Second tier engines have a reputation of poor low converting traffic that is riddled by click fraud and this case is no different. SEJ noticed referrer sources coming from notorious spyware sources and contacted the engine with data to prove his point. While he was promptly issued a refund, there was no explanation on how this publisher got into the network and was able to display the ads. After a month there has been no response which throws up a huge red flag for me.

So you want to know the one reason why you should use 2nd tier engines like Miva, LookSmart and Findwhat? There isn’t one. Under no circumstance can I recommend using a 2nd tier engine over any of the majors (Google, Yahoo, and MSN). If you can’t get accepted to the majors and you need to use a supplementary traffic program, id rather recommend an affiliate program, but if you must use 2nd tier advertising be sure to pay explicit attention to referral reports and use as much technology as possible in hopes to identify click fraud before it gets out of hand.

Google Finally Pays Up

Reports are going around that people are finally receiving their checks from the click fraud settlement a few months back. There is a webmasterworld thread about it here but for the lazy folks, here is what people are saying

” Got my $6.45 wohhooo! headed to McDonalds for a Big Mac on google. “

” After spending $480,000 over 3 1/2 years I see a $280 credit in my account for Oct 20.

You have got to be FRIEKING KIDDING me.”

“I’ve spent $188,000 this year alone, and my payout is only $30. This is a joke. I bet I’ve had more click fraud than that in on hour. “

“Got my G award today, $300… Spent in 3 years $326,000. Go, figure! “

” Yuppers, got my $97 credit this morning on $180,000 spent during the period the suit applied to.”

“I got back 0.1% of my total spend during the relevant period. Some estimates of click fraud are as high 20% and even the most conservative estimate would be around 5%. But 0.1% is an absolute joke.”

ok enough you get the point. If you expected a financial windfall from Google, you better think twice, and hope for enough to get a happy meal!

Google Adds Invalid Click Reports but What For?

Google annouced yesterday via Adwords Blog, that there is a “new tool” in your account that will show you in real time the number of ivalid clicks that affect your account. I use the term “tool” loosely even though industry experts around the web have dubbed it “Google’s Click Fraud tool.”

While I don’t consider it a “tool” but more of an interesting peice of data. I don’t see how this “tool” in it’s current form can help manage any accounts, sure it provides an interesting piece of data but it is more of a report then a tool. What this report does is show you the number of invlaid clicks as determined by Google’s measures and the percentage of clicks that are “invalid”. The good about this report is it provides a warm fuzzy feeling to advertisers who no longer have to just take Google’s word for it when they have said ‘yes, these clicks happen but you arent charged for them.’ That is about the extent of the good as far as I can see it. To really make this report useful they should allow you to break it down to the keyword level rather then just the account and campaign levels.

By seeing invalid clicks on a keyword level basis I believe it will allow advertisers to see which words drive the most click fraud in their account. Here is an example: if the highest traffic keyword in our account gets 100 clicks per day and Google catches 10 (also the most of any keyword in the account) of them as invalid, I would presume that there is more click fraud happening on this highly competitive keyword and may want to adjust my bids accordingly.

A setup like this would really make this interesting but as it stands right now, this “tool” is exactly not a tool and mildly helpful. Google needs to break out this report with more information on keyword activity, network activity (content or search) and even time of day activity. Although there have been click fraud reports that estimate click fraud around 12-15% I firmly believe that the majority of click fraud comes from Google’s Adsense network and if they seperated this report out to search and content network I’d bet we’d see a large percentage of “invalid clicks” coming from the content network.

Why not givs us more data Google? If you really care about your advertisers and are confident in your click fraud measures, give us more data, I challenge you to!

To see this report for yourself
Log into your Google Adwords account
Click on the Reports tab
Create a new report
Select Account Performance or Campaign Performance report
Click on Advanced Options
Select Invalid Clicks and Invalid Clicks Rate
Invalid Clicks Report
Fill in the rest of the report info
Run the report
and your results should look similar to the one belowInvalid Clicks Report

Click Fraud on the Rise

Greg Sterling points out this press release by Click Forensics that documents the rise of click fraud. According to Click Forensics

Google and Yahoo are better at weeding out click fraud than smaller Web sites, but Click Forensics still concluded both companies are being hard hit. About 12.8 percent of the clicks on ads served up by Google and Yahoo are deceptive, up from 12.1 percent three months ago.

While it maybe fairly obvious or assumed by all that click fraud is on the rise, it is always nice to have some data behind it, although I believe fraud is more rampant then anyone believes.

I would love to see the data behind this such as if content network was included or if this was only search networks because I have a sneaking suspicion that content network click fraud is well higher then 14.8%.

The most amazing thing we can take from this study is that if we do the math by examining the rise in click fraud, vs the rise in PPC spend compared to last years estimate $800 million fraudulent clicks, we can see that click fraud is about a $1 Billion dollar a year industry. I wonder how many more click fraud tanker ship are sailing around out there!?