Identifying optimal patent prosecution strategies requires capitalizing on prosecution experience and understanding an applicant’s patent-related business goals. More specifically, the strategies should be ones likely to result in the desired outcome. There is seldom a one-size-fits-all strategy, as applicants vary markedly with respect to, for example: how important it is for applications to quickly issue as patents; the amount of budget that can be devoted to a particular matter (during a particular time or overall); a scope of desired protection; and technology-area focuses. To illustrate, a strategy that is associated with fast allowances may be advantageous for an applicant seeking quick patent protection and disadvantageous for an application seeking long patent term adjustment.
Additionally, prosecution strategies may be as diverse as applicant objectives. For example, prosecution strategies may involve determining an application focus, a number and type of claims, whether to use one of many available USPTO prosecution programs, how to interact with an examiner, whether to initiate an appeal cycle, and so on. Practitioners may accumulate experience with particular strategies to determine whether they are likely to result in specific outcomes (e.g., an allowance while preserving broad claim scope). However, such experiences are typically built on small sample sizes and underrepresent the total strategy-outcome space. The picture becomes even more complicated upon considering that outcomes depend, not only on strategy, but also on multiple, external and dynamic factors, such as which examiner was assigned to an application and what case law is being established.
Recently, big data has afforded practitioners with the ability to conduct a client-tailored, data-specific analysis to identify optimal prosecution strategies. Thus, practitioners need not any longer rely merely on their own experience or that of their professional networks. Rather, practitioners can benefit from the aggregated, filtered and processed data stemming from past prosecution of all published patent-application matters.
As a first simple example, LexisNexis PatentAdvisor® provides examiner-specific data that identifies the examiner’s allowance rates, average number of office actions issued per patent (or abandonment), typical post-RCE response, probabilities of responding to appeal briefs in various manners (e.g., issuing an Examiner’s Answer), and so on. A practitioner may use this data to predict the examiner’s response to various types of actions (e.g., an RCE filling, an after-final response, or a Notice of Appeal filing) and to recommend a course of action to meet a client goal (e.g., to attempt to secure an allowance as quickly as possible, with the desired claim scope, with the fewest number of office actions, or other client-desired outcome).
As a second example, PatentAdvisor™ also provides data that may be used to conduct a portfolio-level analysis. For example, applications in the portfolio that are failing to achieve desired results (e.g., that are receiving many office actions or that are resulting in abandonment) may be identified, so as to find which characteristics are predictive of the poor outcomes. To illustrate, characteristics may include use of a particular USPTO program, a technology focus, use of an appeal strategy, a low-allowance-rate examiner, etc. Subsequently, the portfolio may be managed in a way to bias against or avoid those characteristic(s) if they are controllable or to set up alerts to allow for an intelligent response to detection of any non-controllable characteristics.
As a third example, big prosecution data may be used to generally assess prosecution strategies. An applicant’s portfolio may not be sufficiently large or diverse so as to be able to predict outcomes of particular strategies. For example, an applicant may not have previously filed a Track-1 application or a request for a Pre-Appeal Brief Conference. Such data may also be unavailable or small for a given examiner (e.g., due to a lack of examiner assignment or an examiner’s lack of experience with the approach). In these circumstances, it may be advantageous to consider data from a broader data set. A statistical analysis can be performed so as to explore probability distributions and significant dependencies for a particular prosecution technique. The analysis can be tailored to consider pertinent constraints, such as data from a recent time period and/or art unit of interest. Accordingly, rather than a guess-and-check approach in an attempt to sample a prosecution-strategy space, big data can allow practitioners and applicants to engage in a selective and targeted course of action.
In sum, the big data that is now available from LexisNexis PatentAdvisor provides users with the benefit of “experience” that no individual attorney—or even law firm—could have. Further, the statistics generated based on the data facilitate generation of objective predictions as to how effective use of various strategies would be in achieving particular goals. Integrating this type of number-based analysis into the daily work of practitioners should improve the ability to provide advice to clients as to how to most efficiently and effectively reach their patent goals.
Kate Gaudry, Ph.D., is an associate in Kilpatrick Townsend & Stockton LLP’s Washington, D.C., office.