We recently spoke with Marius Domokos, an IP attorney specializing in strategic counseling for technology companies, who also worked as an in-house IP attorney for KLA-Tencor and Intel. Here Marius shares some thoughts on how the MEMS industry should approach IP protection.
MEMS Investor Journal: What are some of the biggest mistakes companies make that can jeopardize their IP?
Marius Domokos: Many companies equate IP with patents and dedicate most of their IP efforts towards initiating or expanding a patent program. There are three common mistakes to this.
The first common mistake I see is that the scope of the patents tends to track existing technology too closely and the patent portfolios are not optimized to support the business and commercial needs of the companies as they participate in the global economy.
A second error is the failure of companies developing cutting edge MEMS designs to realize that trade secrets are at least as important as patents, and therefore their protection deserves significant thought and effort, from the corporate formation stage through technology commercialization.
A third mistake occurs when companies that have developed excellent IP portfolios experience leakage of IP rights directly through overbroad outbound license agreements, or indirectly in connection with commercial relationships.
Shortly put, companies need to develop IP portfolios aligned with their business strategies and then must be careful not to dilute their IP rights through contractual frameworks.
MEMS Investor Journal: What steps should companies take at the outset to minimize exposure to third party IP infringement challenges?
Marius Domokos: Assertions of IP infringement made against an early stage company could be devastating both in terms of financial drain, distraction of management, diminished valuation and decreased appeal for investors.
First, a startup company should ensure that the founders and other key personnel hired subsequently do not bring to the company any clouds of uncertainty with respect to IP ownership (e.g., the founders did not work on the new company while still employed in their previous jobs).
Second, a MEMS company bringing to market a new design should be sophisticated in terms of defensive patent infringement analysis. This means that the company should consider the issue, but should not overreact unless it can articulate specific goals for conducting such analysis (e.g., designs-arounds are rarely feasible at the MEMS design level). As a general rule, I discourage startup companies in the semiconductor space from conducting patent landscape surveys (whether internally or by engaging outside counsel), unless specific and persuasive reasons for doing that exist.
Third, a MEMS startup that will rely on an external foundry for manufacturing or will enter in a commercial relationship with a larger company for the development or commercialization of the startup’s MEMS technology should devote significant attention to IP indemnification, liability allocation, and IP licenses.
MEMS Investor Journal: How should a MEMS startup company’s approach to IP be different from what a large and established company would do? What are the specific differences?
Marius Domokos: A high number of large companies do not appear to have a cohesive patent strategy – they seem to just file as many patents as they can fit in the allocated budget. Statistically, a number of these large companies will experience negative outcomes (e.g., progressive erosion in relative IP balance compared to better managed industry peers, increasing inefficiencies in patent portfolio costs, etc.). Fortunately for them, most large companies are able to absorb the losses associated with such outcomes, and often do not even correlate such losses with past deficiencies in IP strategy.
In contrast, a startup MEMS company will generally be faced with a limited IP budget and high costs of technology development, and cannot afford significant missteps. MEMS startups should define crisp IP strategies from the beginning, and then execute these strategies against systematic milestones and quantifiable performance metrics.
To define an IP strategy, a MEMS startup should consider the following types of issues:
(a) the relative balance between patent coverage and trade secret protection of the company’s technology; The answer to this question will depend on how easily the company’s MEMS designs can be reverse engineered from end products, the budged available for the patent portfolio (both currently and in the future, given the inevitable up-trending in patent related expenses), and the expected leakage of trade secrets in connection with foundry manufacturers, employee turnover (especially in countries with less efficient trade secret protection legislation) and technology development engagements with third parties.
(b) the expected evolution of the company; For example, if the company pursues an IPO strategy, the company should envision itself in the future as a stand-alone entity in the industry, in which case a comprehensive patent portfolio that takes into account relative patent-revenue metrics against anticipated competitors would be paramount. In contrast, if the company expects to be acquired by a larger entity in the short- or mid-term, the patent portfolio could track the technology of the company a bit closer, should focus on being complementary to existing patent portfolios of potential acquirers, and should include an offensive component directed at potential acquirers. If a MEMS startup contemplates the need for additional capital raising, the patent filing strategy should be modulated to provide a positive impression to potential investors, but without overspending (especially given that VC firms have become very sophisticated in not overestimating the value of patents).
(c) the structure of the patent portfolio; For many companies in the semiconductor space, depending on their specific technology and business direction, I tend to suggest a multiple-stage patent strategy, beginning with a defensive approach centered around the company’s technology, then adding offensive prongs that target existing and prospective competitors, and then expanding to target specific entities in other layers of the commercial chain. This strategy includes additional dimensions, including the distribution of foreign patent filing (which will vary for each of the phases described above), optimization for anticipated participation in SIGs and Standard Bodies, patent purchasing and licensing components for both defensive and offensive purposes, aggressive and continuous patent portfolio pruning to remove low-yield patent assets, and systematic public disclosures to produce prior art against threatening external patents
MEMS Investor Journal: Specific to MEMS what has been the trend in IP litigation – up or down?
Marius Domokos: I am not aware of any specific statistics regarding IP litigation in the MEMS space. Patent litigation tends to be somewhat countercyclical with the main economic indicators, although this statement is probably less accurate these days than it used to be 10 years ago.
As a general trend, it is clear that patent assertions have increased in the MEMS space, from the design level to the manufacturing stage, although most of these assertions do not actually end up litigated. Many semiconductor companies, universities and other entities now have aggressive projects for monetizing their patent portfolios, and startup companies in the MEMS space are certainly receiving their share of written cease and desist letters that are precursors to licensing discussions. The rate of such patent assertions, and the amount of money spent by MEMS startups on settling them, will likely remain high for the predictable future as part of a broader trend in the semiconductor industry. In some of these cases, however, patent assertions evolve into more sophisticated technology transfer transactions or into acquisition discussions, with positive value for both parties.
MEMS Investor Journal: What costs are involved in taking an IP complaint through trial and what are the main drivers of these costs?
Marius Domokos: The characteristics and costs of litigation involving actions for misappropriation of trade secrets, copyright infringement and patent infringement differ significantly relative to each other.
Copyright litigation tends to be more predictable, although cases involving copying of designs or software may involve significant expert testimony costs, and the technical analysis necessary to evaluate the scope and extent of copying could result in escalating expenses.
In patent litigation, the costs will generally be higher for a defendant, and will tend to be proportional to the number of patents asserted, the complexity of the technology allegedly infringing, and the scope of the defensive invalidity and noninfringement analysis that is performed. When the defendant counter-asserts patents, the costs also escalate for the original plaintiff. Further, when litigation expands to include foreign jurisdictions, the costs will experience another significant jump. The main costs in patent litigation are attributed to analysis of patents and noninfringement positions, expert testimony and opinions, preparation of witnesses for depositions, conducting depositions of adverse parties, and analysis and briefs submitted in connection with claim construction in Markman hearings. When cases progress to trial, significant costs arise in connection with trial preparation by the litigation team, and sometime even mock trials, jury consultants and other sophisticated procedures for assessment of probability of success and fine-tuning of litigation strategies. The disruption of internal business for companies defending against patent assertions is something that is often not quantified and taken into account, but is nevertheless a major cost in terms of diversion of management’s attention away from the company’s business. As a matter of fact, strategic patent infringement lawsuits are sometimes brought exactly for such secondary reasons, putting forth only a colorable case of infringement.
Trade secret litigation is inherently unpredictable, and MEMS companies should always spend some time thinking through the risks when constructing a legal framework for any technology-centric engagement. At one end of the spectrum, there are easier and less costly cases, where the offending acts are relatively clear and the defenses are unconvincing (e.g., the cases of former employees who are found to be in possession of written confidential documents taken from their previous employer to be used for the benefit of a competitor). At the other end of the spectrum, there are cases where the circumstances suggest some possible improper leakage of trade secrets, but the technology involved is complex and the defenses are plausible (e.g., an allegation of trade secret misappropriation brought against a company that was incubating and developing that same technology internally). The former case could be cheap to bring, and may even be passed on for prosecution by government authorities (therefore also possibly avoiding negative PR for the company). The latter case could be very expensive to pursue, and may very well come down to a jury decision that awards only nominal damages.
MEMS Investor Journal: What is the cost of developing and maintaining a patent portfolio? What are the drivers of these costs?
Marius Domokos: Patent portfolio costs can be predicted with reasonable certainty. At the application filing stage, the costs of outside counsel plus the filing fees can range between $9K and $20K+, depending on the fees charged by patent counsel, the complexity of the application, and any synergy with other patents or applications prosecuted by the same counsel. During the patent prosecution stage, there will generally be between one and three rejections issued by the USPTO, and the cost of responding to such rejections will normally vary between $3K and $5K+. Expanding the patent family outside the US (e.g., via PCT applications) results in an initial increase in legal fees in connection with the international application, and subsequently produces a cost multiplicative effect once the application enters into national phases in specific designated countries. Once issued, patents incur maintenance costs, which vary from country to country. Techniques exist for deferring costs for foreign prosecution in connection with PCT applications and for decreasing maintenance fees in select foreign jurisdictions (e.g., UK and Germany).
MEMS Investor Journal: Why do some established MEMS companies have extensive patent portfolios with hundreds of granted and pending applications while some have very few?
Marius Domokos: A rigorous answer to this question would require targeted analysis for specific companies. Nevertheless, some general trends can be ascertained in the semiconductor industry (e.g., companies with semiconductor manufacturing facilities do not tend to sue each other but they still develop large patent portfolios, patent litigation is more common in the semiconductor equipment and materials space, etc.).
I would like to believe that companies that have few patents have undertaken careful analysis and have concluded that the cost-benefit of a significant patent portfolio would be negative, but I know that this is not clearly the case for some companies. Generally, before developing a significant patent portfolio or deciding not to do so, a company should project the costs of the portfolio, try to quantify the expected benefits of the portfolio in defensive and offensive terms (e.g., ability to deter lawsuits in the first place, ability to force more favorable settlements if suits are filed, etc.), and compare the two. This analysis is not trivial by any means, given that it involves expected values and compounded probabilities that can be determined only with limited degrees of confidence, and may very well be impossible to conduct early in the life of a MEMS startup. But by the time the company has filed 10-15 patent applications, the company’s Director of IP or outside IP counsel should attempt to develop some relevant metrics and quantify the patent filing strategy for the company.
MEMS Investor Journal: In terms of IP litigation matters relating to MEMS where do you see the most action in terms of enforcing IP rights and defending against claims of infringement or misappropriation?
Marius Domokos: Most of the IP litigation involving MEMS companies relates to patent infringement. Traditionally, semiconductor companies used to be sued by competitors, lawsuits were rarely surprising, and settlements tended to follow predictable paths (e.g., “You are not really seeking to get an injunction against our products, so how much do we have to pay to make this go away?”).
In the modern world of globalization and increasingly-efficient patent trading markets, lawsuits can be brought by unknown entities, universities or research institutes, can involve trans-border injunctions and strategic lawsuits in foreign jurisdictions, and companies are faced with the possibility of multiple patent assertions by different parties targeting substantially the same technology. Needless to say, defending against patent infringement claims can be very costly, and as a rule, it actually becomes progressively more expensive for successful companies as they increase product offerings and revenues.
To a lesser degree, MEMS startups get involved in trade secret disputes, generally in connection with key employees that previously worked for other companies and departed from the previous employer on less than cordial terms, or as a result of unsuccessful commercial relationships.
Copyright disputes involving MEMS startup are less common, given that MEMS companies do not produce much software and that their MEMS designs are usually shared with third parties in encrypted format (e.g., GDS files), so are harder to understand and copy.
MEMS Investor Journal: With the growth of technology and the availability of information resources to anyone in the world, can you comment on it becoming easier or more difficult to (a) identify and (b) prosecute IP infringement?
Marius Domokos: Technology is certainly facilitating all aspects of IP infringement identification and prosecution, but not necessarily less expensive. In any event, it is clear that the Internet is making it much easier for a patent owner in any country to search for potential infringers around the world, identify those companies’ suppliers and customers in the commercial chain, and track the final products that are entering that country as potential targets.
In terms of trade secret misappropriation, litigation complaints tend to be based on circumstantial evidence (e.g., a former joint development partner abandoned the joint project and suddenly brings to market a product substantially identical with the one previously contemplated by the parties). Modern technology could help in this case by facilitating analysis of the products suspected to incorporate misappropriated technology (e.g., decompiling/decrypting a design).
With respect to patent infringement, a plaintiff could bring an action based on limited information, with the expectation that additional relevant data will be obtained during discovery (a process that is more extensive in the US than in other foreign jurisdictions). In many cases, however, a prospective plaintiff will use its internal technical resources or hire external consultants to ascertain the probability of patent infringement based on the information available in the public domain and the functionality of the target products.
As a side note, many companies do not realize that by building elaborate websites to market and support specific products, they are indeed making it easier for their customers to reach the respective information, but they are also increasing the probability of patent assertions against those products by making it easier for patent owners to build an infringement case. In some cases, a patent owner will engage an independent lab to perform destructive analysis of a suspected infringing product, and will then use images produced using SEMs or other inspection or metrology equipment as a basis for the infringement analysis.
MEMS Investor Journal: In general terms if not MEMS specific, what if any countries or regions have particular notoriety in terms of IP infringement worldwide.
Marius Domokos: It is hard to conclude that any particular country or region is more active in IP infringement since there are many issues that come into play. With respect to patent infringement, infringement is usually unintentional, with the infringing company not even aware of the existence of the patent. Whether companies in some countries tend to engage in more systematic intentional copying of patented technology is a matter of public debate. And this also applies to trade secret misappropriation.
As a general rule, we could probably conclude that countries with less mature legal systems tend to have more inefficient mechanisms and less reliable forums for prosecuting and stopping actual IP infringement. The main strategy for a MEMS company with respect to these countries should be to minimize disclosure of information in unencrypted format in those markets, but should not seek to avoid doing business there given that those countries are becoming important
commercial partners in the global economy and their governments are making good progress in closing IP protection gaps.
MEMS Investor Journal: What are the strengths and weaknesses in support by the FTC and ITC toward IP protection on a national and international scale?
Marius Domokos: The ITC has emerged as a powerful forum for bringing strategic patent infringement assertions and expecting faster and tangible effect. From a plaintiff’s perspective, a complaint with the ITC is often a good way to bring a potential patent licensee to the table for meaningful commercial discussions. From a defendant’s perspective, the prospect of a fast ITC decision to prevent importation into the US of the defendant’s products is a scary prospect, and the pace of procedures sometimes appears too fast. Given ITC’s ability to effectively block importation of products into the US, the ITC’s actions have implicit global reach.
The FTC is also a powerful forum for addressing disputes involving unfair competition, and together with the DOJ, antitrust. MEMS startups are unlikely to raise issues of antitrust in their early stages of corporate evolution, but need to pay attention to their marketing programs and comparative advertising to avoid unpleasant letters from their competitors or admonishment from Government entities.
Marius Domokos is an IP attorney specializing in strategic counseling for technology companies, with a focus on patent portfolio strategy, technology transactions and patent litigation. Marius received BS and MS degrees in Electrical Engineering from the University of Southern California and a law degree from U.C. Berkeley Boalt Hall. Marius currently works for Wilson Sonsini Goodrich & Rosati, where he spends a significant portion of this time advising startup companies in the semiconductor space. Marius previously worked in house as an IP attorney for KLA-Tencor and Intel, where he acquired a deep understanding of the problems facing companies in their day-to-day existence and developed pragmatic solutions for addressing corporate IP needs. Prior to law school, Marius worked as a systems engineer for Hughes Space and Communications. In his spare time, Marius writes a blog focused on IP issues encountered by companies in the semiconductor and electronics industries (http://semiconductorlawblog.com). Marius would be happy to address your questions and comments relating to IP on his blog.
The opinions expressed in this interview belong to Marius Domokos and are not necessarily shared by his employer.
Copyright 2008 MEMS Investor Journal