So, how much worldbuilding do you really need to do before you can start writing a novel?
The answer is, of course, it depends. The less you do in first draft, the more you will likely have to do in revisions. However, having some basic things sketched out can be an enormous help, especially since specificity helps increase meaningful word count. For that reason, then, I thought I'd lay out some worldbuilding parameters here, especially for those among my readers who will be starting NaNoWriMo shortly! Remember - you can put an enormous amount of effort into worldbuilding (and for that I recommend my series of worldbuilding hangouts), but you can get a great deal of excellent effect up front with a small amount of concerted thinking.
1. Do I need a map?
Maps can be very helpful in several ways - they can get you in the mood to write in your world, and they can help you with logistical questions like, "If my characters have to get from A to B, on horseback, how long will that take?" Even if you're not working with a questing party, having a general map of the layout of the city or town or village where the story takes place will help you be able to describe your characters moving through it more fluently, and reduce the possibility of strange errors. I ended up mapping the suite where my noble family lived, down to the furniture! Because I had to fit a lot of different rooms into a relatively small space, and know where one room was relative to the others (and make it all fit!), the map became really important. You can, of course, write away and figure out your maps later, but in the interest of increasing word count in a non-fluffy way, I recommend a bit of mapping before you start. Making your descriptions of travel and motion more accurate makes the map a good investment of time.
2. How much setting?
My first recommendation is to have a sense of the world climate, and especially the local climate to where your story takes place. Choosing a climate won't just give you general weather conditions, and details on how comfortable it will be to storm the castle, or walk 50 miles, etc. (though those things are useful!). It will also give you a sense of what the local vegetation will look like, and what people will grow on what kind of farms, and thus what people will eat. It will also let you know what kind of building materials are available, and that will help you know what the local architecture looks like. If you don't deliberately choose a climate up front, then there is some risk of inconsistency between, say, food and weather. You may find yourself having to go back later and realign all of the details of agriculture, architecture, weather, road conditions... Since climate has such an enormous influence on what you're describing, and reaches into so many areas of life, it's worth setting this one up early.
3. Do I need to figure out the economy?
The short answer is yes. You should really know where people are getting their food from, and their marketable goods from, and what kinds of things are basic and what are luxuries, etc. In terms of working quickly, this is one where you can often work with a sort of "default" economic setting, by deciding whether your economy is like a complex capitalist economy, a medieval feudal economy, a form of socialism, etc. This leads me directly to...
4. What's my technology level?
If you're planning to work fast, this is another area where default settings can be very helpful. Pick a period of human history and use its model for a sense of what kinds of tools people will possess, how valuable they will be, and how often used. Bronze age technology, Iron age technology, Industrial revolution technology, computer age technology - each of these has sets of expectations that readers (and you!) will be able to keep track of relatively easily. Pick one, to allow you to orient yourself. If you want to change things, and work in a way that doesn't follow the easy defaults derived from our own history, you're looking at a longer-term project. It might be good to use defaults for now, and save alterations from those defaults for a later draft.
5. What's my social structure?
For me, social structure is a large part of what drives the conflict in my stories. This could be why I have to take so long to write them! When you're embarking on a novel, it's good to think through some basic stuff about how your people work. Do they have a monarchy? What is the government system like? What activities are valued? What kind of achievement might allow someone to rise in social standing (military success, etc.)? Think also about how many different cultural groups you're working with. Is it just one? More than one? Either way, large-scale cultural groups will come with subgroups inside them. For the purposes of a quick novel-writing experience, keeping the social structure to broad-brush strokes is probably the best idea, and then you can use your notes on the smaller details of social value and interaction later. You may end up unconsciously writing stereotypes if you're really in a hurry. That's not a problem, so long as you take a closer look at social nuance later.
This - social structure - is where you're first going to encounter your characters in your worldbuilding. Given that characters are the best way for social structures to demonstrate themselves, it's a good investment of time to sit down and figure out the explicit connections between your world's social structure and your characters, their resources, their assumptions and behavior. This again is going to increase your ability to capture specificity, this time in character behavior, and also in...
You may not want to get fancy with your voice or dialogue before you jump in to a NaNo project - being experimental takes time. But whenever I want to experiment with voice, for prose or for dialogue, I do at least one (and sometimes more than one) experiment beforehand. If you come out of your social worldbuilding with something in mind, try a micro-scene to test out whether the voice you've imagined is sustainable, especially at high speeds. I don't recommend plunging headlong into a voice that will take you half an hour to get your head into every time you sit down (yes, I have done this. I'm not really cut out for NaNo)! Dialogue is probably the easiest thing to fix or restyle later.
What if I don't have something figured out in time?
Use defaults. Chances are, even if you haven't fully thought things out, you're using some sort of defaults from our own world already. The patterns of your story will tend to conform to your expectations of what might happen in your own life. It's really not a problem to use a default technology setting, for example, and go back and retool later. That's what revisions are for!
Overall, I would say that going through a few basic elements of worldbuilding before you start will help you sharpen your focus and will help your wordcount stay high (being vague doesn't require many words). Be sensitive, though, to the point at which you start spending more time for less result. That's when you should slow down and say, "I'll save these details for draft 2."