Analysis Methodologies

Ultimately, the value of any patent rests on the scope of rights secured and the ability to enforce those rights...But not all patents are created equal. PatentRatings, LLC develops and validates quantitative measures of the patent quality and value, as well as the relative proximity of patents and technologies.

Publications & Presentations

• "The Quality Conundrum" by Jonathan Barney from October/November 2006, Intellectual Asset Management
• "A Study of Patent Mortality Rates: Using Statistical Analysis to Rate and Value Patent Assets" by Jonathan Barney from Summer 2002 AIPLA Quarterly Journal
• IPQ® Report sample
• Due Diligence Report sample

IPQ Score

The "IPQ®" objectively scores and rates patent assets based on a proven statistical methodology. The IPQ® score is mathematically determined from the identified metrics for each patent using our patented and proprietary rating system.

The IPQ score system has a median of 100. Patent assets with higher IPQ scores are statistically more likely to generate economic returns. As a person's credit score is used by banks and lenders to determine financial risk, a patent asset's IPQ score is used by IP professionals, investors, and corporate strategic planners to determine value and quality.

IPQ® Scores are unique to each particular patent examined and are updated monthly. Scores are based on certain identified predictive variables determined to have statistically significant correlation to patent quality and value.

Typical metrics include, for example:

• Number, length and type of patent claims,
• Amount and type of prior art cited,
• Number of forward citations or references made by later-issued patents,
• Presence or absence of limiting claim language,
• Patent prosecution history
• 50 other relevant internal and external factors

Market Validation of the Value of IPQ Scores

A recent independent blind study conducted in partnership with a major Fortune-100 company determined that IPQ scores were statistically correlated to the probability of rated patents producing economic benefit (a licensed or commercialized patent) or no economic benefit (non-licensed, non-commercialized).

The obvious implication of the study (see graph below-left) is that IPQ scores can be effective at helping companies identify and leverage economically significant patents or technologies while avoiding or culling those that are less promising.

Class Percentile Rank

Patents are ranked on a percentile-basis according to IPQ score. The percentile rank represents the proportion of patents in a given population that have IPQ scores less than or equal to the reported score. For instance, if a patent receives an IPQ score of 125 and this is greater than or equal to 90% of other patents in a given population, and then the patents' percentile rank would be 90% relative to that population.

Our search tool and reports rank patents according to five different populations or groupings:

• Overall - all patents
• "Field" - same field
• "Class" - same class
• "Subclass" - same class/subclass
• "Assignee" - same assignee

Rating

Letter grades are assigned to each patent based on percentile rankings of IPQ scores among issued patents.

Relevance Rating Score

The 'relevance' score is the probability that there is a citation of the patent of interest. Calculated on a scale of 0 to 1, the score provides the first objective, statistically accurate measure of relevance. Highly related documents will have scores close to 1, while non-related documents will have scores close to 0.

Using advanced statistical algorithms that analyze all citation records from the USPTO, we are able to determine the probability that the documents are related.