If your problem is small, viable solutions to that problem are vast. If your problem is large (scale), there's usually only one or close to only one way to solv

If your problem is small, viable solutions to that problem are vast. If your problem is large (scale), there's usually only one or close to only one way to solve it. So choosing the right path is important. Since you can't predict product market fit (you cannot predict value), you have to figure out how to iterate and test as close to continuously as possible. This last thing is what most people just can't do. Programmers included. So you can see there's a technology side and a culture side and they have to be well-aligned.