There is an occasionally had conversation at the club between GMs about various approaches to writing an adventure and devising an overall plot. We were chatting about this again recently and sharing our various perspectives and experiences on both sides of the GM screen.
For those new to running a game there is the default road of starting with a published module which is definitely the way to go when facing a new career and not yet having built up a game management instinct. This ticks a lot of boxes straight away regarding npc detailing and overall environment but the GM has to be very clear on the details and adhere to the script tightly as the adventure will have plot threads dependent on the cause and effect of actions within the module context. This all entails reading over a source book several times. The disadvantage, particularly so for a club, is managing the time span of a fixed adventure - the club rotates games on quite a fierce basis given that there are only a couple of hours role playing in a week, games really should end every six months. Typically campaigns can be sliced into chapters somewhat like a Game of Thrones series which is a very good way to keep players hanging in suspense and not letting fatigue set in. Having said this, I have seen several games advertised in Meetups using the exact same campaign which can be a bit frustrating if you've already done it.
So there also has to be an option of bringing a game to a climax at short notice and this can be easy or hard depending on the plot. In this respect the other way to go is to heavily modify an existing campaign of create one from scratch. For my part I prefer to build a game from the ground up and generate several sub plots with the option that not all of them have to be completed to result in an overall success - this works quite well for me as it never ceases to amaze me how long a party can take to complete one small task or how quickly they can drive though another. In addition I can manage off roading easily when I have built something myself but it can take a while for the players to build up enough knowledge and stake in a game not to feel they are just part of some GM centric ego project.
In the final analysis, its often the case that if a player is getting unhappy then you can just pay them off.