You can ship a car with an engine that goes 0-60 in 10 seconds. Then you can ship the next car with an engine that does it in 7. If you try to develop an exact

You can ship a car with an engine that goes 0-60 in 10 seconds. Then you can ship the next car with an engine that does it in 7. If you try to develop an exact thing by an exact date, you're bound to either ship something that's broken on the date, after a lot of heart-ache, or to ship something that works late, after a lot of heart-ache. Then, good luck getting your people to make the next product better. All the cards were played; all the bridges were burned; and you got a shitty product. Instead: let's ship the best working car we can by <date>! And here's why that date. Who has ideas for how to get an even shittier completely working first-iteration done six months before that? Now, go figure it out and make it happen. I trust you. (To self: and I'll trust you even more as I see each shippable iteration that you enjoyed making.)