I'm running a game that does entire campaigns in 5 sessions:  Prologue then 4 Acts.  The theme is Gothic Horror and the players/ characters rotate every campaign (yay for being one DM with dozen+ players).

I'm looking for some thoughts on the following story arc for our next "book".

Prologue - TBD - Get the players into the mists and into the small county of East Lake

Act 1:  Players are tasked to investigate a banshee that preys during the full moon and drains it's victims dry.  They are not drained but escape the encounter without beating it.  Question for end of session:  What do we need to take with us to give ourselves a fighting chance.

Act 2:  Return to the forest, better armed and better prepared.  They defeat the banshee, destroy it and return victorious to town.  But, oh no!  The Warden's Daughter has been attacked in the night as well!  Her life force is drained and she is but a husk, slain by another banshee no doubt.  Question for the end of session:  How were there two banshees?  How did it get into the manor undetected?

Act 3:  Investigation of the attack.  Through cunning, stealth, guile or just plain brow beating they discover that the daughters are themselves the banshees and the warden has been dead for many years.  Each month one of them is split from her body to prowl the country side for a victim to drain so her own youth can be maintained.  The eldest knows the ritual, the younger two are simply victims of her father and sister's machinations.  The father has been gone but the eldest has kept up the rouse by insisting he is "indisposed" while she sees to her sister's care and "feeding".  Session ends with the question:  What do we do now?

Act 4: Resolution as the party sees fit.

Comments?

Prepare for the possibility that your party won’t run from the first banshee. A lot of D&D players, especially newer ones, come in with the video-game idea that every enemy is a fair fight. Be sure to remind your players beforehand that, since this is a horror game, many enemies are better to flee from if the fight’s going ill. Maybe they realize their weapons aren’t doing any damage, and need to get the warriors’ weapons silvered? That’s an easy way to communicate that a fight’s beyond the group’s current ability, lest they commit suicide by bravery.

Some more passive players also don’t know exactly when or how to start an open-ended act, so it wouldn’t hurt to end each of your acts with the same questions you have here! Otherwise, many players won’t even think to start an investigation in act 3, for example.

Yeah, almost I stopped reading after act 1. Never assume you know how a fight is going to end. Naivara was right that players will rarely, if ever run from a fight. Maybe your table is different, only you know that, but its usually a big stretch to build that in.

Moreover, if you do already know who's going to win, then what's the point of the players even showing up? This is you dictating a story to them, not crafting it with them. How do you know they won't be drained? Are you just planning to fudge die rolls? (and they're saves, which the players will be making, so they get really hard to fudge anyway) And what does drained mean? There's no level draining mechanic in this edition. I don't think at all (could be wrong there), but certainly not by a banshee. 

Then as I read on, the same issues crop up. Maybe they don't want to go back to town after they beat the banshee. Maybe they don't care that the warden's daughter has been drained (again, not sure what that means) and decide one banshee was enough and they're leaving town. You've put them on a very tight leash, you're assuming what they're going to do next and who they're going to talk to. You should just be providing them with problems, leave it to them to figure out the solutions.

If you want the PCs to fail at their first encounter with a boss, the best way of doing that is to give the boss victory conditions that don't require the PCs surrender, such as 'boss arrives, kills some NPC/steals some McGuffin, leaves, possibly beating the stuffing out of the PCs in the process'. A separate problem, specific to your choice of monsters, is that a Banshee encounter has very high randomness.

Without sounding unnecessarily harsh - it doesn’t sound like a full campaign arc - even a 5 session one.

it reads like a single episode of an anime and as others have said relies heavily on your players doing exactly what you want rather than what they want.

Try and build a locale instead, somewhere for it to take place, you have your BBEG, so build into it some red herrings and false leads. Create people for them to interact with.

think about the characters your players will be playing, put in activities, potential character moments or plot points for the characters. Create situations that challenge their values or expose flaws.

you are trying to build a ghost train when you should be build a fun fair

Everyone is correct, but my players would need the structure that has been shown in the OP post. 

Seems to me though that the OP is trying to tell a story with some dice rolls, not have a campaign as such. That is fine, but if you are doing that you need to make sure the story (and the meta story) is worth telling and worth hearing. 

Good luck. 

I’m going to chime in a little by saying that as long as your table is having fun, you’re doing something right. Running short term adventures like you describe is fantastic for getting into the habit of prep work. Also, you can run encounters from multiple points of view when you’re running groups back to back like that. 

my question would be, is there a way you can weave these all into an over arcing plot that engages your entire group for a big bad final battle session? Let them know it’ll be primarily combat and tactical combat at that, but the interplay between the more role play focused players can be a joy to behold.

I have ran 12 person groups before and while it seems daunting, if you prep for combat it can be satisfyingly done. 

I really wanted to thank everyone for the general feedback.

I do think I need to massage the first fight with the Banshee and work up some kind of clear "you are not being effective with A, you will need to pursue B if you want to defeat this" that can be communicated throughout the fight so that the players have a clear reason to back off their fight and "go learn more".  I wanted to use the tool of "you are not yet ready to defeat this enemy" you see in a lot of literature to help set up tension for later.  You are right though, that players tend NOT to want to end a fight until everything is dead, even if that everything is themselves.  :D

I'm also not sure anyone at our virtual table would NOT want to see the "right thing" done.  My goal in the story craft was to build to a larger moral question at the end of Act 3.  In Book 1, it is asking "do we side with revenge over justice?" or "what is justice?"  Here I wanted it to be "do we destroy something innocent because it's existence causes harm?"  Or something like that.  But along the way I can't imagine the players NOT wanting to help the daughters of the Warden to rid their community of it's monthly (ish) plague.  Everyone at the virtual table knows that this is only a 5 session game and not a full sandbox campaign.  That said, I also don't think anyone would say 'well, hey, clearly there's an adventure here, but we'd rather to explore the woods".  Part of my thinking is that that could also be that as we're streaming the games live, players are more likely to opt into being "heroes" and less likely to opt out of doing the heroic thing like saving the town ect.

There is a note to be taken, though, in the idea of how I can give them more agency on what happens in Act 2 (the figure out the bane of the banshee and going to kill it once and for all).  Maybe the solution is not in town but some other relic they can get.  I do have to be careful as I only have 2-3 clock hours to a) figure out the bane, get the bane, go back and kill the banshee, get back to town to report success and go "wait.. wha?!?".  I could move the learning of the Banshee Bane into Act 1 and make Act 2 the quest to get it and then to use it and that spares me some of the time issues.

I do appreciate the feed back and I realize that a clear story arc isn't everyone's game.  What I'm trying to craft is a blend of solid story with a clear beginning, middle, and end, with enough decision points along the way that the entire thing isn't completely on rails.  They players need to hit the story beats but I do want to give them some choices as to how they're approach those beats.

Quote from DM_from_1975 >>

I’m going to chime in a little by saying that as long as your table is having fun, you’re doing something right. Running short term adventures like you describe is fantastic for getting into the habit of prep work. Also, you can run encounters from multiple points of view when you’re running groups back to back like that. 

my question would be, is there a way you can weave these all into an over arcing plot that engages your entire group for a big bad final battle session? Let them know it’ll be primarily combat and tactical combat at that, but the interplay between the more role play focused players can be a joy to behold.

I have ran 12 person groups before and while it seems daunting, if you prep for combat it can be satisfyingly done. 

I would consider that except for one big thing which is that this is over a stream on Twitch.  Getting 12 people in a zoom call alone is a challenge but to keep them all feeling engaged, live on camera, is a whole different beast.  

That said, I am SORTA working towards a big bad evil guy in the form that all of these adventures are set in, roughly, Ravenloft.  It's possible that over the year I'll work up to running Curse of Strahd as an adventure for a group of players that can then take the half dozen or so previous "books" in the world and build towards a big final showdown with the arch nemsis of the world himself.  Of course once the Mists are lifted and the darkness ended, it also, sorta, destroys the campaign world as it is and puts me into exploring a different campaign setting which... I'm not ready to do yet.

Along these lines,

No battle plan survives the first shot.

Act 1. I would say make the encounter with the banshee challenging but winnable with good tactics or preparation. If they win, you can skip Act 2.

Act 2. I wouldn't reveal right away that there was a second banshee. Maybe give a few light clues. Maybe the players guess it now, or maybe they'll be a few steps into the investigation before they figure it out.

Act 3. Nice twist.

Quote from NaivaraArnuanna >>

Prepare for the possibility that your party won’t run from the first banshee. A lot of D&D players, especially newer ones, come in with the video-game idea that every enemy is a fair fight. Be sure to remind your players beforehand that, since this is a horror game, many enemies are better to flee from if the fight’s going ill. Maybe they realize their weapons aren’t doing any damage, and need to get the warriors’ weapons silvered? That’s an easy way to communicate that a fight’s beyond the group’s current ability, lest they commit suicide by bravery.

Some more passive players also don’t know exactly when or how to start an open-ended act, so it wouldn’t hurt to end each of your acts with the same questions you have here! Otherwise, many players won’t even think to start an investigation in act 3, for example.

You can run an unwinnable fight even if your players won't run.

You can give your monster a reason to flee. Maybe they are turned away by the rising of the sun.

You can have a powerful ally ride to the rescue when your party is about to TPK.

Maybe your villain wants them alive. While they are imprisoned, the players will have a non-combat encounter with the chance to escape.