My previous article showed how expensive patents are – from $220,000 to $440,000 to file and maintain a patent in the United States and 9 other major industrial countries. I also discussed one experiment that showed that data driven patent analytics could predict manually determined patent maintenance fee payment decisions over 70% of the time while dramatically reducing the time and cost to conduct the manual review. In this current article I describe best practice Patent Maintenance Fee Payment Decision Processes.
IPVision has worked with many companies to implement evidence-based data driven processes 
for the cost-effective evaluation and management of patents. The specific processes that are adopted differ somewhat in the details because companies differ in their culture and structure and because of their differing industry technology and competitive environments.
Although the specific processes differ there are core components in each of these programs. The following is a step-by-step compendium of the best practices I have seen.
Step 1: Determine Renewal Decision Period
I have found that it is most efficient to select a “Patent Renewal Decision Period” – i.e. period of time during which some of your patents require payment of U.S. maintenance fees. This period should start sufficiently in the future to enable you to run the evaluation process before the period commences. For on-going Patent Payment Decision Processes it seems to work best to evaluate patents at least one calendar quarter in advance.
Step 2: Gather the Patents for Analysis
U.S. Patent Maintenance Fee Windows where no surcharge fees are required are as follows:
- 3 to 3.5 years after the patent issuance date
- 7 to 7.5 years after the patent issuance date
- 11 to 11.5 years after the patent issuance date
Search for your patents that have 3.5 year, 7.5 year and 11.5 year issuance anniversary dates in the Patent Renewal Decision Period – this will be your “Analysis Set”.
NOTE: At this point some companies separate the Analysis Set into two groups: (1) the 3.5 Year Group and (2) the “Later Age Group” – i.e. the 7.5 year patents and the 11.5 year patents. The reason for this is that the original “business/technology thesis” has probably not changed much for the patents that are only 3.5 years old whereas the business/technology bases are more likely to have changed for the Later Age Group. Also the maintenance fee for a 3.5 year patent is substantially less than for a Later Age Group patent. As a result some companies do only a cursory analysis of the 3.5 year patents, especially if they are using a manual review process. Because of the efficiency and cost-effectiveness of algorithmically based IPVision patent analytics I recommend to our clients that they run the same process on all age groups.
The goal of the next two steps is to identify whether a patent is strong or weak and whether is it “important”. Understanding these key components provides a strong base for decision making.
Step 3: Analyze the Patent Claims
The goal of this step is to identify the claims strength of the patents under review. IPVision Claims Analytics rate the independent claims of a patent for “broadness” and for “structural issues”, - i.e. claim structure that may limit the enforceability or ability to license a patent. IPVision Claims Analytics uses an expert system and rules developed in coordination with experienced patent litigators and patent licensing experts.
NOTE: People often ask - Why there would be “structural issues” in patent claims? The reason is that in many, if not most, cases the attorney who prosecutes the patent application is focused on getting a patent issued. Prosecution Attorneys very rarely also litigate patents and do not see how claims are interpreted and applied in litigation and licensing situations. IPVision Patent Claims Analytics look at a patent from the litigation and licensing perspective.
Step 4: Conduct a Patent Citation Analysis
Patent citations are an important consideration in the patent renewal decision process. In order to obtain a patent the inventor must demonstrate that the invention is “novel” meaning it is “new” and has not been invented before. The patent applicant must cite all “prior art” of which he or she is aware that is relevant to the determination of whether the claimed invention is novel. Prior art includes patents as well as publications. A patent can be invalidated if the applicant fails to cite prior art of which he or she is aware – this has been referred to as “inequitable conduct” or “fraud on the patent office”. Bottom line – patent citations mean something. Research has shown that there is a correlation between the number of patent citations and the importance of the patent cited.
IPVision algorithms use the concept of “relative citation frequency” to evaluate the importance of a patent. This approach looks at the citations of a patent relative to its “technology peers” of similar “age”.
Step 5: Review Claims and Citation Results
I recommend visualizing the results of Step 3 and Step 4 in a two by two matrix form. This visualization makes cross-discipline discussions easier because it presents the information in a more intuitive form.
- Valuable Patents = Strong Claims and High Relative Citation Frequency
- Little Value = Weak Claims and Low Relative Citation Frequency
- Potentially Valuable = Strong Claims but Low Citation. These patents could be like a “key in a lock” – narrow but vital
- ???? = Weak Claims and High Citation. Not clear what this means. It may depend on whether it is part of a larger family of patents.
Step 5A Optional: Consider the Entire Patent Family
For a fuller picture of the IP situation I recommend that the entire patent families of each of the renewal patents be assembled at this point. The metrics described above can be run on each family member and can be used to assess the overall value of the entire patent family. These results provide additional insights to Step 6. NOTE: If the company uses a Stage Gate approach it is less likely to consider the entire patent family before a first pass at Step 6.
Step 6: Review Results from Business Perspective
The Patent Evaluation process of Steps 1 to 5 characterize important attributes of the patents up for review. It is an evidence-based, repeatable process. What it does NOT do is make business decisions. A few examples will help frame this:
- Valuable Patents. You might say that the renewal fees should automatically be paid on these. That may be but for what business purpose? Do these patents support our current or planned strategies? If so, great. If not, what should we do with these - perhaps these are candidates for sale or out-licensing. What does the larger patent family look like?
- Little Value Patents. You might say that you should not pay renewal fees on these, if they are Weak and have Low Citation then why keep them? One reason is that they may be part of a larger patent family that collectively is strong. Dropping the “Little Value Patents” might weaken the entire family from a licensing or sale viewpoint. On the other hand if the entire family is large and strong we might save some money by letting the weak ones go.
Bringing a business perspective to bear is crucial to an effective patent management program. Using Patent Analytics and the approaches described above can cost-effectively frame the business issues with evidence-based repeatable metrics.
A valuable outcome of the process described is that it can provide IP insights to the company’s business and technical people in the context of specific “digestible” cases. Cumulatively over time this can raise the IP intelligence of the entire company.
Author: Joseph G. Hadzima, Jr. | Senior Lecturer, MIT Sloan School of Management | Co-Founder & President of IPVision, Inc.
