When starting a new game, how often to play is an important consideration. Personally, I'm a fan of playing weekly as developments are kept fresh in people's minds. A week is long enough for a player (or players) to talk and plan, but not long enough for an enthusiastic to grow too frustrated because they can't move things forward. Weekly games help players remember details such as the names of characters and locations and if the occasional session is missed the schedule is not pushed back by months. Playing weekly allows a "continuous TV serial" style of game. Sessions don't all have to be crammed with plot because there is are more coming along soon.
A slower schedule would have different strengths and weaknesses. A monthly game may breed more excitement for an individual session and can avoid cancellations as people prioritise playing higher. Playing monthly encourages more of a "cinema" style game with infrequent instalments, each of more importance to the whole. Each session needs to advance the plot while also giving all the players the space to play their characters. On the other hand, each session will likely be longer than a weekly game and has more time for the GM to prepare. Rather than being a continuous plot, a game with long gaps between sessions is likely to be written episodically. Everyone forgets details, so each session will need to be relatively self-contained so there is less to forget.
There isn't a "right" way of running a game and the decision will likely be taken on an entirely practical level based on everyone's availability. The danger comes when not taking the frequency into account when writing the game, or when circumstances force a change.
In between the excitement and adventure of beautifully crafted gaming sessions exists the tedium of real life and that has a habit of intruding on and disrupting the game. My group are all post-university (and hence in-work) age and lead busy lives. Consequently, our weekly game skips roughly one in five sessions. This isn't a problem when it is one a month, but it is all too common for us to go through a period of one session, two weeks off, another session, three weeks off. This is a sad, inevitable fact of getting older but results in the weekly game - written to be continuous - at times running monthly. The ongoing challenge is how to bridge those periods without re-writing into an episodic game (not that there is anything inherently wrong with an episodic game, but that would change the character of what I'm writing at the moment).
Big gaps mean forgotten details. Obvious things first - a recap helps everyone get back on track. Starting each session with a "previously on" sets the scene, highlights important details and neatly doubles as a signpost for the beginning of the session. Keeping good notes during a session is essential for this - there will always be times when you come to write the recap and you have no idea what a particular player did last time, regardless of how important it was.
Another good way to help people's memories along is with an NPC cheatsheet. For each plot arc I write my own sheet to help me remember names, motivations, personalities, physical traits and political allegiances. Next time I do this I'm going to produce a modified one for each of the players. This will help not only with remembering everyone - and because we're playing Legend of the Five Rings and we are not Japanese speakers, many of the names are complicated - but also be a starting point for their political manoeuvring as it will aid them visualise the courtly battlefield (as noted in Feedback a few weeks ago). A simple and obvious idea and one that I'm kicking myself for not doing earlier.
It's important to remember that characters have memories as well as players. A detail from two months ago in real time might be something experienced an hour ago for the character. Various mechanics exist for character memory (most of them similar to character deduction, or GM hints) however before reaching for the dice it is worth asking "what happens if they fail this?" If the objective is to bridge a perfectly reasonable gap in player memory a poor dice roll is going to lead to an unreasonable blocker to the story and frustration so maybe skip over the check entirely.
I don't want to completely rewrite the plot to something more episodic, but I am looking at ways to restructure what is going on to help keep things moving. A sequence played weekly can afford to be slow and have some false starts, but if those same sessions are less frequent it is very easy for that sequence to feel like it is taking forever - months in real time even though it's only two sessions of game time. I want to have plots which can be fast-forwarded when needed and have escape routes which I can use to temporarily "episodify" (not a word) the game when real life gets in the way.
Finally, I'm re-evaluating the experiences of the sessions themselves. Memorable NPCs stick in the memory (obviously) so what makes a memorable NPC? An outlandish character can certainly be memorable, but doesn't fit the character of the game so I'm working on enhancing the descriptions and interesting traits. The same is true for the locations. These are things that need to improve anyway and with a focus to that improvement I'm looking forward to seeing what happens.