A series is the most reliable structure in genre fiction, because a reader who finishes book one already knows they like book two. It is also the easiest thing to break, and almost all the breakage happens in book one, before anyone knew there would be a series.
Decide the shape early
Two broad models, and they behave very differently.
Episodic: each book is a complete case; the character arc moves slowly underneath. A reader can start anywhere. Easier to sell, easier to sustain, more forgiving of a gap between books.
Serialised: a larger arc runs across the books, with each resolving part of it. More gripping, and considerably less forgiving — a reader who starts at book three is lost, and if the series stops early it stops unresolved.
Most successful mystery series are episodic with a serialised emotional thread: the case closes each time, the relationships do not.
Build a world that generates plots
The single best predictor of whether a series can run is whether its setting produces new stories on its own. A small town does, because everyone is connected to everyone and the past is never finished. An occupation that puts your protagonist inside strangers' lives does, because each client is a new closed world.
Before you commit, ask: can I name eight more books' worth of story in this setting? If you cannot get past three, you have a trilogy, and there is nothing wrong with that — but plan it as one.
What to leave open in book one
New writers close too much. Book one should resolve its case completely and leave three things unfinished:
- A relationship in motion. Not a cliffhanger — a direction. Two people who are somewhere different at the end than the beginning, and not yet settled.
- A personal wound not fully healed. Whatever your protagonist is carrying should be lighter, not gone.
- A corner of the world unexplored. A character mentioned but not met, a place referred to and never entered, a piece of the town's history nobody will explain.
Keep a bible from book one
Start it the day you start writing, not the day you sign a three-book deal. Names, ages, eye colour, cars, the layout of the protagonist's house, who is related to whom, what month each book takes place in, and every detail of backstory you have committed to in print.
This is not administrative fussiness. Readers of series fiction remember everything, and they will write to you about the dog that changed breed. More importantly, a bible lets you plant something in book one and pay it off in book four with confidence that you have not contradicted yourself in between.
Time is the trap
Decide early how time passes. If each book covers a month and you publish two a year, your protagonist ages a fraction of the rate your readers do — fine, and most series quietly do this. What you cannot do is be inconsistent: a character who was "nearly forty" in book one and has a fortieth birthday in book five, after three years of story time, is a problem readers will notice.
The related trap is the escalating body count in one small place. Address it with humour, geography, or variety of crime, but address it.
Let the protagonist actually change
The reason series go stale is that the author, sensibly wanting to preserve what readers liked, freezes the character. By book four the protagonist is doing an impression of themselves.
Change them. Let the thing they were afraid of in book one be something they are competent at by book four — and give them a new fear. Let a relationship progress rather than circling. Let a supporting character leave. The reader's loyalty is to the character being alive, not to the character being identical.
Write book two before book one sells
If a publisher takes book one, they will want book two quickly, and the gap between contract and delivery is shorter than you think. If nobody takes book one, you have learned enormously more about the world by writing two, and the improved version of book one is easier to write than the first.
Either way you are better off. This is the closest thing to universally good advice in this article.