I've always subscribed to the idea that your players should be the most interesting characters in your game, and this advice applies to them as well. Starting from a simple plot creates an opportunity for your game to grow in directions that reflect you and your players.
If you want a practical way to go about this, consider the Underpants Gnome school of adventure design.
Satire aside, the 3 step plan is useful for almost any plot. Start with a villain, whoever it is, and give them a plan that really is as simple as:
- Do something simple
- Do something complicated
- Achieve goal
This is usually easiest if you start from the goal, since that tends to suggest the previous steps. With that in mind, I strongly suggest a concrete goal - "power" (or even "profit") tend to be so amorphous as goals that they don't really suggest a course of action. If a goal of that sort is what you're looking for, then try to pick some manner of specific implementation of it, like leveling up or stealing a particular treasure.
This process is made much simpler if you embrace the cheese. There is a natural inclination to try to make the plots smart, coherent or clever, but realize that a lot of great plots have almost embarrassingly simple underpants structure. Y'know - Take Ring, Throw it in a Volcano, Free Middle Earth. Look at that example and consider how far short of the true complexity of the story that falls - the good parts lie in that difference.
Thus, start with something like:
- Kidnap Orphans
- Sacrifice them to Orcus
- Gain Undead Army
On paper, this looks like the basis of something pretty cheesy, but it need not be. Challenge yourself and consider how this framework might make for a good story. The villain might be interesting, the orphans in question might have compelling stories, the sacrifice might require all sorts of logistics to pull off, maybe the use the army will be put to is interesting. Whatever. The point is it can be done.
The trick is that you don't need to solve all of the problems up front. The underpants plan should seem unworkable on the face of it because it leaves unanswered questions. Answering those questions is a driver of play.