One of the many writery things I did at Swanwick was attend a four-part course on building a novel, by Emma Darwin. She writes historical literary fiction, so some of what she said was counter to what I understand of writing genre (which is often not character-driven), but it was interesting to see things from a new perspective.
The first sessions were about characters and plot. We did a group exercise, and we created a character using a spider diagram. Emma put the name "Jan" in the centre of the board and people shouted out ideas. What we ended up with was a gay, transvestite, Polish, rapping, piano playing body-builder, with a tattoo, a chihuahua, and an unexplained wedding ring. (At this point I started to wonder if some of the more straight-laced attendees were shouting out all the things they didn't dare to put in their own fiction.)
When we meet a character in a novel we know nothing of them, just like in real life. We intuit what they're like by what they do, and what they do is driven by their needs - love, security, whatever. (In Jan's case it was to improve his English so he could get a job and feed his chihuahua.) The key to a character is what they do and why, and what they do drives the plot. Of course, it's never as simple as characters getting what they want. The stakes need to be raised - what gets in the way and what's the effect of it? If the character doesn't get what they want things should be different even if they're the same, if they do what's the price and is it worth it?
One suggestion Emma made was that if a scene won't work, try writing if from a different point of view, or switch from 3rd to 1st person or vice versa - if it still doesn't work it doesn't belong there. Looking back, I can see dozens of times when doing this would have helped - or I did it but was too stubborn to admit what the scene was trying to tell me.
Emma also mentioned "The Thirty Thousand Doldrums". Apparently it's very common, whether you write short novels or long ones, think in chapters or don't, to get stuck at around 30k. She also suggested that this is a good time to stop and produce some kind of outline, whether you're an outliner or not, to give you an overview of what you've got and where it's going. It could be that your novel's no longer doing quite what you thought it was.
She provided a handy example of the structure she used when writing The Mathematics of Love to keep everything organised, a grid with the chapters down the side and each element with its own column against them (also useful for keeping track of information you need to know but that doesn't necessarily appear in the book). She goes into more detail in her blog post here, and also how she thinks about novel structure in her post "Building the Bridge". Anyway, this post is getting be quite a bit longer than I expected, so I'll finish it in another sitting.