Resources/Help Topics/Analytics & Reporting

Identify suspicious clicks

Faith Hoefner
posted this on January 14, 2011 17:39

The first time a respondent clicks on a LiveBall traffic source, LiveBall records the time and date of their "first contact". An invisible <img> pixel is included at the bottom of the initial page of the creative. When LiveBall hears the request for that <img>, it updates the "last contact" timestamp for that respondent. A respondent is considered "suspicious" if they requested an initial page on a LiveBall creative, but then they never fully load the page.

There are two common explanations for this:

  1. The respondent arrived at the traffic source URL with no intention of loading the entire page -- they were (a) clicking on ads just to rack up charges for the advertiser, (b) clicking on affiliate links to rack up rewards, or (c) running an unauthorized "bot" that's scanning pages without loading all the associated resources (i.e., <img> tags). This is generally unwanted traffic.
  2. The respondent arrived at the traffic source URL "honestly," with no ill intent, but they decided to bounce before the landing page loaded completely. This can become more common in scenarios where the page takes more time to load: (a) the page has lots of Javascript or image assets, (b) the respondent has a slow Internet connection, or (c) both. This may be more pronounced in markets such as China where connection speeds can vary significantly. 


You have the ability to delete suspicious clicks from your console under the global Settings menu and the 'Manage Capacity' menu item. You can plug the IP addresses from your Suspicious Clicks report into the 'Delete by IP' field.