“A combination of the right processes and legal technology can help law firms and companies improve patent quality.”
“If you can’t measure it, you can’t manage it.” — Peter Drucker
Companies spend considerable sums of money to develop patent portfolios that protect their valuable innovations. Given the large stakes, it behooves companies to obtain high quality patents. I’ll start this article with an example of a patent mistake that resulted in a bad outcome for the patent owner.
iRobot lost a patent infringement claim against a competitor that perhaps could have been avoided. The issue was that important concepts of the claims were not described in the patent, and the meaning of the claims was not clear. The independent claims included the phrase “instructions configured to cause a processor” but the only use of “instructions” in the patent related to operational instructions for a user.
Because the patent did not sufficiently describe the “instructions” in the claims, iRobot did not obtain its desired claim construction, and the Federal Circuit found no infringement. It seems plausible that better claim drafting might have avoided these errors and achieved a better outcome for iRobot.
To avoid outcomes like the one above, a company could measure the quality of patents in its patent portfolio and take actions to improve quality.
Some measures of patent quality may relate to the overall patent portfolio. For example, the technology areas in which a company has patent protection. Other measures of patent quality relate to individual patents to increase chances of success in patent disputes, such as the one above. While both are important, I will focus on patent quality metrics for individual patents in this article.
While there are many aspects of patent quality for individual patents, they can generally be categorized into patent scope and satisfying legal requirements.
A patent has value when it helps a company meet its business objectives. A patent that broadly covers important features of a product may prevent others from implementing that feature. Such patents may provide the owner with a competitive advantage that allows it to sell more products or charge higher prices for those products.
While the scope of a patent is clearly important, it is not easily measured for individual patents. Indicators such as the number of words in a claim or backwards citations are coarse measures that provide unreliable information.
In my view, there are no useful, objective measures relating to the scope of a patent. While artificial intelligence (AI) is evolving rapidly, the best way to determine the value of a patent’s scope is to ask a patent attorney to review it, and different patent attorneys may come to different conclusions.
Statutes and regulations set forth numerous legal requirements for patents. In some situations, violating these legal requirements may cause the patent to be invalid, unenforceable, or less valuable (e.g., narrower scope). Further, violations of legal requirements provide opportunities for defendants in court proceedings and can greatly increase legal costs.
Some legal criteria are not amenable to being measured. For example, at least with current technology, it is not possible to compute a reliable measure of whether a patent specification provides written description support for and enablement of the claims.
Some legal criteria, however, are objective and easy to measure, such as whether the claims are properly numbered. Patent Bots recently published Patent Quality Scores using several measurable legal criteria to determine a quality measure for individual patents, law firms, and patent owners.
The Patent Quality Scores are based on the following:
- Numbering errors (non-sequential or repeated claim numbers)
- Dependency errors (a method claim that depends from a system claim)
- Antecedent basis errors (no instance of “a widget” prior to “the widget”)
- Word support errors (claim words that do not appear in the detailed description)
I am using the term “legal errors” somewhat loosely here. The presence of any of the above does not necessarily render a claim invalid, but it does make it more likely that a claim will be invalid or not infringed (as illustrated by the iRobot example).
To rank law firms and companies according to the Patent Quality Scores, we downloaded an entire year of 306,738 issued patents and proofread them with an automated patent proofreading tool. By counting the number of errors in each patent and cross referencing them against the associated law firm and company, the tool computed an average quality score and ranking for law firms and companies.
Although the Patent Quality Scores measure just some aspects of patent quality, there is no doubt that eliminating the legal errors tracked by tools such as this will improve patent quality. Further, law firms that implement processes to have higher Patent Quality Scores are likely doing well with other aspects of patent quality that are not tracked by the Patent Quality Scores.
These scores may not be an accurate reflection for law firms that file a large number of patent applications drafted by others (e.g., from non-U.S. law firms or companies) and are not able to control for quality. For these law firms, the Patent Quality Scores reflect the patents they file as a whole, but these law firms may do higher quality work for the subset of patents actually drafted by them.
Reduce Error: Measure
Measuring patent quality is the first step to improving quality. A combination of the right processes and legal technology can help law firms and companies improve patent quality. An automated patent proofreading tool should be used at all stages of preparation and prosecution (before filing, with each office action response, at allowance, and after issuance). Patent attorneys should carefully review proofreading reports to ensure that identified errors are fixed. Review by a second patent attorney will also improve all aspects of patent quality.
A very small percentage of patents are ever litigated or part of a legal dispute. In those situations, however, a legal error in a patent can cause the patent owner to lose important rights to their innovations. To maximize the value of their patent portfolio, patent owners can measure the legal errors in their issued patents and take action to reduce those errors.
Image Source: Deposit Photos
Jeff O’Neill is the founder of Patent Bots, a company that provides tools for patent attorneys, such as automated patent proofreading and patent examiner statistics. Jeff is also a practicing patent attorney at O’Neill Patent Law, where he assists startups with technology relating to speech recognition and machine learning. Previously, Jeff was in-house patent counsel for Amazon, an associate for Boston law firm Wolf Greenfield, and a clerk for the First Circuit Court of Appeals.