The first question I ask myself, as an engineer approaching a new problem, is "How can I cheat? What's the 90% solution that will take 10% of the time?" (The next question is, "Is 90% enough?").

I'm always surprised by engineers who don't think this way.