Fumflockerkin - Breaking Open the Google Black Box

7 comments

QuadsZilla over on SEOBlackhat has outlined a nice little plan of action on how to get a handle on what Google is really doing, and end all the mystery surrounding Adsense, Smart Pricing, Adwords, and the quality score.

Fumflockerkin - The PPC Study the World Needs

1. Create 25+ imaginary keywords like “fumflockerkin” (But not fumflockerkin)
2. Set up 5 adword and 5 adsense accounts under 10 different names
3. Create a series of landing pages for the keyword
4. Track what happens when you bid against yourself.
5. Track what happens with quality score for various settings. (CTR, conversions, page layouts)
6. Track how much Google is really taking.

As he notes the plan is without risks, however I wonder if you might minimize some of that by spreading it over few different people and not just one, as long as they shared the data. I'm sure publishing this data has to be violation of the TOS, but it does make you think now doesn't it.

Comments

#6

Done this experiment before and pretty much know the answer to most of those questions, esp #6, but it would be nice to see someone else produce data for comparative study.

I do not think this will

I do not think this will work as there will not be a real data set of users ctr/etc running through it.

Also one thing this test lacks is the historical aspect of an account as well as the historical keyword performance across google.

Good in theory but I think it will be a worthless test. I think it would be better to apply it on an industry you are already in and know where you once performed and where you currently do not now. Not only that it would actually lead to diagnosing a formula that works for you, plus you could capture leads or sales in the process.

My $0.02 anyway.

Hot and Not

Good point I think testing for real words like "car insurance" in conjunction with with fake words might give some interesting comparisons too.

Meh

Create one ad for a very off season ticket industry related ticket. Such as Houston Rodeo right now. NO ONE even thinks about Houston Rodeo until February. You can have three of the top 10 listings for it in all the three SEs and you'll get maybe 5 or 10 clicks a month. It's that cyclical.

Yet, you'll notice something strange: Your PPCs, if you're in the top 5 or so, will go through the roof, and sometimes at *really* round numbers ;p In August when I tried this, I got 1500 clicks (exactly) for Houston Rodeo, on an undisclosed engine. All three are affected (i'm a member of overture, msn adcenter, and adwords) but one has gotten dramatically worse since January 2006. Of course, of all these 1500 clicks, not a single damn one of them downloaded the CSS stylesheets; a clear give-away that a bot was utilized.

also you probably want to

also you probably want to make sure that
#1 - you do not have the google toolbar installed
#2 - you do not log into the accounts from the same ip
#3 - watch your a**

I think you would want the TB

I think you would want the toolbar installed if you were to fake the traffic though...

a clear give-away that a bot was utilized

By way of an expansion it might be worth setting up a similar project or several to tie some hard numbers to the actual amount of click fraud (or, at the very least, useless money draining clicks) around.

Not to mention the fact that such a testified report might qualify for superb linkbait, of course. :)

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.