Report, U.S. Small Business Administration, contract SBAHQ-05-M-0292 (2006).

Diana Hicks, Dirk P. Libaers (Georgia Institute of Technology)
Alan L. Porter, David J. Schoeneck (Search Technology, Inc.)

Published: 2006


Among the population of small firms, "serial innovators" are relatively rare, but appear increasingly important in an economy undergoing structural change. This study develops a taxonomy that captures key differences among the technology commercialization strategies of such firms. Existing classification schemes of innovative firms are usually built from large firm data, and so ill-suit highly innovative small firms. In this project, we studied 407 innovative U.S. small firms, across sectors, with at least 15 granted U.S. patents over a 5-year period.

We explore the potential for gathering information on firm strategy from their websites. Web addresses were located and verified, then 89 potentially informative keywords related to business models were identified, with theoretical bases. We built a special "Innovative Firms Application Programming Interface" (IFAPI) to use Google’s search capabilities. Using IFAPI, we obtained hit counts for each term on each firm’s website and normalized by size of website. Several iterations helped refine the useful terms. Factor analysis detected a pattern in use of the keywords, delineating factors that appear to reflect six distinctive high technology commercialization strategies.

The six business models identified for these serial patenting, small firms are:

  1. R&D organization or contractor
  2. Science- based product/service firms
  3. Highly specialized component supplier (high volume production)
  4. Specialized subcontractor firm (one-offs or very low volume production)
  5. Product solutions provider
  6. Service solutions provider (technical consultants)

The same analysis was run on a group of control firms, identified in Hoovers as small firms who compete with the highly innovative small firms. Factor analysis failed to converge on a solution; that is no underlying factors could be identified that explained the pattern of keyword use across the control firms. This supports the conclusion that high-patenting small firms pursue distinct business models to commercialize their technology.