The Click Forensics scoring engine uses multiple inputs to generate a specific output, the ClickScore. ClickScores can be aggregated across different dimensions, often by source or destination categories. The inputs required to produce a click score fall into three main categories:

Attributes: technical and economic attributes of every click include IP address, proxy server traffic, bot characteristics, originating location, HTTP header characteristics, visitor profile characteristics and cost-per-click. Attributes such as IP address, UserAgent, and source of click (referrer) are captured directly from the click log and combined with derived attributes such as geo-location, demographic data, and site reputation.

Behavioral: click-through behavior such as visit depth, time on site, viewing patterns, time between clicks, visit frequency and conversion information.

Community: comparative data is available across a community of advertisers, publishers, and ad network for validation. Community-based information includes known offenders, visit patterns or anomalies that span multiple sites.