@BlazikenLover
Another thing I forgot is to try to instill your dungeon with a sense of urgency. Have a steadily ticking clock with escalating bad things.
Maybe its a lava dungeon and the heat is slowly killing the players. Maybe its a sunken ruin that is flooding. Have the big-bad completing a ritual. You probably don't want them constantly under the gun, but make sure they understand that sitting around arguing is wasting valuable time.
One of my favorite things to do is have in-game things, like heat exhaustion from the lava heat, or how much longer till the BBG completes summoning the demons, tick up not based on game time but real world time. Every hour of real-life time, they need to make an a progressively more difficult endurance check; every two hours the BBG gets a couple more HP. stuff like that.
Honestly, the best tricks to "speed" up a dungeon is to limit the scale.
One of my biggest vices when it comes to DMing is I absolutely love designing dungeons and experimenting with them. I love especially the ones that are designed as ancient ruins and lost cities; and can go crazy with size due to that. I also learned very early that any dungeon larger than say a 30 x 30 map will take at least two sessions to clean out, and any one that has more than one level will take way longer.
So yeah, limit how big your dungeons are if you don't want them to be multisession explorations.
The next thing you do is you actually have to limit encounters, since encounters always suck up at least a half hour even with experienced players depending on the size. I often find that traps and environmental hazards tend to have the same effect but flow better. Throw in monsters when they make sense for the location, or if they are dawdling and arguing too much in an area.
Last thing I'd argue is if you use puzzles; as long as the party shows a decent logic to their solution, then accept it. If you're going to be persnickety over how they solve it, they're going to be stuck there for a while.
But yeah, unless the dungeon's small, it's going to be at least a couple sessions dude.
When I look at
most puzzles I make sure there are three solutions - The right way, an alternate way, and a brute force way.
Unimportant puzzles, like opening a small treasure box I probably won't go this far, but anything required for progression; and if its a hardcore meatgrinder, approved solution (or related) or get fucked.
The right way is the the officially approved way. The alternate way is another way of getting through the puzzle that more-or-less meets the conditions. And finally there should be a way for the party to throw in the towel because they just aren't getting it.
For the right way, make sure the party has sufficient information about how to play the game. You don't have to literally spell it out (but you sort of do) but if the statues need to rotate, make sure there is something like scratches around their bases that show that they are supposed to rotate. If they need to pull levers, make sure there is something they've encountered that lets them tie lever colors = squence. IF there is danger, put a corpse somewhere.
The alternate way is another way the players can solve or get around the puzzle. This can be as simple as "They take their axe to both magic doors and ask them 'Did that hurt'?" or maybe earlier they found a key in the Guard Captain's secret stash that unlocks the door. Or maybe one of the doors is able to be flattered into giving up the solution. If the alternate solution requires equipment, put that equipment near by.
Thinking up an alternate solution/work around to the puzzle doesn't just give the players another way to get through, but it gets you thinking about other things the party might try and lets you be ready to either counter munchkins trying to exploit the game rules, or be ready if the players do try something similar.
Finally, have a cost ready for the players just being completely stumped.
And, again this is really only for puzzles that are needed to progress through the dungeon. Don't feel a need to let them eventually guess the gold vault combination.
But Figure out a time/HP/Resource cost for just "We're stuck and want to move on" - they take all the consequences for wrong answers, maybe get the unoptimal path, but they can move forward to something less frustrating.
My group has found a good way to make a dungeon feel densely populated without having a billion encounters: if the "encounter" would be of trivial difficulty, don't even tell the players to roll for initiative. Just roll one attack against each and declare they quickly dispatched that group of monsters.
The undead-infested catacomb goes ever deeper. The stairs deposit you it a large hall with a handful of decrepit skeletons milling about in a mockery of their duties in life. Some sweep, some tend to the magical torches affixed to the walls, others still move ruined holy symbols from one side to the other. As your presence is noticed, their empty eye sockets turn towards you in unison and they stagger and clatter towards you. Disorganized and weak, the skeletons are no match for you. (Roll attacks.) Ragnar the Barbarian takes a small scratch to the arm while sending bones flying across the hall (that's 3 damage to you, Jeff), and Mara the Priestess has a quick scare as a previously unseen skeleton manages to grab onto her robes and claw at her before her holy mace reduces its skull to dust (that's 2 damage, Dave).
With the room secured, you find that it has exits on both sides and a heavy-lidded sarcophagus in an alcove in the far wall, behind an altar. What do you do?
Obviously, this works better when the party is already powerful enough that those threats are suitably small. It's also a trick that then demands the actual encounters to be more interesting than just "X monsters (type A), Y monsters (type B)" and no additional mechanics, because it implies trash-clearing to be not worth the
players' time.
That's a good idea, I'll have to steal that. I've been doing something similar with converting encounters into traps/hazards and just having the trap roll an attack, but I like the idea of just a one-roll random combat.