Skip to content

Creating a database for the network analysis of global business leaders

Finance session

Dominick Sutton (Boardex)
Creating a database for the network analysis of global business leaders

Network theory and analysis underpins the creation of a database and associated business products that are used by many of the leading investment banks, executive search and professional advisory firms to support key business activities and decision making. The database grew from a product originally targeting corporate governance and which had a focus on corporate leadership analysis. The database is of very high quality and is used by many leading universities in their research of, inter alia, business leadership networks.

There are many services that provide consolidated data on a variety of business individuals. Most of these are based on either web scraping or subject-contributed data. This presents many difficulties with respect to data veracity, accuracy, completeness and timeliness. Owing to the critical and strategic nature of the data and the analysis use by clients such an approach is not acceptable. The data must be of extremely high quality, as any failure to meet these high standards can cause significant revenue loss to the clients and an abrupt termination of contracts with the company.

This high quality requirement for a global database presents many challenges. These include language and national institution barriers, access to up-to-date data, difference in corporate compliance standards, the vagaries of the Data Protection Act, and differences in approach to personal data disclosure. In addition it is necessary to account for the risks of co-mingling data on different individuals who have the same name as well as separating different parts of one individual’s profile owing to different international or institutional naming conventions.

A key part of the database process is building in high quality at every stage. This starts with the selection of reliable data sources and the continual monitoring of sources of new data. Ensuring accuracy requires the creation of strict business rules regarding the handling of different data classes as well as defining standard operating procedures. In addition, high standards of accuracy requires a rigorous quality verification process that is driven by statistical analysis of the database, of the different types of data, and a methodological framework that can allow for the immense variety of individual data. Ensuring completeness is a combination of analytical rigour, data collecting procedures and understanding of different classes of data subject. Timeliness is also dependent on reliable data sources but also encompasses strict rules on accuracy, as updated information is often the most valuable to clients and therefore also the most risky for the company.

The use of network analysis is key to developing this database, not only in its applications but also in the selection of the data subjects for inclusion, the types of data collected, and the management of the internal data processes. Network-based output is also a fundamental part of the current data offering. Its integration into client-based systems allows them to integrate their internal systems with the company’s and so take advantage of its proprietary analytical processes. Extending the reach of a client’s internal data is a very attractive part of the offering, but also one that presents its own challenges. These include data concordance in the absence of a universal key as well as the need to maintain this linkage in the face of dynamic changes to both the company’s and clients’ databases.

Future projects include the development of a greater understanding of the dynamics of this specific type of relational data by means of statistical analysis and probability modelling. Creating an index to classify the network of an individual in such a way as to allow other business applications to interact is also considered a key strategic aim and is one currently under investigation.