I had a reader ask questions about revision, and since that's the writing phase where I'm currently mired, I figure it's a good time to talk about it. Maybe I'll learn something!
First, a disclaimer: There is no one right way to deal with this. You have to find the way that works for you. I know very successful (as in bestselling) authors who revise as they go -- they edit the previous day's work at the start of a day's writing session, and if they come up with something that needs to be fixed earlier in the book, they go back and do it immediately, so that when they reach the end, it really is the end. That terrifies me because I'm never entirely sure what the book is really about, no matter how much plotting and planning I do, until I've written the first draft. So, if you've found your method and it's working, then ignore me. If you're still trying to find your method, then you might want to give my suggestions a try and see if they work for you.
The first thing I recommend is putting a book aside and letting it rest. It's difficult to change things when you're still emotionally connected to the work, when you remember why you wrote something a certain way. If you thought something was wrong in the first place, you wouldn't have written it that way. I think this is especially important with a first book because there's so much glee at finishing it that you don't want to change it. The idea is to get to the point where you can look at the book objectively, like you might look at something written by someone else and without a lot of the baggage about the book that's in your head but not on the page. I'm always running into jokes I wrote that I thought were funny in the first draft but that I don't get at all when I come back to them because I don't even know what I was thinking, which means readers would likely be scratching their heads, too. How long this takes varies. I prefer to work on something else between drafts, like write a first draft or revise another project to really clear the book from my mind, but whether that's feasible depends on deadlines. If it took you a long time to write the first draft, you might be able to go back to it fairly quickly because you may have forgotten the beginning.
The second pass should focus on the story. Until you've got the big-picture plot fixed and the scenes in the right place, there's no point in fine-tuning and wordsmithing, and if you are fine-tuning, you might not even see the big picture.
For new authors, I recommend doing a read and taking notes before you even try to fix something. Read the book like you're a reader, and if something bothers you or if you have a question about it, make a note but don't try to edit. Once you have all your notes, you can come up with a plan for your revisions, whether it requires reworking some of the main plot or just fixing some scenes.
How do you know what to fix? You can learn that over time. It may help to get critiques from someone more knowledgable and experienced than you are. Once you see what changes they suggest, you may be able to spot those things for yourself. I seem to channel my agent when I revise, so I can already tell what she'd criticize. You can find online and in-person critique groups, find critique partners through writing organizations, enter writing contests that offer feedback or even pay for a professional critique (though be very careful about checking credentials and determining exactly what you're paying for and what you'll get, because there are some scams out there). If you're paying for the service or entering a contest, I'd recommend getting the book to the point where you don't think you can make it any better before you submit. If you have an ongoing critique relationship with a group or partner, you might be comfortable with getting feedback on a first draft. I'm more likely to get it to the point I think it's perfect before I let anyone else see it. I may discuss problems while revising if I'm stuck, but I don't hand out the text itself until I reach a certain level of "done."
Here are some things to look for when evaluating a manuscript for revision:
* If you find your attention waning or distracted when you're reading your own work, that's a bad sign. Rule out things like hungry, tired, or a parade passing by your house, and then take a look at why your attention drifted. If your attention drifted, then a reader is likely to put the book down at that point and go do something else. Or that may be when an editor or agent gives up and rejects.
* Have you inadvertently established a pattern? Sometimes, patterns can be used for effect -- things happen a certain way often enough that you establish an expectation, and then you can create surprise by changing the pattern. But sometimes you do it without meaning to, and it makes for a boring book. When my agent gave me the feedback for the second book in my series, she noted that almost all the conversations between the characters happened when they were going to and from work together. Half the book seemed to involve them walking to the subway station while planning what to do next. I needed to cut a lot of the conversations because characters should be doing things more than they talk about things, and I needed to mix things up for the remaining conversations. You can also run into this when the plot becomes "nearly be caught by bad guys, escape, lather, rinse, repeat."
* Does everything that happens have some reason, especially some reason involving the main plot? Do you have scenes that are just cool rather than serving a purpose? You don't necessarily have to cut them. You can find a way for them to have a purpose. I came up with an entire story thread that ended up tying the whole plot together in the fifth book in my series just to justify a scene I really liked that was just there to be funny and suspenseful. If I'd cut the scene that was cool but had no purpose instead of finding a way to justify it, I'd have undermined my plot instead of strengthening it.
* Take a step back from the book and think about what these characters would really do. If the villain is mostly offstage and we're just seeing the results of his work, what is his plan, really, and how do events fit into that plan? What would the villain's response to the hero's actions really be? Would the hero really take that action? Or did you write it that way to nudge the story in the direction you wanted it to go? Do you understand the characters' actions, and is that clear on the page or just in your head?
* Are the plot turns too predictable? Can you come up with anything more interesting or surprising? I frequently hear the advice to make a list of things that can happen and then throw out the first ten you come up with because those are the things readers are most likely to expect, but whether that works depends on how your brain works. If you always think out of the box, the first thing you come up with may be something no one else would ever think of. But it's still worth thinking about. Have you relied too heavily on a tried-and-true trope with a predictable outcome? Can you come up with a fresher, more interesting way to do things, or can you subvert the trope in some way? Feedback can be helpful here. Give someone the setup and ask what they think will happen next. If they come up with what you've written, you might want to change it. Or you can use the fact that it's expected. In one of my books, I thought the identity of the villain would be a surprise, but it turned out to be way too obvious. Instead of changing the villain, I let the heroine notice the villain and came up with a reason no one but the heroine seemed to notice who the villain was, and then her frustration about that led to a lot of conflict and some suspense.
There have been whole books written about revision, and there are entire workshops on the subject, so this is just a starting point. It really is something you learn by doing. The more books you write and revise, the better you'll be, but you'll never stop learning or revising. Once you instinctively avoid writing the stuff you'd have had to revise, you'll start noticing less obvious things that need to be fixed.