Granted, that is oversimplified, but the ease of use is definitely there. Obviously this is illegal and defrauds users, advertisers, and Google, none of which are happy when it is detected. But how do you detect it?
Detecting click fraud reminds me of an Apple Training Class I attended. The trainer said, "You'll want to read your log files on a regular basis, that way you know what to look for if something goes wrong."
I don't know about everyone else, but I instantly realized that I don't want to go reading log files in my free time.
However, the way to detect click fraud is similar. Read your stats, that way you know what to look for when something goes wrong.
The way you do that is to go through the history when you smell something fishy. AdWords, adCenter and Y!SM all provide a history of your account. That history includes spending, clicks ,CPC, CTR, etc.
If you notice an increase in clicks or in spending and thing something is up, do this:
1. Print out the 3 previous months stats. Make sure you get average CPC, CTR, ads viewed (number of ad views compared to number of clicks) and any other stats that might be relevant.
2. Print the current month. Then multiply the time that you noticed something, and take it out for a full month. For example, if you have 5 days of unexplainable activity where you are spending 4 times your normal amount, but not selling anything additional, take those 5 days and multiple them out to get a full month of stats (5 x 6 will give you 30 days, enough for averaging).
3. Compare your numbers. Take the previous 3 months and compare them to the fake month you created based on partial stats. Keep in mind that fluctuations do normally occur. You may have one or two consecutive days where your spending increases.
4. Email the PPC company. If you definitely have some strange behavior, email AdWords, adCenter, Y!SM. In your email include your research. They probably don't need your help, but this way they know why you think something is up. If you have a dedicated account rep, you might want to email this person before you even go through the hassle of calculating all of those stats.
Include in your email the amount you think you have spent that was click fraud. They probably won't come up with the same number that you did, but in the least they know what you think about the whole fraudulent situation.
If you have any tips on detecting click fraud I'd love to hear them!
