A few tips to help you detect fraudlent clicks.

Discussion in 'Pay Per Click Advertising' started by FCM, Jan 23, 2010.

  1. #1
    Before I even start I'm just going to clarify that there is no true way to detect every single invalid click. We can only use best guess and statistical data to help us come to a conclusion.

    What is the easiest way to detect fraudulent or invalid clicks?

    There is no easy way to detect these types of clicks as we can never know
    the meaning behind a click, or the intent. However, the biggest piece of information we can use to detect fraudulent or invalid clicks is by the IP, and the IP range. Ip Address can tell us a lot about a user, where they are coming from, whats there ISP is, and other types of data we can use.

    By knowing the IP we can ask our self questions such as

    Is it normal for X to be clicking on Y at Z time? But this can only come after we have gained data from the past.

    We can also see if the user has ever clicked or viewed an advertisement before. We can also see if a specif IP range has clicked before as well. We can even use IP's to detect if a person is coming from a proxy.

    What other methods can we use?

    We can use the users interactions with the advertisements to help us detect invalid clicks.

    How? Well, if you think outside the box you can see if it was an actual "click" on the advertisement or if it was a direct hit ( which is common for bots ).

    Bots don't use mouses, nor the keyboard. Remember that.

    Is there anything else we can use?

    We are only beginning to scratch the surface, but I have to keep some things secret but this is going to really make you go wow, I can't believe I didn't think about that.

    Use the users screen size. Yes, thats right use the users screen size to help you detect invalid clicks.

    But how?

    Well its simple really, if the users innerbrowser is less than the advertisement theoretically the user couldn't have seen the site in full, nor could they have seen the advertisement in full thus making it seem as if it where an accident or fraudulent.

    We can also see if the screensize changed. The screen resolution and much more data such as java and flash but I'm not going to get into any more details just use this little bit of what I told you ( if you can figure out how to implement it ) and you should be on your way to reducing false clicks


    Hope this helps at least one person

    Jody Fitzpatrick

    If you think that your advertisements are being clicked fraudulent report it right away to the advertising network. Click fraud cost advertisers millions of dollars a year lets try to put a stop to it.
     
    FCM, Jan 23, 2010 IP