Overview of ROL Group Patent Ranking Metrics

When analyzing a large patent portfolio it is often difficult to determine where to spend diligence dollars first. In order to direct us to the specific assets that are more likely to be interesting we developed an in house heuristic ranking system. It is based on static patent characteristics. The image above describes the specific component scores for each of five characteristics. We used this system to rank the assets in Intellectual Venture's patent portfolio (IAM Magazine, Issue 66 July/August 2014).

These component scores are then summed to create a raw rank (max 100) and bucketed into groups of interest. High > 85, Medium > 75 and everything else is rated as Low.  It should be noted that theses rankings are intended for the general grouping of patents by likelihood of value, not for direct patent-to-patent comparison. It is possible for a specific low ranking asset to be of more interest and value than a high ranking one.

We have used this method across projects for over five years and have found that when compared to both alternative (more complex) scoring systems and initial human review (< 1 min per asset) of 1000+ asset portfolios it yields similar, though not identical, groupings. Overall, this ranking method has given us a very high return while maintaining simplicity.