There will be a lot of learning and adapting, but it should happen in a staged approach; much of it before actually 'launching'. Launching generally means exposing your full product/site/whatever to the entire general public. Before doing that you should do smaller scare, more targeted testing.
Before launching, but after you have a prototype / MVP, you need to gather data on whether people want to pay for it. There are many ways to do this before a full launch. The methods vary depending on the type of product. Without further description of your product I can't help you further in terms of those methods. Send me a message if you'd like to discuss it further,
best of luck,
Having helped a few start-ups in different industries I can tell you that before you launch you need to have planed out everything that you can. Naturally you will not be able to plan for everything but you want to think through as much as possible. Having a business professional help you map out a business plan is the best way to start. This will give you a point from which to deviate for the first few years of operation and also show investors and lenders that you have a plan in place. Feel free to reach out to me and we can get your business plan started as well as take care of any other start-up needs you have.
Less than a formal business plan, you need what I term a "battle plan," that is high level vision,direction and set of objectives with strategies/tactics and milestones to meet them. This is more project-like and your focus of launching, learning and adapting is absolutely correct. And as you do, you modify your "battle plan" according to "battlefield conditions" but always with your eye on your high level targets.