NS8 developed the EQ8 Score™ to give users a simplified view of a complex problem, understanding website traffic. The EQ8 Scoring Engine inspects and scores everything that passes through it, returning a simple number between 0 and 1000.

The EQ8 Score is applied to the quality of advertising traffic as well as the risk of fraud for a transaction. Every user that visits your store, and every action a user takes is scored by NS8’s EQ8 Scoring Engine. Negative attributes such as a deceptive session, where the user is discovered to be lying about their device or location, will push their score down. Positive attributes, such as natural mouse movements, will push their score up.

The scale below demonstrates the ranges of EQ8 Score categories.

0 through 375  (Very Bad - Bad)

NS8 identified a bot session or detected an attempt at deception. A high number of deceptive flags in a session will result in a lower score.

376 through 624 (Neutral)

A neutral score can mean that a mix of weighted good and bad attributes were discovered, or in rare cases, no attributes were discovered. Neutral is below good on the scale and should be considered as such.

625 through 1000 (Good - Excellent)

NS8 identified enough positive attributes to determine that there is a real human behind the session. Each data set includes the relevant attributes through the API for use in your service.

EQ8 Scores are built using dynamic global data sets and big data analytics in a methodology this using machine learning to update as fraud techniques change.

Scoring Attributes

Scoring is accomplished by tracking over 170 attributes, beginning when a visitor enters your website to the time they check out. 

In fact, pre-session data, such as referrers, campaign attributes, and cookies are also included in the scoring model. These attributes are combined to determine the level of suspicion of a user. 

Attributes are broken down into two main categories, user behavior and technology profile.

User behavior attributes are used to determine if a user is behaving like a human or a bot. This is determined by using attributes like navigation speed, mouse movement, session visibility, and some other patent pending technologies.

The user's technology profile reveals if they are using tactics to mask their identity, location, or otherwise act deceptively. By analyzing the information gathered, like browser profile or I.P. address data, along with other proprietary methods, the risk level of a session can be determined. 

Scoring is quite a complicated process that luckily NS8 simplifies by providing the EQ8 score and supporting data. Supporting data is summarized in attributes like a Primary User Risk Factor, Payment Risk, and is demonstrated in this article about order review. It is important to remember that although this article demonstrates order scoring, the EQ8 score can be applied to other events like advertising views, account creation, user login, and many others.

Did this answer your question?