The unfortunate truth about running a website is that web hosting costs money. The ugly, behind the unfortunate truth is that we are pretty much stuck with advertisement to recover the cost. So you set up AdSense and after a while start seeing an ever increasing number of clicks paying in the single digits. Obviously you got some cheapskates hogging your website and no way of figuring out who they are. You do the obvious thing: blocking categories that sound “cheap”, only to make matters worse!
Understanding the problem
AdSense/AdWords was originally conceived as a blind auction in which advertisers would bid on keywords. Highest bid took the top slot, but when your ad was actually clicked, you were only charged one cent more than the guy who won the slot just below you.
That model had a problem. Namely that it was easy to game for brand marketeers: just offer top dollar to show the company’s logo (instead of an actual product). Nobody would click those banners, and even if, you didn’t have to actually pay what you pledged. This allowed big brands to advertise virtually for free.
In order to stop abuse, Google eventually introduced the “quality score”. QS, in a nutshell is a booster that rewards clickbait ads. The idea is simple: depending on how appealing an ad is, it can either be clicked never (0%) or every time (100%) it is shown (realistically it will be somewhere between 0% and 100%). Since Google only gets paid for clicks, they naturally want ads with a high CTR, so the QS is an incentive that makes advertising cheaper the closer you get to a 100% clickrate.
The math is reasonably simple. Let’s assume advertiser A bids $1 on a keyword and has a 10% clickrate. Advertiser B bids only $0.50 on the same keyword, but his banner has a 30% clickrate. At first it seems that advertiser A is preferable over B, but that’s only true for the single click. Factor in the measured CTR and you get: A=$1*0.1=$0.10 and B=$0.50*0.3=$0.15. In other words per 100 banners shown, you earn $0.15 from B, but only $0.10 from A. So even though A pays double, B is more lucrative in the long run and should therefore win the top slot.
Now comes the stupid part: AdWords uses dynamic bid adjustment. An advertiser only specifies what he is willing to spend at most per click. The system then automatically brings that down to be one cent cent above the second best offer. If in the previous example, A would have a 6% higher CTR, his banners would be more lucrative ($1*0.16=$0.16). He would win the bid and you’d get $0.51 per click. However, since B is winning, you are getting one cent more of whatever C is offering (in this scenario, $0.49 tops). If A, B and C are the only bidders, then C will dictate the CPC. So even if you were to write about the famed “Mesothelioma Lawyer”, C could bring down down your payout to less than 10 cents.
As a publisher, you are therefore interested in two things:
- Having an advertiser who bids more than peanuts on slot C.
- Not having an advertiser for B who has such a high CTR that he can easily trump A.
If you didn’t follow the math, don’t worry. The thing to take away from this is that you should never block ads by category. The less competition you allow, the lower your worst CPC will drop. However, you do need to weed out the shitvertisers in the ad review center on a case by case level
Case study: Mind Spark Interactive
You probably have never heard of Mindspark Interactive, but as a publisher, they are your worst enemy. Mindspark is the “toolbar” company, offering a gazillion customized versions of the ask.com (no, I won’t link to them) browser toolbar. The ask.com toolbar is easy to install, hard to remove and utterly superfluous. In a nutshell, it is a trojan horse that promises you some functionality, for the price of being allowed to reconfigure your webbrowser to use ask.com as your search provider.
Mindcraft ads are easy enough to spot. They mostly consist of one giant “Start Download” call to action button (usually green) and very little additional context. In most cases, the user has no idea what s/he is actually suppose to download and the banner can even easily be mistaken to be an element of the page it is showing on. A clueless user is highly likely to click it. Of course, that’s what drives the quality score of Mindcraft banners skyhigh and allows the company to adjust their CPC down to the peanut level.
Mindcraft Interactive plays the game of large numbers. They will buy any traffic as long as it is cheap, in order to resell it later by “infecting” the user’s browser with an unnecessary toolbar. At first this looks you you are making money, but in reality, you are getting paid peanuts to ruin your reputation as being a malware distributor.
Getting rid of Mindcraft banners is almost as difficult as getting rid of the ask.com toolbar they promote. Mindcraft has hundreds of AdWords accounts, each of them linked to its own domain and offering a toolbar with a different functionality. You can only manually ban them from your inventory. The only thing that makes the task a bit easier is that all of the websites use the same favicon. So whenever you see that orange little sun thingie in the the preview, just bring down the banhammer on the entire account in the ads review center.
The fact that they are difficult to get rid of, is be an indicator of how unwelcome their junk is on your website.
In the example above, Mindcraft will always be the B or the C, potentially even both. So regardless of whether or not their banners show, they will always hurt your income.
