October 8, 2002

Central Intelligence

In life sciences as in other industries, sharing integrated business and competitive intelligence is a valuable competitive advantage

by Adrienne Tannenbaum and Elby Nash

Continued from Page 1

Thus, the successful life sciences BI environment depends on an underlying:

• Structure — a defined set of search starting points with interrelationships as appropriate, all well accepted within the internal life sciences organization. Retrievable components should be natural results.
• Structure-based access — Using the underlying architecture, access of the BI-CI component should be second nature and not require substantial thought as to where target information is or how to find it.
• Consistent expansion — The BI-CI environment is never stagnant. Fully implemented frameworks include processes for incorporating additional information on an ongoing basis, and expanding the metamodel or its instances as necessary. In a BI-focused organization, these roles move to the research librarian and require Internet savvy, often in support of standards such as Dublin Core, which specifies XML tags for documents that are to be retrieved over the Web.

Cost-Benefit of BI-CI

How does a life sciences organization decide if a BI-CI project will pay for itself? As with most financial analysis, expense and revenue amounts must be determined and compared. The expenses of implementing a BI-CI environment can be easily measured. Dollar amounts for expense types such as internal and external staffing, software, and training will need to be gathered and used as a clear projection of the project's total cost.

Benefits and revenue are not as easy to quantify. However, because a significant financial impact on the business (missing the chance to introduce a product, spending too many research dollars on an unsuccessful result, misinterpreting financial data and falsely reporting revenue results) usually triggers the desire to institute a BI-CI project, some anticipated dollar benefits are usually predetermined.

Despite the intangible revenue number that may be in mind, inefficient processes or the amount of time expended in information identification is usually at the heart of the experienced dollar loss. Again, the benefits of a well-implemented BI-CI environment are both cost avoidance and additional revenue generation (when the development of new products results directly from the use of the BI-CI environment).

Consider assigning expense to the following activities and using the total as a cost-avoidance amount:

• Locating information: All time spent identifying potential sources of information, accessing (or attempting to access) these sources, as well as acquiring external data has a cost. Staff expense, specific software purchases, the purchase of external data, and so on are all costs that exist before the roll-out of a BI-CI environment and are therefore "projected savings."
• Interpreting information: When information is found, it is often misinterpreted, or not interpreted at all because its true meaning is not derivable. Follow-up research, often used to validate the located information, includes phone calls, meetings, questions, and required further analysis. More important, the results of misinterpretation are often extremely costly. Product recalls, excessive development costs, and late market entry are all attributable to the lack of an organized BI world.
• Integrating and reorganizing information: Without reuse, the same information is often recreated and kept in an isolated location. Woe to the new employee who finds out after months of work that not only was the same work already performed, but the already achieved results were available within a company database.

Finally, many intangible benefits result from the well-implemented BI-CI environment:

• Improved understanding of current and future competitors (the latter being most important)
• Improved knowledge of complementary technologies that may be available for license and bundling with a company's own product
• Knowledge of other scientists with complementary skills who can help a company develop its products
• Existence of market studies in the product or technology niche that a company occupies.

Critical Success Factors

Undertaking the development of a BI-CI environment requires some rules of thumb. Before approaching this effort, it is essential that the scope and objectives be clearly defined and agreed upon. One common characteristic of failed effort is the attempt to do everything at once, so objectives should be phased so that functioning components of an overall environment can be delivered in no longer than six-month increments.

The BI-CI undertaking must be based on business issues rather than a desire of technology professionals to use the latest and greatest technologies. Another characteristic of failed efforts is that the establishment of the BI-CI environment was triggered within IT, was subsequently developed and delivered, and attracted no business users. These projects are the first to be cancelled during budget trimming.

From a technical perspective, the following characteristics are also critical to the successful functioning of the delivered BI-CI world:

• Categorization and use of information and data internal to the company before seeking external sources of information (up to 70 percent of the information that is needed by a CI-BI function can be found inside the company)
• A published "information governance," model that describes the roles, responsibilities, and decision rights regarding information and data throughout the enterprise
• Creation of a metadata structure that facilitates the integration of requirements-based internal and external information for ease of access and analysis.

No More Excuses

In today's highly competitive cost-conscious world, life science organizations — and organizations across all other industries — are looking internally for assets of value. There is never justification for the re-creation or purchase of data and information that already exists. Furthermore, there is no longer an excuse for missing the boat because the supporting statistics were simply "not available." The solution to preventing these costly inefficiencies is a well-planned BI environment.

INFORMATION IN ACTION

An excellent example of a BI-CI environment in environment in the life sciences industry is Eli Lilly's use of patent co-citation clustering as a tool to analyze the major scientific and technology frontiers that Lilly is pursuing. Patent co-citation analysis is a technique that employs sophisticated text analysis with data visualization software to identify relationships between patents. By doing so, a company can answer these questions:

• Where should we be spending our scarce R&D dollars?
• What innovation competencies do we have that we may not be aware of?
• What intellectual property do we have that other companies find so interesting or useful that they are cited as relevant prior art in their own patents?
• How can we use this information for competitive advantage?

The key to this process is a detailed visual map of intellectual property as it is represented in both its own and public patent databases. Clusters represent areas of similarity between the originally filed patents and any subsequent ones that cite the original patent (for example, for an osteoporosis drug such as Evista, patented by Eli Lilly in 1997, and a competing product such as Fosamax, patented by a competitor in 1995). Clustering analysis also looks at self-citation — instances where patents filed by a company cite its own patents. Here potential reuse of internal knowledge or areas of special competency can be identified.

Rate This Article Rating: 1 (best) 2 3 4 5 (worst) Comments: Optional e-mail address:

Lilly's co-citation clustering analysis examined approximately 2,800 U.S. patents issued to Lilly between 1975 and 1998. Lilly was particularly interested in the self-citations (relationships among its own patents). Because the research that accompanies the development of new compounds concerns the identification of known side effects resulting from any of the compound's molecules, the citation of an organization's own products makes it obvious that intellectual capital exists, is identifiable, and is being reused.

Lilly's specific investigation targeted the identification of disease conditions for which estrogen was typically used, such as osteoporosis. Upon discovering citations to seven key patents filed by Lilly in the early 1980s, a number of new drug compounds, among them raloxifene (the active ingredient in Evista) were developed. More important, when this clustering was identified, continued cyclical research yielded 139 other internal patent citations during research on other estrogenlike compounds that were potentially useful as antifertility or breast cancer drugs.

This particular case study demonstrates the importance of metadata and its effect on the organization of patents, almost 3,000 in this case, into knowledge groupings. This organization becomes automatic when the proper metamodel is developed, and there is no prerequisite need for an expert to read and classify each of them.

Lilly's approach employed software that captures the internal structures of many patent and patent citation databases. As a result, not only was intellectual capital identified, but non-IT professionals obtained a deeper understanding of the relationships among these patent-specific databases. The relationships were remodeled and organized, via their metadata, into a portfolio of intellectual property that could be developed and exploited to yield even greater business value over time.

Adrienne Tannenbaum [atannenbaum@dbdsolutions.com] is the president of Database Design Solutions Inc. (www.dbdsolutions.com) a Bernardsville, NJ-based consulting firm.

Elby Nash [enash@taratec.com] is an executive VP of Taratec Corp. (www.taratec.com), responsible for information and knowledge management services.

• Privacy
• Your California Privacy Rights
• Copyright © 2008 United Business Media Limited