Science-based businesses as organizational models

Gary Pisano of the Harvard Business School has a new working paper on The Evolution of Science-Based Business: Innovating How We Innovate. First of all, Pisano differentiates between technology-based and science-based businesses. Technology-based businesses, like software and electronics, develop and apply existing science. Science-based businesses, such as biotech, must engage in developing new science. That difference makes science-based businesses far more risky - since the science may or may not pan out:

Science‐based businesses are at the frontier of knowledge. Technical failure is the norm, not the exception. What is known pales in comparison to what remains to be discovered.

. . .

Thus, not only might the financial costs of exploration be high, but critical technical uncertainties may not be easily or quickly resolvable early in the development process. And, even if an organization can resolve those uncertainties through research, there is no guarantee the resulting intellectual property will be appropriable. "Deeper understanding" may be critical to further development, but it is generally not patentable.

That fact of "science" limits how such science-based businesses can raise capital. After discussing the limits of venture capital and capital markets, he offers this discussion of IP monetization:

An alternative or complementary strategy for a firm to raise capital for its R&D is to "monetize" its intellectual property. That is, rather than trying to develop a whole product and earning revenues on product sales, the company essentially licenses out the project to another firm. Such licensing has become a huge part of the R&D world in most technology intensive industries. There are literally thousands of R&D agreements and licensing deals that occur every year. One of the chief benefits of intellectual property monetization is that it enables firms to manage risks. It also enables firms with complementary capabilities to access know‐how.

Monetization of intellectual property is not a new phenomenon. Firms have licensed intellectual property for more than a century. However, the extent of this IP monetization appears to have grown dramatically in the last few decades. Since science‐based businesses rest on intellectual capital, it stands to reason that markets for know‐how will play an ever more important role in the future. However, we must also understand that monetization of IP has limits as a device for creating the required integration.

Market mechanisms work best when the relevant "modules" of knowledge are clearly defined. Thus, modularity facilitates collaboration (Teece 1982). This is one reason Open Source projects like Linux have been so successful. The modular architecture of Linux enables thousands of software developers from around the world to make contributions without ever having to talk to each other directly or to meet face to face. The IP monetization approach is often predicated on an assumption that the IP in question is a discrete module or asset that can be bought and sold. However, as mentioned earlier, in science‐based contexts, the immaturity of the underlying knowledge base makes it less likely for modularity to exist. This suggests that achieving the required integration through licensing and the market for‐ know will fall short in science‐based contexts.

I'm not sure he has completely grasped the role of IP monetization. In some industries, such as electronics, the licensing process is one of integrating modules. But in biotech, the process seems to have two other roles: division of labor and capital formation. The division of labor function of licensing spreads the work among several organizations, specifically between the new drug development and approval process and the production and marketing processes. Licensing (and sale) of biotech IP also functions in the timeless manner of swapping long term revenues for upfront capital. Licensing and other forms of IP monetization use the revenues from the previous science-based success to fund the next scientific gamble. Thus, it may be perfectly suited to the high risk nature of these types of businesses.

Ultimately Pisano argues that these science-based endeavors require new organizational models -- based on a view from Alfred Chandler that "it is hard to think about technological innovation as anything but tightly intertwined with organizational and institutional innovation." As he notes:

Science‐based businesses in biotech and elsewhere have 'borrowed' many elements of organizational technology (venture capital financing, use of the public equity markets for liquidity, monetization of intellectual property, etc.) that have been used, often successfully, in other technology contexts such as electronics and software. However, as argued above, science‐based sectors create novel organizational challenges around the simultaneous need to manage risk, integrate cross knowledge bases, and leverage cumulative learning. Addressing these challenges calls for new "organizational technology."

Here I would completely agree. But I would not limit the observation to only science-based businesses. Most innovation-based businesses (whether new science-based, based on existing science, or non-technological) face the same three challenges of risk, multiple knowledge bases, and learning. We can look to science-based businesses for clues to the emerging organizational models. But those models, I would argue, will end up being widely applicable in the I-Cubed Economy.