Charlie O'Donnell makes a great case for why it's useful to plan even when you don't know where you're going:
When we launched Path 101, the startup I worked on for two years, my board suggested building a traffic model, including a detailed funnel. That felt pretty pointless because I was taking a shot in the dark--trying to figure out how many questions a user might answer or percent of users that would upload a resume. What Hunter Walk, our board member and investor, said was that it wasn't about getting it right, but it was to anchor the conversation about the results--and to give ourselves a frame of reference to think about it. It forced us to but together our best guess of what would happen, and when that wasn't the case, it helped us structure the conversation of how reality turned out to be very different. We realized that many more people came to the site through the Q&A section than we realized--and so we redoubled our efforts to convert people from that section to the personality test, which also converted higher than we thought. With this framework, we could continuously measure our results and test our hypothesis.
A lot of people try to launch and just see what happens. Inevitably, they wind up in "screw tightening mode", which is what I call it when there's no overarching strategy for where the product needs to go, so you just wind up tinkering on a lot of small things.
This is a classic startup fallacy: "ship it and see what happens." Whenever you use this plan, you are guaranteed to succeed - at seeing what happens. Unfortunately, if you cannot fail, you cannot learn.
Another key point of business planning, which Charlie mentions, is to enable you to plan milestones (working back from that rapidly approaching point when you run out of money) and to see if your current plan can even theoretically get you off the ground.
If you read this far, you should follow me on twitter here.