But you’re dangerously close there to having no predictive power. Business folk may rightly judge a product by its market success, but some business folk need t

But you’re dangerously close there to having no predictive power. Business folk may rightly judge a product by its market success, but some business folk need to decide where to invest resources in building new products. Hindsight is 20/20, and of course products that are widely viewed as having poor engineering quality can succeed and even dominate a market, for a variety of reasons. But that doesn’t provide much evidence to support the claim that, when developing a new product, engineering quality doesn’t matter. Sure, we can take the fatalistic view that new product development is an unguided process like biological evolution, where we will only know which random mutations were successful when we observe their prominence many generations down the line. But clearly most people act as if they believe they can do better than uniform random (i.e. investors generally don’t invest the exact same amount in every new company or product they can find).