"Everyone has a plan, until they get punched in the mouth" - Mike Tyson
The best thing you can do when preparing to transition from postdoc to PI is plan out 5-year research goals. Talk with your postdoc advisor about what projects are yours, think about what questions you are interested in, design experiments to test these hypotheses, and make lists of every construct you need to create or reagent you need. Even though everything around you will be moving quite quickly, there will actually also be some time as you are setting up the lab to just sit and think. We often get asked, if there were unlimited funds, what kinds of experiments would you perform? Starting a lab with a pool of undesignated money (startup) will likely be the closest you will come to this "do any experiment you want" utopian world. I didn't realize the full weight of this until my startup ran out, but having a pool of money for which you don't have to specifically justify each experiment a uniquely powerful situation.
It's great to hit the ground running with a definitive plan in hand, but always keep in mind that the real world can intervene. Many people I've sought advice from over the years have suggested the importance of having multiple lines of research within the lab at any given time. Don't be wed to a single question or system, especially in times like now when funding is tight. Keep reading and don't be afraid to try new assays or experiment with different systems. The early years of your lab, startup money in hand, may be the best/easiest time to branch out and ask completely new questions. However, the flip side of having multiple irons in the fire is that juggling experiments requires a skill not easily learned. While it can be quite easy to dream up the "next" experiment, it's often difficult to know when it's time to pull the plug on failed projects. Sometimes it's just a gut call. At least IMHO, knowing when to stop a particular line of research is one of the most intrinsically important skills for being a PI.
I started my lab with ideas for "easy" projects that would be straightforward extensions of postdoc experiments. After moving from North Carolina to Arizona, I realized that my bacteria and plants didn't behave the same in the dry air of Tucson as they did in soupy Chapel Hill. It was frustrating to say the least, and I was stuck with the decision to slog through and figure out a way to carry out these experiments or to cut bait and try a new direction. I moved the plant based experiments somewhat to the backburner, which is a bit tricky because I'm housed in the School of Plant Sciences, and decided to focus on investigating interactions between microbes. We just started reading papers and trying stuff, building off of research interests shared across all lab members. Looking back (over the last four years), there have been a lot of starts and stops, but I'm quite happy at how things are turning out. There are still experiments that I know I could get to work given more money and time, and strains sitting in my freezer for experiments I haven't come close to trying yet. These side experiments fail much more frequently than they work, but you have teach yourself to do the cost/benefit analyses to know the difference between when to stick it out and when to move on.
All of this in mind...a brief sidenote. As I mentioned above, you will never be as free to experiment as when you have startup funds. The tendency can be to bring in people (technicians/postdocs) to carry out exact experiments written down in your 5-year plan. Although this may work in many situations, I made a conscious decision to do the opposite. I tried to hire independent postdocs whose interests overlapped with mine, but who wanted to branch out into completely new research directions within the context of my lab's interests. At the outset I had no clue how this would go, and there were certainly some nervous moments. Looking back, I can honestly say that this plan worked out about as well as possible. We have developed numerous new research directions and both postdocs have also contributed greatly to more "basic" projects. For the young PIs, don't be afraid to leap in directions that are uncomfortable because you might just find yourself in interesting new places. For the postdocs, given the lack of funding opportunities, don't be afraid to find your way into the labs of young PIs. They will often have freedom to spend money whatever way they chose (unlike if you are brought in on a grant), and you might have a greater chance of developing your own independent research programs.