Business models and technological innovation

Mar 08, 2019
0
0

In an earlier post I discussed how biotech companies can earn a return on a technology either in the product market or the market for ideas. Although this appears to be a dyadic decision, it is more helpful to think of a continuum of choices, between plugging into the value chain early and full vertical integration, with many different ways in which a firm can interact with its value chain. The best strategy will depend on how well the market for ideas works. Important factors include the degree of information asymmetry between the seller and the buyer, the need for investments in specialized assets, how easily knowledge can be transferred between parties, and how strong the intellectual property protection is.

Looking at pharmaceutical development, there is a broad range of technologies and projects that span these factors, suggesting that different business models may be appropriate for various technological innovations. In Science Business, by Gary Pisano (published in 2006) the author provides a useful examination of four broad classes of technological innovation and the common business models associated with them:

• novel research methods and tools (such as high-throughput screening, combinatorial chemistry, bioinformatics)

• identification of novel mechanisms of action or targets (angiogenesis, RNAi)

• creation of novel compound types (rDNA, MAbs)

• identification of novel treatment modalities and therapeutic markets (gene therapy, xenotransplants, or drugs for rare genetic diseases).

There is broad variation within the technology categories described by Pisano, so we have to be cautious in generalization. The technology does not necessarily determine a firm’s business model, but rather influences it. Other factors, such as a firm’s ability to access capital, also influence the choice (and success) of a business model. Common business models across the four technology classes are discussed below, as is their popularity. This discussion draws heavily from Science Business.

Novel research methods and tools

Several business models are available to these companies, including simply licensing the use of the technique or tool to other drug companies that would then use them in their own discovery process. A second model would be to sell drug discovery services, while a third strategy would be to vertically integrate forward into drug R&D and develop proprietary molecules.

The market for ideas usually does not work efficiently in the first strategy, because the difficulty in fully sharing background information may make it difficult to convince potential licensees of the value of the technology or tool. Furthermore, the licensee would probably have to invest in specialized equipment (complementary assets), raising their risk. Difficulty may occur in transferring knowledge that is not easily codified, impeding the adoption of the new technology by licensees. If the intellectual property protection is not air-tight, the innovator could expose itself to imitation. Under the second business model all of these risks and issues are removed.

In the late 1990s this service model was followed by many platform technology companies. However, many of them, such as Millennium, Celera and Human Genome Services, abandoned this strategy to vertically integrate into the development of proprietary. Vertical integration (FIPCO/FIBCO) based on a platform technology is likely to be overkill and may even be suboptimal if the firm lacks downstream capabilities such as manufacturing, marketing and distribution. However, being a service or tool company offers a very different risk-reward profile than drug development. The problem that platform companies faced in the late 1990s was not their business models, but the environment created by the genomic bubble in which unrealistic valuations could not be sustained with a service or tool model.

Novel targets or mechanisms

Innovation here is concerned with the identification of new disease targets or mechanisms of action implicated in diseases. The market for ideas is not fully efficient is this situation. It is unlikely that intellectual property can be completely secured on a mechanism or class of targets. Often a lot of prior art exists and the intellectual property is based heavily on the kind of knowledge that cannot easily be transferred from one company to another, so it is unlikely that a firm in this innovation category can simply license its innovation. It is therefore more likely to pursue a drug discovery and development strategy. But how far down the drug development value chain should it integrate? This depends on the characteristics of the drug and the market. If it is a small-molecule drug candidate targeting a well established therapeutic market (hypertension, for example, or diabetes or depression) the rationale for full vertical integration is weak, assuming the innovator is able to secure IP protection on the molecule. A licensee would likely have the necessary complementary assets and capabilities required to take the drug candidate down the development pathway to market. Tacit knowledge (the knowledge that is difficult to write down and pass on) may be an issue in designing or interpreting clinical trials in some instances but can be overcome through close collaboration with the licensee. A long-term commitment would have to be made, and would probably be a more efficient solution than full vertical integration.

Novel compound types/novel treatment modality and novel markets

Biotechnology is bringing us new types of therapeutic molecules, such as rDNA, stems cells, monoclonal antibodies and new treatment modalities. Novel market opportunities are also being developed such as those for orphan disease for personalised medicine.

These types of innovations can be difficult to license due to lack of knowledge and capability on the part of would-be partners. Also, importantly, they typically require significant investments in downstream assets (development, manufacturing, distribution). Full vertical integration may be the logical strategy for these types of innovation. Collaborations have been seen with these opportunities, but the risks are high, disputes common and collaboration may be a second-best strategy. While vertical integration reduces the risks of operating in an inefficient market for ideas, it raises other risks. The level of capital required is huge, and may preclude R&D portfolio diversification. Younger firms pursing a FIPCO strategy may have everything resting on the success of a single (first) therapeutic candidate.

Pharmacogenomics is the branch of pharmacology behind ‘personalised medicine,’ in which drugs and combination therapies are optimised to an individual’s genetic makeup. Pharmacogenomics looks at the influence of genetic variation on drug response in patients by correlating gene expression or mutation with a drug’s efficacy or toxicity.

Whilst personalised medicine has been slow in making its mark on healthcare, it will probably be an important pillar in the future of drug therapy. The model for personalised medicine is evolving. It is likely to be more than one unique model, varying with the underlying technology, involving discovery companies, pharmaceutical company collaborators and clinical laboratories and may involve participation by healthcare providers.

It is clear from the preceding discussion that technology type has an important influence on business model by the bearing it has on the need for specialized assets and the difficulty in transferring underlying knowledge. Other factors influencing the choice of business model include how well the intellectual property can be protected and the firm’s ability to access capital.

In my next post I’ll talk about the strategic issues facing biotech start-ups and the kinds of strategies firms employ in response.

Janette Dixon

No comments yet.