Ad Traffic Quality Resource Center

You may have heard the terms 'click fraud' and 'invalid clicks' and wondered how or if they're different. Here's a definition of each:

Click fraud refers to clicks generated with malicious or fraudulent intent.

Invalid clicks are clicks on AdWords ads that Google suspects may constitute click fraud. Advertisers are not charged for these clicks.

There are two primary incentives for committing click fraud:

• AdWords advertisers may try to attack competitors by raising their costs or exhausting their budget early in the day.
• AdSense publishers may click ads appearing on their own websites in order to inflate revenue.

Here are the primary methods for carrying out click fraud attacks:

• Manual clicking
• Click farms (hiring individuals to manually click ads)
• Pay-to-click sites (pyramid schemes created by publishers)
• Click bots (software to automate clicking)
• Botnets (hijacked computers utilized by click bots)

*Google builds in a high false positive rate by marking a much larger number of clicks as invalid than are believed to be generated with bad intent.

Note: Not to Scale

At a basic level, we protect advertisers against click fraud by not charging for suspicious clicks. The intent of a click is difficult to determine with a high degree of scientific accuracy. We therefore create a high false positive rate by marking a much larger number of clicks as invalid compared to the number of clicks we believe to be generated with bad intent.

Undoubtedly, there will be some instances of click fraud which will go undetected by our proactive systems and processes. By casting the net of invalid clicks wide enough, however, we are able to effectively minimize that proportion.

At a high level, we detect invalid clicks utilizing a three-part system:

Proactive detection happens before advertisers contact us. Our goal is to proactively detect as many invalid clicks as possible. Despite the media attention on click fraud in 2006, our rate of reactively detected invalid clicks remained extremely low (less than 0.02% of all clicks), even as advertisers sent us more cases. This low rate can be attributed to the significant number of invalid clicks we detect proactively.

Our filters are particularly apt at detecting invalid clicks due to their flexibility. Rather than ignoring a fixed percentage of clicks, they react to various traffic patterns and indications of click fraud attacks. Real-time monitoring and filtering means our invalid click rate is constantly fluctuating. On average, invalid clicks account for less than 10% of all clicks. At our current revenue run rate, the number of clicks we filter and therefore do not charge our advertisers for is potentially worth hundreds of millions of dollars.

How can advertisers protect themselves against click fraud?

Reactively detected invalid clicks on AdWords ads are rare. However, we recommend taking advantage of ROI-monitoring techniques to check for undetected click fraud as well. In particular, make sure that you do the following:

1. Track campaign performance on the most granular level possible, preferably by keyword
2. Use Google Analytics and auto-tagging for any click data analysis
3. Become familiar with click fraud false positives
4. Report any suspicious activity for investigation

Most importantly, advertisers should have a working definition of ROI that is trackable. For advertisers that are not selling products directly online, try to define some intermediate step (e.g. filling out a form) as a proxy for a conversion.

Read more about our Ad Traffic Quality team.