Going With the New Project :
How Information Needs Can Drive Product Decisions
Associate Professor of Economics
Minnesota Mining and Manufacturing (3M) has had a company rule that 10 percent of annual sales must come from projects introduced in the past year and 30 percent of sales from projects less than four years old.
It was faithful to that policy even though such a rule might, in some cases, result in a less-profitable newer project being selected over an existing product with an established profit record.
The history of corporations holds a number of such examples, and Dongsoo Shin, associate professor of economics in the Leavey School of Business, is offering some reasons for this phenomenon. It is a part of his overall research emphasis on extending research on information-gathering issues within organizations using rigorous analytical tools and game theory.
In a paper titled “Information Acquisition and Optimal Project Management,” which has been accepted for publication in the International Journal of Industrial Organization, Shin argues that in many cases such decisions have a lot to do with the processes and incentives that govern a corporation’s acquisition of information.
“There needs to be a reward for the agent who gathers information on the new project to get the best information and report it truthfully,” Shin says. For the fact-finding agent, often a sub-unit within the company, that reward takes the form of “information rent.” In other words, having gone to great effort to get the information, the agent gets a payback on his or her knowledge, which becomes more valuable to the company.
But “information rent” can be provided to the agent only if the company moves forward with the new project, thus creating value for acquiring the information. If the information-gathering agent anticipates that the company is unlikely to choose the new project, there’s less incentive to aggressively collect necessary information (or aggressively engage in R&D activity).
Shin contends — and has developed a detailed mathematical mechanism to make the point — that this, in turn, creates an incentive for top company executives to significantly commit to the new project or technology earlier on, simply to get the best possible information. Having done that, they are unlikely to change their minds once a new project is chosen over more conventional ones. An example would be Sony’s insistence on using the Chromatron picture-tube technology in the 1960s, even when its planning team warned that it would be more expensive and less profitable than the tubes the company was already using.
Looking at the options, Shin concludes that the optimal process for acquiring information on new projects — that is, the one likely to generate the best outcome — is most likely to occur when the information-gathering task on a new project is partially integrated to the implementation task, thus partially linking the information-gathering reward to the final outcome.
The reason the partial mix approach works better is twofold, he says. If the tasks are totally separated, the agent who gathers the information never gets to implement the project, and so has no incentive to gather information in the first place. On the other hand, if information-gathering and implementation are always integrated (the fact-finder gets to implement the new project), the company is providing too much information rent. The optimal mechanism, therefore, is “partial integration of the two tasks.”
“As the top management you must commit to this mechanism,” Shin explains. “The moment you do so, the expected payoff for information becomes higher and cost of information becomes lower, and that drives the sub-unit of the company to pursue it and report it better.”