However, knowledge management in a biotechnology company must encompass more than patent filings and staff "gagging". A company's information and knowledge assets arise from many quarters—both internally and externally. For example, an increasing reliance on outsourcing and data sharing (particularly in genomics) means that much of the potential knowledge resources are generated outside a company. However, whether from internal or externally generated data, knowledge and information is held initially by individuals and not the body corporate.

Knowledge harnessed
The end product of good knowledge management procedures is "harnessed" knowledge—knowledge that can be readily put to use on behalf of all relevant members of the company (and its clients). "Harnessed" knowledge is held within an organization, but can be mobilized even after the originator has departed. Patents are one good example, or a highly cross-referenced database of expert advisors, or a searchable manual of specific laboratory protocols.

The balance of knowledge management requirements changes as a company develops. In the startup phase, the few founding staff are concentrating primarily on R&D, and informal exchange of information ensures that knowledge is disseminated throughout the organization. However, small companies have only a few "information miners", and may lack sufficient market or competitor intelligence to adjust their commercial or research strategy appropriately. The coherence and compactness of an early-stage firm makes it relatively easy to disseminate knowledge among relevant employees. By the same token, however, a small company does not have the resources to try to fill its knowledge gaps. A small company, therefore, will find it relatively easy to "harness" most of the knowledge it has, but the large amount of "unheld" knowledge still represents a significant risk.

Late-stage companies may have more knowledge-gathering agents, but because of their size often have difficulty mobilizing that knowledge within the organization. Segmentation of functions and more rigid definitions of responsibility can act as barriers to internal information flow. Unless appropriate knowledge management systems are in place, a larger company runs the risk of losing coherence. Collectively, the company's employees will be able to reduce the amount of unheld knowledge, but if that intelligence rests only with particular individuals or within departments, then the company has failed to harness it optimally.

However, much of corporate endeavor—like human endeavor in general—progresses empirically. To either side of the well-lit path to corporate success are the false trails and beaten paths tried but not taken. This "negative knowledge" accumulates not only during R&D but also in manufacturing and commercial interactions. A company with an effective knowledge management process will be able to build a permanent body of awareness of "negative knowledge"—those approaches that didn't work out as hoped or anticipated. Insight gleaned from early disappointments is as valuable as the rare moments of success: each product launched is a survivor among thousands of failed new chemical entities; and for each headline-making pharmaceutical deal, there are many rebuffs. Thus, there is every reason for a company both to retain those hard lessons within its corporate memory, and to withhold those lessons from competitors pursuing a similar approach.

Elements of knowledge management
One measure of a company's ability to manage knowledge is the alacrity with which it identifies and pursues "unheld" knowledge, and the efficiency with which it converts "unharnessed" into "harnessed" knowledge.

In a company that has its knowledge management systems under control, "unharnessed" knowledge has only a fleeting existence—during that brief period before the new research finding or commercial tidbit (previously "unheld" knowledge) becomes "harnessed" for the company's benefit. Inadequate knowledge management results in short-lived "unharnessed" knowledge becoming "unheld" knowledge, and remaining in this intellectual "purgatory" before being rediscovered at a much later date.

That process of defining the information needs will help employees understand better their role in the knowledge machine that is the company they work for. For instance, researchers understand that the results of their experiments are important to the company. However, if they have come from academic backgrounds, they may be less clear on what they can say about their own work at scientific meetings or what might be of interest in other people's presentations. A company's management can clarify these definitions, and provide, through the knowledge management system, a reporting outlet that will channel information around the organization.

Clearly, much of the information generated in a company is not needed by all employees. Information flows need to be filtered so that raw data from every microtiter plate well or the notes from the day's sales calls do not obscure more important messages.

A knowledge management system does not have be burdensome: indeed, if it is, it will not be used. The system will also encompass natural hierarchical information flows. Research staff will naturally inform their group leaders of progress and significant developments; sales and marketing staff speak with their supervisors about new leads. But a knowledge management system will help ensure that that occurs consistently. It will also contain triggers and alerting mechanisms that draw out information that might not otherwise be forthcoming.

Equally importantly, a knowledge management system needs to have an element of self-sustainability. Management can urge or incentivize employees to use the knowledge management apparatus, but the greatest encouragement will arise when an employee finds something in it that is useful or vital to him or her.