Human behavior is an interesting subject. Ask a behavioral psychologist or an FBI profiler what they can tell you about the people they routinely observe and you’ll get an answer that sounds something like “People are walking, talking volumes of information. The more they talk and act, the more we know about their behavior and the better we can predict where they’ll go and what they’ll do next.”

Evidently, Google knows this as well.

At the end of November, the folks from Mountain View, California picked up U.S. Patent #8065296 to automate and expedite the process of evaluating search results based on “observed” behavior and allowing Google to return a greater level of relevant search results. Since Google originally filed for the patent in 2004, it’s a fair assumption that this behavioral evaluation process is something that they’ve been working on for quite some time and it is highly likely that it will allow Google to also return advertising results in the same manner.

The reason for this patent? According to the documents filed at the USPTO, Google asserts that a better model for predicting likely search results is needed “because the amount of information on the web and the number of new users inexperienced at web searching are growing rapidly.”

Google goes on to explain. “It may be desirable to monitor the quality of the search results provided to users in order to notice general trends of improving or declining quality and to identify specific problems that might suddenly cause a drop in quality. Manual evaluation of search quality can be laborious and time consuming, typically allowing for a small number of searches and search results to be evaluated to determine overall quality of a search engine.”

So what are we talking about when we speak of behavioral quality signals? In short, behavioral signals are those signals generated by human interaction on the web. It sounds simple, but the patent document muddies the issue:

“A method comprising: providing items during a time period; and determining an indication of quality of the items provided during the time period using a time series model; and taking remedial measures when the indication of quality of the items is below an expectation, wherein the remedial measures comprise: automatically removing a modification associated with the providing; wherein the items include search results.”

To make this easy to understand, I’m going to give you a dead-simple example.

Let’s say your grocery store manager tells you where to find the high-fiber cereal down the massively-long aisle with all the cereal boxes. He stays on to watch you make your selection, noting as you pull one box off the shelf, quickly scan the package and stuff it back. He makes a mental note that you only spent about two seconds with the first box. He then observes you as you select another box of high-fiber cereal and spend a full two minutes reading the product information on the back of the box before putting it in your basket. Again, he makes a mental note and comes to the conclusion that, for a person such as yourself who wants high-fiber cereal, box two was a more favorable brand. Then he waits around the cereal aisle to see if anyone else fitting your description does the same. In the end, his hopes are to be able to move the preferred cereal boxes to a more high-profile location along the cereal box aisle, allowing high-fiber cereal enthusiasts to save a little time in their shopping.

Given the example, it’s easy to conclude that behavioral quality signals are a complex mechanism used to predict complex behavior. Every behavioral style is unique. Every search result is therefore unique. It’s well known that Google has been using panels of human testers to judge the quality of their search results and assess whether one algorithm tweak resulted in better or worse search results than another (and here they’re talking about measuring the quality of the search engine results pages – rather than an individual page). Bing, the big competitor on the block, also uses quality signals in ranking. But the human testing component is a laborious effort, not practical. The patent is for an automated process. If Google is looking to improve their speed to market, automation would be enormously helpful.

So why is Google jumping into the game only now? The answer: Remember, the patent for this process was applied for in 2004, so it’s been a long time coming. Google has been waiting on this patent for some time. For Google, it’s another tool to allow them to deliver more relevant search results than ever before. For web marketers and site owners, it’s a clarion call for quality content.