Over 60% of new product efforts are scuttled before they ever reach the market, and of the 40% that do see the light of day, 40% fail to become profitable and are withdrawn from the market.
Christensen points out that though the new-product failure rate is high, 'failures are not really random.' They are a result of the difficulty of the task: How to connect disruptive innovations with the right customers to create a foothold in the market, then grow profitably along the sustaining trajectory. And identifying those disruptive footholds means 'connecting with specific jobs your customers are trying to get done in their lives.'
Interesting discussion about market segmentation, which he defines as the 'categorization stage of theory building.' And here's correlation vs causation again - according to Christensen, "attribute-based categorization of either/both products or customers can reveal correlations between attributes and outcomes - but only "circumstance-based categorization (ie., segmentation schemes) tell causality" what features, functions, and positioning will cause customers to buy a product.
In other words, customers "hire" products to do specific "jobs," so it's best to segment the markets to mirror the way customers experience life. The critical unit of analysis is the circumstance, not the customer, which to me suggests qualitative, not quantitative, research. My instincts tell me this is right. And it actually also "fits" with the way we already structure our ideation projects, so that makes me all the happier about it!
Bottom line: One disruptive strategy is to compete against "nonconsumption" for 'nonconsumers." Traditional quantitative market research won't identify these folks or the jobs they are trying to do. The best way to determine this market is to observe what people seem to be trying to do, then ask them about it. And only after you have identified those needs would you then move into quantitative research to determine the size of the market. Until you know what's needed, you can't figure out how many people might have that need.
Corante :: Tech News Blog