Opportunity Spaces in Innovation: Empirical Analysis of Large Samples of Ideas

A common approach to innovation, parallel search, is to identify a large number of opportunities and then to select a subset for further development, with just a few coming to fruition. One potential weakness with parallel search is that it permits repetition. The same, or a similar, idea might be generated multiple times, because parallel exploration processes typically operate without information about the ideas that have already been identified. In this paper we analyze repetition in five data sets comprising 1,368 opportunities and use that analysis to address three questions: (1) When a large number of efforts to generate ideas are conducted in parallel, how likely are the resulting ideas to be redundant? (2) How large are the opportunity spaces? (3) Are the unique ideas more valuable than those similar to many others? The answer to the first question is that although there is clearly some redundancy in the ideas generated by aggregating parallel efforts, this redundancy is quite small in absolute terms in our data, even for a narrowly defined domain. For the second question, we propose a method to extrapolate how many unique ideas would result from an unbounded effort by an unlimited number of comparable idea generators. Applying that method, and for the settings we study, the estimated total number of unique ideas is about one thousand for the most narrowly defined domain and greater than two thousand for the more broadly defined domains. On the third question, we find a positive relationship between the number of similar ideas and idea value: the ideas that are least similar to others are not generally the most valuable ones. This paper was accepted by Lee Fleming, entrepreneurship and innovation.