I've run across one too many articles and how-to-write books that tell me what to start the story process with.
Yes, if you have inspiration for a story that can be summed up in a one-sentence summary of the plot, you can start there. If you have a nasty antagonist who will give your protagonist real trouble, you can start there. If you sit down and devise your hero and then your heroine and then plot out according to the rules, and it works, you can start there
.Indeed, some writers write one sentence summing up the novel, expand it to four, etc. in a fractal manner. If you are one of them, that's one way to write a story.
But you start where you get the idea. Trying to force yourself to start from one place only adds unneeded difficulty to writing.
I've started with plots -- either retelling a fairy tale or ripping off one that the original author just didn't do justice to (and may even have made backstory). I've started with nifty plot twists or clever bits of magic. I've started with -- well, actions of characters is a better term than scenes or characters, because it can be a short portion of the scene, down to something that could be contained in a picture (especially, of course, if the original source is actually a picture), and can not fully explicate a character in so small a space. I've started with themes and abstractions (though those can be a nuisance), and I've started with titles -- and sometimes the title fit the finished story.
All of these have their challenges in that they all have something that needs development. The fairy tale plot needs elaboration; the ripped off plot needs to have the serial numbers filed off and then elaboration. The actions need an entire setting and plot to lead up to them, and lead away -- and some are inciting incidents and some climaxes, and still others must happen in the middle. In particular, characters have to be more complex than would be manifested in a single moment. An abstraction or title makes a moment look detailed with the need for elaboration, with characters and plot and setting all built up to match it.
Though I advise people to try to get as many ideas as possible to work together. It does simplify the invention process if a title and a plot twist and an inciting incident work together.
Then, this is a logical outcome of my typical inspiration. My ideas tend toward the bright shining moments, and the more of them I can hook together, the better, because the more that is carried by inspiration, the less the need for invention. There's no reason, even why it has to be a very dramatic moment in the story. One story of mine was triggered by reading a summary of a movie that mentions how the legendary hero's daughters meet him for the first time (while still children), and built up from there, turning the girls into minor characters, without dialog, in the process. Another turned on the notion of how to break a certain kind of magic in a nifty way -- notice that one doesn't even specify anything about the characters, the setting, or the plot except that the magic must clash (at the climax, as it turned out, so that was a dramatic moment). Nothing so broad as the overview. Indeed, many of them could appear in many different works, if only developed the right way. They need nurturing and additions, developing before and after, before it is clear what the over-arching story is.
Even so, I add a caveat to that advice: "as possible" is important. If the ideas don't work together, they don't. It harms the story to force in incompatible notions.
I also note that I have run across advice in various forums that start with "So you want to write a fantasy/superhero/hard science fiction novel." Frequently with a list of things you go through: superhero means powers and origin and backstory, etc. But --
Does anyone actually decide, "All right, I'm going to write a superhero story?" As the first thing? I suppose it's possible, but I have never heard of things except when chasing trends. (Don't. The time gap between when the trend starts and when you can actually get the story done is too wide to hit the trend.)
I generally figure out the genre from the thing I start with. I have a scene. I have a cool meta origin for superpowers. I have an idea for a magical book. I have a cool curse that I can put on a library. I have a beef with a popular but incoherent superhero trope. And so they get sorted between superheroes and high fantasy. Sometimes the starting idea is not sufficient to dictate genre, but then it meets another and the story crystallizes about them both. The more abstract an idea is, the harder, but I can manage because it's never so abstract as "A superhero novel."
On the other hand, maybe that's just me. If you start with "a superhero novel" and can develop it into a novel, you started with what you started with.
I was at the National Zoo, and I looked down into an enclosure and thought, "What if I put some zerglings in there? Could I keep them in?" (They are ferocious alien pack-hunters from a video game.) I went back to the hotel and started thinking about that. I needed a zoo-keeper, clearly. I wrote a first scene where her boss tells her she has to get the exhibit ready, and then she and I brainstormed the exhibit security.
I thought it was a short story, but they never are for me. I got to the "end" and it wasn't the end (in fact that was about 1/3 of the eventual novel, and boy was I surprised where it went next. 1/3 of a novel was never going to have been a short story anyway).
Years ago an agent told me he didn't want my fantasy novel but if I'd write SF, he'd look at it. So I know I'm not capable of writing a story "of a particular genre" with nothing else to go on: I spent quite some time thinking about that, to no avail. Evidently I had to go to the Zoo.