In case the title is confusing, I will say that this is a method of world-building which depends on little, tiny threads braided together at a time, or bricks of many different kinds piled on top of each other, or dozens of different kinds of food made into a meal (choose your favorite metaphor for writing). It does not rely on big honking tapestries and walls and meals that are just descriptions, or exposition, or history infodumps, or scenes of character introspection, or circular conversations. Some people can work like that. I can’t, and I also prefer to read books that don’t; while a book with big honking tapestries or walls or meals of [insert writing technique here] might have plenty of virtues, I consider the big chunks a neutral feature at best, not a recommendation. To add yet a fourth metaphor, I prefer books that are like Arabians rather than Clydesdales.

1) Make casual mentions mean something.

A favorite trick of fantasy is to toss in a casual mention of a “Dhozkendrian carpet” or “Filandrian wine,” or to show ships unloading bales of cloth “from Far Easternmost,” or to have a crowd scene with “dark-skinned foreigners from Arillliith, gold-ringed dancers from the Darkened Hills, women from Far Westernmost with corpse-pale cheeks and hair red as rouge.” This is fine, as far as it goes, and the mentions can get accepted and enjoyed as minor tricks of worldbuilding.
But what if they weren’t just mentions, just tricks? What if you took the opportunity to actually put Dhozkendria, Filandria, Far Easternmost, Arillliith, the Darkened Hills, Far Westernmost, on the map? Then you took the opportunity to come up with histories for those people, religions, languages, weapons, social relationships, that weren’t solely aimed at producing a convenient trade good or reflecting one stereotypical appearance? Then, when you needed a random group of bad guys to bust into the narrative, they don’t have to be random at all, or yet another ninja school/assassins’ guild/political cabal crowding your already crowded (I bet it’s already crowded) domestic political scene. You know they’re from Arillliith, and they want to kill the hero not because he’s the answer to their prophecy or for some random, ill-explained reason, but because his trading success just made a bunch of people in Arillliith lose their jobs, and they take that pretty seriously. He’s a criminal in their eyes, and they’ve hunted him down. Now they want to capture him and take him back to Arillliith to stand trial.
Return to the small mentions. Build them up. Have them exist and get tugged into confluence with the rest of the world (see points 2 and 3). Throwaway lines hinting at greater wonders can help, but if you feel the need to turn and explore those greater wonders, it helps even more if you’ve built them up some and not just tugged another set of clichés out of thin air.

2) Remember that no groups, institutions, or organizations exist in isolation.

They may be isolationist, but that is not the same thing. They will have allies, neutral acquaintances, enemies, people they are watching in case they become enemies, groups they were specifically formed to act or react against, trade partners, rivals. Group and institution interaction can be great for building up your world, not to mention your plot. Rather than basing all of the action on the shady past between two churches that had a schism, while the other churches of the world sit around untouched, consider how they act on each other. New ideas will promptly come springing out of the woodwork.
Does this take work? Yes. Does it lead to complicated plots that often serve longer fantasies best? You betcha. But there’s an easy solution to not letting it get too complicated: just restrict the scope and setting of the story instead. If it deals only with the events in an isolated village, there may be no need to bring the Arillliith church from a thousand miles away into it, any more than people writing murder mysteries set in Oklahoma usually need to consider what effect the consistency of beach sand in Hawaii would have on their plot. What I do think is silly is pretending to create a story that embraces the whole world, but really only embraces a few sharply defined actors, or one sharply defined institution rife with the usual human error, while its opponent is absolutely evil, and the other groups are shadows and plot devices. If you do mean to build a world, build a world. And part of that is considering what stakes the groups you’ve created are going to naturally have in what’s happening, whether or not you want them to have them.

3) Treat human/human-like societies like an ecosystem.

This is a metaphor, of course. It’s a useful one. Many people are now comfortable with the idea of an ecosystem in which every component acts on every other component, in which the killing of wildcats may eventually lead to a certain kind of flower dying, because the wildcats usually ate the mice who eat the butterflies who pollinate the flowers, and without the wildcats, the mice have multiplied and eaten too many butterflies. Think about a world like that, and you can not only weave diverse ideas that you may have together—which seems to be a pretty common method of worldbuilding—but let the ones you already possess grow new connections and fill holes. A wildcat naturally has to have something to eat, a flower naturally has to have something to pollinate it, and in creating a religion that naturally has to have an antecedent and a village that naturally has to have a dark secret in its past, you may be able to link them just like the wildcat and the flower.
One nice thing about fictional human society ecosystems is that the author doesn’t have to exercise the same delicacy that is best with natural ecosystems. In fact, story conflict and a great deal of potential comes from upsetting them. And if you’ve grown yours, you’re going to have an excellent idea about what the weak points are, and what ripples might spread and, say, tug a character who would normally have no reason to participate in the action into the action.

4) Remember to keep the world in motion.

That is, things don’t stop happening just because the point-of-view character is currently sleeping, or making a long, hard, boring journey from Point A to Point B, or overhearing the conversation in which the villain explains his clever plan. The world is still spinning. The sun is still rising and setting. People off-stage are still living their lives. And things will occur, or start occurring, that the point-of-view character would give his eyeteeth to help or witness or stop. Keep them in motion. This is a great way to remind your audience that there is a wider, wilder world out there than just the one they’re witnessing right now.
As to what that world is doing? Well, you will know better than one point-of-view character who, third person or first person, is often limited. Perhaps it is now midsummer, and that just happens to be the day that the villain (unknown to the protagonist) has picked to launch his evil plan. Perhaps there have been hard harvests in the neighboring kingdom, and thus famine, and that is going to make for a tidal wave of refugees hurrying into the protagonist’s kingdom. Perhaps there has been a death in the family while the hero was away saving the world, and he comes home to find that changed.
World embraces character embraces setting embraces plot, and you can change all those terms around and still have a sentence that makes sense. If worldbuilding is the most difficult task a fantasy writer faces (and I don’t know if I would say it is; at least as hard, I think, is resisting the temptation towards simplicity and fixing every problem that comes along with magic), then keeping track of time and happenings elsewhere can help you to ease the burden.

5) Update old legends.

A fantasy world can be static, with the same old legends and rituals and traditions and history, with nothing ever changing, until the protagonist comes along. And a bee can sting you, too, but I would bet you prefer that it didn’t.
Another way to world-build, this time focusing on the legends and mythic history that form such a huge part of many fantasy worlds, is to think what effect shifts in language, history, weather, geography, migration, politics, and time will do to legends. Yeah, maybe the legend of how the gods made the world, and of the savior who will come to rescue it, really is the same after six thousand years. But that’s the road back to Staticland, and to making your world harder and harder to distinguish from other worlds out there that may be drawing on the same mythic tradition you’re drawing on.
Work in changes instead. Perhaps the original group who believed in the gods creating the world and the savior coming to save it splits apart because an earthquake alters the terrain, and begins migrating. One group may come into contact with another civilization that believes in just one god, and adopt that belief or mingle their own with it. A second group may retain the legend of the gods but no savior, because people decide they like that story better. Another group may encounter people who speak a different language, start speaking it themselves, and change the names of the gods until they’re unrecognizable without careful detective work. And meanwhile, ages pass and people die and forget things and abandon old traditions in favor of new ones. Soon you have a fantasy world with different, flourishing traditions, not one “truth” carefully preserved down the ages.
Maybe, when the savior comes along, he won’t be what anyone expects. Wouldn’t that be fun?

6) Let your characters do some world-building for you.

Sure, it can be hard, remembering that you have such and such a viewpoint character and therefore have to tailor description, memory, dialogue, plot events, and many other things to what he would notice, remember, say, get involved in, and so on. But I don’t think this should be viewed as a struggle between author and viewpoint character. The character can also cut out a lot of dead air and white noise by lending you his unique viewpoint. He can fill up sections where you would otherwise put yet another palace with marble spires with a building that he would expect to see instead.
Look at your character’s background. It may be utterly necessary to the plot to have him reared in a log cabin, in an environment reminiscent of the American frontier in the eighteenth century. Don’t just use it for plot convenience, though, any more than you use the mentions of other people and places (see point 1) for sheer exotica. Start exploring what it means. What kind of necessity made his ancestors have to go out into the frontier? What kind of home country would produce such settlers? What kind of buildings would that home country have, what accents, what after-dinner customs? And what would your character, venturing into the home country for the first time, take for granted, and what would he stare askance at, and what would he compare to the stories and find wanting?
Layer your character. Take what you’ve already created because you had to, and build on it. Dig deep. Chase the fleeting observations he makes, the noises of disgust or interest, and ask where they come from. This can be as profitably applied to worldbuilding as characterization—perhaps more so, because of the disconnection that exists between some protagonists and their fantasy backgrounds, and the tendency of many authors to turn to marble-spired palaces and their kin without thinking.

7) Sacrifice cliche to sense.

You want the final confrontation to take place on top of a tower hundreds of feet high. You’ve been planning it since the beginning. The hero and villain will argue, then duel while the hero’s friend the wind mage hides below. The wind mage will hurl a particularly vicious gust at just the right moment, and knock the villain screaming to his death.
Then you start worldbuilding, and realize the problem. The setting you want for the final confrontation is a series of plains, with no trees or high hills to stop or slow the wind down. So far, so good. Unfortunately, given the technology and magic that the plains-dwellers have, there is no way that they could build a tower as high as you want; the first really good storm would turn it to matchsticks.
Now, you could stick with the tower existing in defiance of all sense, or even being an ancient and mysterious magical artifact that the Builders™ left behind. Or you could alter the landscape or the site of the final confrontation so that it matches better with the world that you have. Make it practical, make it fitting, make it yours.
A lot of fantasy’s bad reputation comes from authors just not thinking before adopting whatever pops into their heads. But it does no good to think if you think and then decide that you’re going to persist with the cliché anyway. Build a world that is not the same as a dozen potted epic fantasies, whether that’s through justifying the cliché so it’s not cliché anymore or altering it. Make it yours. And pierce every seemingly simple and shallow puddle in that world to reveal the depth and other layers lying underneath.
A lot of this is tied up with plot, character, and setting, I notice. Well, done properly, a book will be that tapestry or brick wall or gourmet meal, not a bunch of loose ends or tumbled stones or sardines next to ice cream. And Arabian-like fantasies that move, touch on aspects of world-building, race away and circle back, light and swift, have always been my favorite kind.