[phamilton]

GSD for us is less of a management tactic and more of a trait we find in engineers. Some engineers spend too much time in the design phase, planning out features nobody is going to use. GSD is about finding the fastest route to a solution. The quality of the solution is orthogonal to GSD. Bad GSD results in spaghetti code. Good GSD means providing the minimal solution without painting ourselves into a corner. It means a clean interface and set of behavioral tests with perhaps a hackish library implementation that can be pushed to production. Building out that hack vs taking the time to "do it right" can mean 10s or 100s of thousands of dollars of revenue. Sure it builds up a bit of tech debt (which we resource immediately), but that's often well worth it.