Case Study: i-mobile achieves real-time, highly accurate ad fraud detection and major time savings with Spider AF

Being able to analyze large data more efficiently and enabling highly accurate fraud detection

I-mobile, Japan’s leading mobile and PC ad network, was facing a time-consuming, manual fraud-detection process for impressions, clicks and installs across hundreds of thousands of partner sites. Extracting, processing and visualizing logs took about three days and more than eight staff-hours per cycle, so I-mobile turned to Spider AF to improve efficiency and accuracy in ad-fraud detection.

Spider AF was implemented to analyze large volumes of logs, learn from flagged cases, and perform real-time detection during ad delivery so fraudulent impressions and clicks can be excluded immediately. The Spider AF solution steadily improved detection accuracy, made dashboards usable by less experienced staff, and “almost completely eliminated” the technology-side work time that had previously taken days — with about ten team members using the system to streamline ongoing fraud patrols.

I-mobile

Kai

Chief Executive Officer