Three value creation models in the OR field [dual]

I a previous primal post, I have described three generic types of value creation configurations: the shop, the chain and the network, as characterized in Stabell & Fjeldstad (1998). In this post, I apply these concepts to key players in the field of Operations Research. While names used are fictional, some players may recognize themselves in these archetypes.

Prof. Smith

In many ways, a professor is much like an entrepreneur. He serves multiple customers in multiple markets: (1) he designs, manages and  provide lectures for undergraduate and graduate classes, (2) provides individual tutorship to graduate students, (3) delivers cutting-edge research to the academic world through publications in journals, and (4) sometimes do consulting ($$$$) or technology transfer ($) work with companies. Some would also argue that the professor fills bureaucratic paperwork for the funding agencies and the offices of the deans. The only common thread to these product-markets is the professors’ individual expertise, as well as those of his small team of graduate students and research fellows. Everything done in the lab runs through the professors’ processor and memory. A typical professor’s research provides a state-of-the-art algorithm for a specific problem. Whether he provides the source code or not may depend on his willingness to use it in consulting (4) too. But as these applications are powerful, they tend to be tailor-made for each problem, thus limiting reusability and portability. Prof. Smith is a paramount example of the shop business model.

GuroPlexPress (GPP)

GuroPlexPress is the state-of-the-art mathematical programming solver, resulting from the merging of several key players in the OR industry. It solves everything, from linear programming to quadratic and conic and beyond. It can solve problems with zillions of variables and two zillions of constraints. Its beautifully designed console interface, API and IDE allows you to to model just any optimization problem as long as it fits in the modeling language formalism. GPP then outputs a solution in a standard format. Although GPP is often bested in journal papers by Prof. Smith’s custom algorithms for very specific problems, it needs to be able to solve very different problems. It is also noteworthy that it makes more money out of the market than all Prof. Smiths combined because the vast majority of optimization problems either (1) do not require the degree of customization and punch power of Prof. Smith’s algorithms, or (2) require adding “yet another constraint on the fly but just this once”, and Prof Smith’s algorithm just can’t handle that. GPP is an example of the value-adding process (VAP) business model.


StackORverflow is a place where students and practitioners can learn the tools of the trade of becoming an OR specialist. You have a question, someone answers it, then one day you are on the other side of the fence: someone asks a question, you know the answer and pay it back. StackORverflow is merely a platform: it does not provide content by itself, but it allows for different customers and suppliers of knowledge to meet and exchange value. As StackORverflow is not-for-profit, you earn gratitude and reputation rather than cold cash for providing value to others. The more people use the network, the more profitable it is for each members: the customer has a higher probability that someone has the knowledge he seeks and has time to answer his question, and the suppliers reputation and visibility increased. StackORverflow is a paramount example of the facilitated network model.


  1. Good post. The natural evolution of Prof. Smith is grow, hire others Profs. Smiths and transform his business model in VAP or network.

Speak Your Mind