Your problems usually should be measurable, and sometimes %80 of the work is creating systems that can measure it. By thinking of how do I solve the problem fi

Your problems usually should be measurable, and sometimes %80 of the work is creating systems that can measure it. By thinking of how do I solve the problem first, instead of features, you might skip making a lot of stuff that is unnecessary in actually solving the problem. One example is application startup time. Or revenues. User survey scores and so on. The problem roadmap informs the feature roadmap.