Allocating patent rights in collaborative research agreements: prospective partners can...

The number and scope of collaborative research agreements are increasing rapidly as R&D executives actively manage the costs and risks of their research project portfolios. In today's markets, products contain more diverse technologies than ever before and customers demand capabilities unanticipated

Planning and negotiating a collaborative R&D alliance requires creative thinking based on a disciplined and structured methodology. Leading alliance firms have formalized this methodology and built it into their corporate processes (1).

A central theme of every methodology is that intellectual property issues are an essential part of both the pre-agreement and post-agreement stages. In this article, we describe a useful approach to one important intellectual property issue: the allocation of patent rights between partners.

Note that we sometimes use the phrase "intellectual property" in referring to patent rights. With modifications, many of the patent rights ideas in this article are applicable to other important intellectual property matters such as allocation of rights to technical know-how or proprietary marketing information. These vital types of intellectual property are not specifically covered here.

We describe several concepts and tools intended to enable business and technical leaders, working closely with their legal counsel, to reach mutually acceptable patent rights in an R&D alliance agreement. These concepts are based on a simple but crucial assumption: the guiding principle for allocating patent rights is a clear understanding of both firms' technical and business strategies, which in turn is driven by the marketplace intents of both firms.

If the marketplace intents of the parties are aligned (a fundamental requirement for reaching a satisfactory alliance agreement), then the partners can reach agreement on patent rights allocation. In this context, "aligned" does not mean that the parties intend to cooperate in the marketplace under all conditions. Rather, "aligned" means that the parties agree to 1) the circumstances under which they will cooperate in the marketplace and 2) the circumstances under which they will proceed independently of each other in the marketplace (including the potential for competing with each other).

In what follows, we will demonstrate how marketplace intents can be best revealed by an initial focus on "rights to use" patent rights rather than on "ownership." We will provide several key definitions that must be understood as the parties develop rights to use rules that reflect marketplace intents. We will then discuss a key issue that must be addressed in every collaborative research agreement: to what extent should rights to use patents depend on the source of an invention made during the collaboration? We then introduce a straightforward tool, the Intellectual Property Needs matrix, which both prospective alliance partners can use to develop and negotiate their rights to use positions. At the end of the article, we will mention some important issues that must be considered as rights to use are translated into the terms of an agreement. Throughout, we will emphasize the critical need for research and business managers to get effective and ongoing support from their legal counsel as marketplace intents are translated into agreement terms.

We begin now by describing the common pitfall of collaboration prior to a clear understanding of patent rights allocation.

The Risk of Preventive Collaboration

Frequently, the technical staffs of partners in a prospective R&D alliance begin to work together before an alliance agreement is reached. Sometimes this work is done under a non-disclosure agreement or letter of intent, and sometimes it is done without any clear understanding at all between the prospective partners. The temptation to start work can be overwhelming. Aside from a desire to move the project forward, there may be a perceived need to conduct exploratory cooperative work to determine the potential value of full-fledged collaboration.

Nevertheless, collaborative work prior to an unambiguous agreement on intellectual property rights is always a bad idea. The usual non-disclosure agreement or letter of intent does not adequately address the allocation of intellectual property rights based on exploratory joint work. Neither party has carefully considered the problems of determining the source of inventions or the phenomenon of stimulated invention between the two technical staffs. More importantly, they have not resolved these problems in terms of each firm's marketplace intent.

Collaboration without an agreement can have unintended consequences. As just one example, in a situation in which two firms create a patentable invention absent an agreement, United States law defines the invention as jointly owned. When an invention is jointly owned, either partner can exploit the invention without the consent of the other joint owner. This includes licensing or selling their intellectual property interest to anyone they wish, including the partner's significant competitor (2).

We believe that collaborative R&D should never be done before a clear allocation of intellectual property rights, including patent rights, is agreed upon. Often this requires the completion of an entire alliance agreement. To those who argue that the pressures of time make it unwise to delay collaborative work until intellectual property rights are agreed upon, we have a simple answer. If the parties are sufficiently aligned in terms of marketplace intent, they should be able to agree on the allocation of rights without undue delay. If they cannot readily agree on allocation of rights, the parties have a marketplace misalignment that must be resolved before collaborative work starts. The time to resolve that misalignment (and the accompanying allocation of intellectual property rights) is before an invention occurs. Otherwise there can be negative marketplace consequences for one or both parties.

Linking Patent Rights to Marketplace Intents

Experienced alliance managers understand the connection between patent rights and "marketplace alignment." Marketplace alignment means that the potential alliance partners reach an unambiguous understanding regarding how each will use the patent rights flowing from the collaboration to carry out their own business strategies. The parties agree under what circumstances they will be working together in the marketplace, with each playing defined roles. They will also agree on the circumstances under which they will not be working together in the marketplace, but rather proceeding independently of each other. "Circumstances" may refer to products, applications or geographies.

In the next section we introduce the Intellectual Property Needs Matrix, which helps potential partners connect marketplace alignment to several different dimensions of patent rights. But first we will present a simplified example to introduce the principle.

Suppose that Ajax, Inc. and Brown, Inc. are considering an alliance that includes development of technology with potential application in two product areas: widgets and stegdiws. The initial driving force behind the alliance was their common interest in widgets and the realization that both lacked the resources to compete effectively against larger players in the widget market.

The alliance involves technology development, widget product design, and cooperative widget manufacturing and marketing. But the parties recognize that the technology will also be applicable to stegdiws, and both sides agree that they have no interest in collaborating in stegdiws. On the contrary, each views the other as a competitor in the stegdiw market.

This is an example of reaching "marketplace alignment": Ajax and Brown reach a clear understanding as to their relationship in the relevant product markets--in widgets, close collaboration to compete vigorously with larger firms; in stegdiws, head-to-head competition.

Given that marketplace alignment, Ajax and Brown are able to agree on the allocation of patent rights for any invention arising from the alliance. For widgets, each will have the necessary rights to use such patents to carry out their responsibilities in the alliance. That is straightforward and obvious because Ajax and Brown are cooperating in widgets. But stegdiws are outside the alliance. Ajax and Brown will compete in stegdiws; both want to apply inventions created in the alliance for stegdiws, to enable them to proceed independently in the marketplace for stegdiws. That can be achieved if both have "rights to use" (RTU) such patents for stegdiws, independently of the other.

Note the emphasis on "rights to use" rather than "ownership" of patents. By focusing on RTU, both parties can describe their required patent rights directly in terms of marketplace intent. For example, given Ajax's need to proceed independently of Brown in the stegdiw market, Ajax needs RTU patents that might be invented by Brown in the course of the R&D collaboration. Similarly, Brown needs RTU patents of Ajax for application in stegdiws. However, in this situation the grant of licenses for application in stegdiws will be nonexclusive.

For example, Brown will grant only non-exclusive RTU to Ajax for Brown's patents. Brown wants to exploit its own patents for stegdiws. It may also decide to enter into alliances for stegdiws with a third party. Brown may need to grant that third party the necessary RTU. Ajax will have the same requirement for its own patents.

As we will discuss below, patent ownership is relevant and cannot be ignored. But experience shows that efforts to connect marketplace intent to allocation of patent rights (as well as allocation of rights to other types of intellectual property) are much more productive if R&D executives concentrate first on RTU. Once the necessary RTU are agreed upon, counsel for both sides can develop mechanisms (such as appropriate licensing terms) to achieve the agreed RTU.

Key Concepts in Patent Rights Allocation

Effective negotiation of the allocation of patent rights requires understanding of three key concepts: background and foreground intellectual property, alliance boundaries, and the "joint vs. sole" question. We will describe each concept.

Background and foreground intellectual property

"Background intellectual property" is usually defined as relevant intellectual property that either partner created prior to the start of the alliance. In some circumstances, the parties might agree to define background intellectual property to include post-start intellectual property developed by a party outside the scope of the work being done under the alliance. For example, this might apply where one party has extensive R&D operations independent of the agreed joint work, and where that party wants to avoid having the fruits of such R&D work treated as work under the alliance.

This extension of the definition of background intellectual property creates opportunities for conflict during alliance implementation, and must be handled with care by both parties. We will not discuss this matter further in this article, and we will use the more usual definition of background intellectual property (i.e., developed prior to the alliance).

"Foreground intellectual property" is defined as relevant intellectual property created during the alliance. Using the usual definition of background intellectual property as that created by a party prior to the start of the alliance, foreground intellectual property is that created by either, or both jointly, during the alliance term.

Boundaries of the alliance

Boundaries are a clear statement of the scope of the alliance. Inside the boundaries, the firms are allied. They will play by a set of rules defined by the alliance agreement. Outside the boundaries, the firms are not allied and a different set of rules will apply.

Severe problems can arise in alliances when boundaries are not clearly defined. Consider an example of two firms collaborating to develop one or more products, under an agreement in which the firms agree on significant restrictions on the use of each other's background intellectual property:

The management of Company A views the product boundaries of the collaboration as the circle on the left in Figure 1, while the management of Company B believes the product boundaries are represented by the circle on the right. The first product being cooperatively developed is in the overlapping area, which both firms agree is inside the boundaries. Now suppose Company B uses Company A's background intellectual property in a different product represented by Point 1. Company B believes that such use is allowed under terms of the alliance agreement that permits use of the partner's background intellectual property in products within the boundaries. But Company A does not recognize the product represented by Point 1 as inside the alliance product boundaries, and is angered because the alliance agreement does not permit the use of each other's background intellectual property outside the boundaries. Company A's managers react badly since they view Company B's behavior as a deliberate violation of that agreement. Managers in Company B do not understand Company A's reaction because they believe Point 1 is within the boundaries.

[FIGURE 1 OMITTED]

Actual disputes can be more complex than this example, since the dimensions of boundaries can encompass not only products, but also underlying technology, classes of applications or customers, or geographical regions. To avoid such arguments, which sometimes stem from genuine misunderstandings, it is essential that prospective alliance partners carefully define the boundaries, in all dimensions, as part of the negotiation.

The "joint versus sole" question

Firms enter into collaborative research agreements in order to create new technology. Both parties expect to be able to commercially exploit that newly created technology under agreed terms. But there is a sticky problem built into all collaborative R&D alliances, expressed as the "joint vs. sole" question, which must be faced squarely during the negotiation of the alliance agreement. As applied to foreground patents, the joint vs. sole question is: Will the partners' RTU foreground patents depend on which firm's scientist or engineer is the legal inventor of the patentable invention? The answer to that question has a profound impact on the working relationship during the R&D collaboration and on the marketplace impacts of inventions.

An illustration of the joint vs. sole question is seen in the Ajax/Brown alliance. Suppose that Ajax and Brown agree to grant RTU each other's foreground patents outside the boundaries, with broad sublicensing rights, on a royalty-bearing basis. With that understanding, there will be a financial consequence of determining the source of a foreground patentable invention because there will be a royalty flow from the non-inventing party to the inventing party.

Where the RTU patentable inventions are dependent on the source of the invention (as in Ajax/Brown), we refer to the parties using the "sole" option for allocation of patent rights. The word "sole" is used as shorthand to describe this RTU option since many inventions will be made "solely" by one or the other party, and the RTU will reflect that.

Two aspects of the sole option must be understood: First, some inventions may be actually "joint," in that the invention was made with significant contributions from both firms; the RTU terms must allow for that possibility. Secondly, nothing in the choice of the sole option requires that the non-inventing party have no RTU the patents of the inventing party. Rather, the RTU in the sole option are different for the inventing and non-inventing parties, usually more favorable for the inventing party. In the Ajax/Brown situation, the source of the invention determines the direction of the royalty flow.

In a nutshell: The defining characteristic of the sole option is that each party has an important stake in determining the source of each foreground invention, because the source will affect the RTU of each party. While at first glance the sole option might be the obvious way to approach allocation of patent rights, it can have a chilling effect on collaboration. Scientists and engineers ask for openness from the partner's scientific staff and are reticent to share information of their own when RTU and inventorship are linked. The "joint" option is an alternative.

In the joint option, the RTU of each party are independent of the source of the invention. For example, suppose Ajax and Brown decide that each will have a royalty-free RTU inventions of the other outside the boundaries, and that the terms of those RTU (such as sublicensing rights) are identical. In that situation, neither Ajax nor Brown cares which party is the inventor--they have agreed that the rights to exploit a foreground invention outside the boundaries will be the same for both, regardless of the inventing source.

Assuming that Ajax and Brown agree that all other RTU (for example, after alliance termination) will be independent of the source, they have chosen the joint option. "Joint" is used to describe this option because all inventions are treated, for RTU purposes, as though they were joint. This option fosters collaboration because the RTU are unaffected by inventorship.

There is another aspect of the joint option that must be understood. In the joint option, the RTU foreground intellectual property are often, but not always, identical for the parties. Suppose Ajax was vitally interested in Product Z, while Brown was vitally interested in Product ZZ, both products being outside the boundaries. Under a joint option, Ajax could have RTU all foreground patents as applied to Product Z (but not to Product ZZ), while Brown could have RTU foreground patents as applied to Product ZZ (but not to Product Z). In this example of the joint option, the RTU are independent of the inventing source, but are different for the two parties as applied to Products Z and ZZ.

In a nutshell: The defining characteristic of the joint option is that the parties do not have a vital stake in determining the source of each foreground invention, since the RTU are independent of the source of the invention.

Making the choice

Having understood the difference between the sole and joint options, which is the better choice for a R&D collaboration? There is no single right answer--there is a tradeoff involved in the choice, which the parties must address in agreeing on the choice. If the parties neglect to specifically address this choice in the alliance negotiation, the sole option is the legal default and will apply to any patentable invention that results from the collaboration.

In the sole option, since RTU depend on inventing source, technical people will tend to be circumspect in their interactions with the other firm. Open technical discussions may be guarded, until new ideas are carefully documented. Brainstorming and informal idea exchange will be limited. Some of the benefits of cooperative work may be lost.

In the joint option, interactions are more collaborative. While technical people desire credit for inventions (just as in normal intra-company work), there is much less reluctance to have informal and open idea exchanges with staff members from the other firm. But the downside of the joint option can be severe. With equivalent RTU, a non-inventing party might be able to grant a sublicense to an inventing party' s patent to the inventing party's most significant competitor. The non-inventing party may not even have to inform the inventing party of that sub-license or share licensing revenues. Essentially, the inventing party may lose significant control over the marketplace use of its own patents.

Because of this issue, alliance partners using the joint option will sometimes agree to specific, equally binding limitations on sub-licensing rights. Like all aspects of intellectual property rights, the nature of any such limitations is closely tied to the marketplace intents of the parties.

In making the tradeoff between the sole option and the joint option, the parties must consider such matters as the scope of the R&D collaboration, the nature of the relevant marketplace, and the potential for agreeing on mutual constraints on sublicensing consistent with applicable law. For example, a close collaboration involving cooperative work on the same technical issues, perhaps with exchange of scientists, may favor the joint option. The marketplace risks may be acceptable, especially if the parties agree on lawful constraints on third party licensing.

In contrast, in a collaboration where the R&D tasks are well separated, with each party working on a different set of tasks and with minimal informal information exchange, the sole option may be best. Each party retains more control over its own inventions while not unduly constraining the effectiveness of the R&D collaboration.

The Intellectual Property Needs Matrix

With a clear understanding of its marketplace intents, the definitions of background and foreground intellectual property, the most appropriate boundaries for the collaboration, and the joint vs. sole question, research managers must develop the firm's patent rights positions and communicate them both to internal stakeholders and to the prospective partner. The Intellectual Property Needs Matrix (Figure 2) is a tool to help managers do that.

The goal of the Matrix is to fill in each cell with clear statements of each firm's required RTU. Filling-in Cells 1 and 2 is usually straightforward. Refer back to our example of two firms collaboratively developing a product. The statements in Cell 1 provide each firm with the background intellectual property rights it needs to carry out its assigned tasks in the alliance. Therefore, each firm will have the rights to use the other's background intellectual property in order to carry out its roles.

Cell 2 deals with foreground intellectual property. Each firm will have the RTU foreground intellectual property needed to carry out its roles in the alliance.

Cell 4 describes each firm's rights to use foreground intellectual property outside the boundaries of the alliance. The RTU terms in Cell 4 are sensitive to the firms' choice of the sole or joint option. In the joint option, both firms may be reluctant to see the other have "joint rights" to exploit foreground patents broadly outside the alliance boundaries. The solutions to these issues are directly tied to both firms' marketplace intent and may include a variety of restrictions, including field of use, geography and time.

The terms agreed upon in Cell 4 lead directly to the terms in Cell 3, dealing with use of background intellectual property outside the boundaries. One solution for this cell is for each firm to grant the other licenses to background intellectual property, but only as needed to exploit foreground intellectual property outside the boundaries. This solution reflects one of the realities of alliances: foreground intellectual property builds on the background intellectual property of one or both partners.

The solutions to Cells 5 and 6 are based on similar reasoning. The firms must understand their marketplace intents after the termination of the alliance. These intents are translated into a set of foreground intellectual property RTU in Cell 6 and a set of background intellectual property licenses in Cell 5 needed to exploit the foreground intellectual property.

One issue complicates the negotiation of the RTU in Cells 5 and 6: the reason for termination. If the parties terminate the alliance because all of the alliance objectives are achieved, then Cells 5 and 6 will contain a particular set of RTU rules. However, if the parties terminate the alliance because one party breached the agreement, then a different set of RTU rules apply. There are at least six reasons an alliance can be terminated:

1. The objectives of the alliance are successfully achieved and the partners want to proceed in the marketplace independently.

2. One of the partners changes its strategic direction or decides that other projects have a higher priority.

3. One of the firms identifies a more attractive partner, and wants to terminate the current alliance and create a new collaboration with a new partner.

4. One or both firms decide that the implementation results are less successful than expected and future prospects are not encouraging.

5. One partner has not performed adequately or has breached the terms of the contract.

6. One partner is acquired or merged with another firm.

Firms may negotiate a distinctly different set of RTU rules in Cells 5 and 6 for each reason.

Translating the Matrix into Alliance Agreement Terms

The vital role of legal counsel

Once the business and research managers clearly understand each firm's RTU for both foreground and background intellectual property in terms of their market intents, legal counsel for both sides translates those RTU requirements into the appropriate legal structure. Fortunately, United States law provides counsel with flexibility in allocating patent rights. The law states that inventions and patents may be assigned in whole or in part, and may be licensed exclusively, non-exclusively, in fields of use, and in specific territories. For example, ownership and patent rights to inventions made under joint R&D agreements may be allocated between the partners in any way the partners agree, subject to specific requirements such as anti-trust law.

We strongly recommend that legal counsel for both parties be involved in intellectual property RTU discussions from the beginning. This allows them to understand both parties' intents and translate those intents into a set of RTU, ownership and licensing provisions. Counsel can identify legal issues, explain the implications of various alternatives and begin formulating contract language. Also, counsel begins work on the following important patent issues: each firm's responsibility related to drafting and prosecuting patent claims, selection of countries to file for patent protection, patent maintenance, and enforcement.

Some implications of joint ownership of IP

Joint ownership has implications in the areas of drafting patent claims and enforcement. Drafting patent claims is how intellectual property attorneys define the legal boundaries of a patentable invention. Experienced counsel ensures that claims cover as many key business interests as possible. Joint ownership of patents complicates the drafting process because claims that are crucial to one partner's interests may be irrelevant to the other. This means that each firm's intellectual property counsel must spend time educating the other on why claims must be written in a certain way. There is always the possibility that patent counsel will not agree. This leads to increased cycle time and poor utilization of intellectual property counsel's time.

Enforcement is an equally important issue. The ability to decide when and how to pursue patent litigation is an important part of intellectual property strategy. Technology-based firms have a valid business interest in protecting their intellectual property. In the United States, courts will not permit a patent infringement suit to be brought unless all of the parties having an ownership interest in the patent are named in the suit. The courts take this position to prevent the possibility of multiple suits for the same infringement. Therefore, if a patent is jointly owned, both firms must be named as parties in the suit. Absent a contractual agreement, one joint owner cannot compel the other to join in an infringement action. If Company A decides not to join co-owner Company B in an infringement suit, Company B finds itself in a difficult situation (3).

There are other enforcement issues as well: Who pays for the litigation? Does adding one's name to the suit invite the firm to be counter-sued, deposed, or otherwise brought into a legal fight the firm does not want?

Additional issues must be taken into account. When patents are jointly owned, questions arise as to which firm will pay to file, maintain the granted patent, and file for foreign (non United States) equivalents. If the patent is commercially successful, can one party unilaterally abandon the patent or will one party be obligated to prosecute and/or maintain regardless of their continued interest (4).

These issues must be satisfactorily resolved and put into the alliance contract to set the stage for a good alliance relationship. For all the reasons above, joint ownership is often considered an undesirable option.

Summing Up

While this is a complex subject, we hope to have persuaded readers of three things:

1. A clear understanding of both firms' marketplace intents should guide patent allocation negotiations with an initial focus on rights to use.

2. Combining the Intellectual Property Needs Matrix with careful thinking about alliance boundaries and the "joint versus sole" question facilitates clear understanding and communication of both firms' RTU needs.

3. Early and ongoing involvement of legal counsel in the planning and negotiation process enables optimum selection of the legal structure that maximizes the value of the collaboration.

Acknowledgments

The authors acknowledge Mary Burch of Rohm and Haas Company and Terry Mohr of Timken (retired) for their able leadership of the Industrial Research Institute "Research-on-Research" project on this topic. Additional thanks to the reviewers of this manuscript for their valuable suggestions; and to Rutgers University and the Federal EDA University Center Program for their support.

Caution About the Joint Option

It is still necessary to properly identify the source of a patentable invention in order to comply with patent law. Inventorship is an important decision that affects such matters as patent filing, prosecution and associated costs. A patent can be held invalid if the inventor or inventors are not properly named, or if a non-inventor is named on the patent. Inventorship may be subject to differing opinions, may be subject to legal interpretation and may legally change due to claim amendments during the patent prosecution process (3). All of these issues must be addressed by lawyers.--G.S, and M.S.

References and Notes

(1.) Slowinski, Gene and Sagal, Matthew W. The Strongest Link: Forging a Profitable & Enduring Corporate Alliance. New York: AMACOM, 2003.

(2.) The actual words are found in United States patent law at 35 U.S.C. 262: "In the absence of any agreement to the contrary, each of the joint owners of a patent may make, use, offer to sell or sell the patented invention within the United States, or import the patented invention into the United States, without the consent of, and without accounting to the other owners." Laws in other countries may differ.

(3.) Dillahunty, T. Gene. How To (And How Not To) Deal with Inventorship in Joint Agreements. les Nouvelles, March 2002, pp. 1-6.

(4.) O'Reilley, D. Patrick. Allocation of Ownership of Inventions in Joint Development Agreements: the United States Perspective. les Nouvelles, December 2000, pp. 168-172.

Gene Slowinski is the director of strategic alliance research at Rutgers University Business School, Newark, New Jersey, and the managing partner of the Alliance Management Group Inc., Gladstone, New Jersey. He has over 20publications in the areas of technology management and business development. He received his M.B.A. and Ph.D. in management from Rutgers University. gene@strategicalliance.com

Matthew Sagal is a senior partner in the Alliance Management Group, Gladstone, New Jersey. Previously, he was an R&D, manufacturing and business development executive at Bell Laboratories and AT&T. He earned a bachelor's degree in chemical engineering from Cornell and a Ph.D. in physical chemistry from MIT. mwsagal@comcast.net.

Figure 2.--The most effective allocation of patent rights occurs when
each firm has the background and foreground patent rights it needs to
carry out its business intent. By filling in each cell of the
Intellectual Property Needs Matrix, managers from both firms develop
a clear picture of their needs. This tool allows them to communicate
those needs internally to key stakeholders and externally to the
partner.

Intellectual Property Needs Matrix

                          Background     Foreground
                         intellectual   intellectual
                           property       property

Inside the Boundaries       Cell 1         Cell 2
Outside the Boundaries      Cell 3         Cell 4
Upon Termination            Cell 5         Cell 6