How to Write a Web Serial

With Worth the Candle concluded at a hefty 1.6 million words over four years, I can finally give some authoritative tips! I can’t promise that these tips will work for everyone, because writing follows some very individual processes, but I think they generalize well. If you want a more general post on the more business/audience/process side of things, I would recommend my Serial Writing FAQ.

Work in Chunks

One of the things that makes web serials more easy to manage is to break things down into their component pieces. The scene is the fundamental unit of storytelling, and scenes make up chapters, which make up plot arcs, which in turn make up books, and if you stack up enough books, you’ve got a whole giant serial. Anything between the ‘chapter’ and ‘whole giant serial’ level is technically optional, but I think playing at several levels is what makes the most sense, both as an author trying to write the thing, and for readers. The classic narrative is peace->conflict->tension->resolution, and one of the primary things you do when making these ‘chunks’ is to ensure that you’re nesting narratives. An example might be starting with a ‘hometown’ arc, which has its own peace->conflict->tension->resolution, but also, taken by itself, serves as the ‘peace’ part of a larger overarching arc.

If you do this kind of division or nesting, you can think of different parts of the story as black boxes whose inputs and outputs are known, but whose internals are not. It allows mindspace to be dedicated to a more granular level of the story, whatever you’re working on at the moment.

Generally speaking, I would set aside a few hours for outlining when I was starting in on a new chunk, though my approach to outlining usually meant being away from the computer and having my body do something while my mind was occupied with the book, which meant that it didn’t much cut into normal fingers-on-keys writing time. In this way, outlining got divided up across the whole time I was writing, with the major plot beats pinned at the start, especially those at the end, but the rest having attention on them only when I was closer to actually doing them.

Know Where You’re Going

When you start a large project, you do it without knowing every single beat that’s going to happen over the course of the next however many years and however many hundreds of thousands of words. It’s definitely possible to outline to that level, but some level of discovery is definitely preferable. Instead, I think it’s best to map out what you think the ending is going to look like, and from time to time, as you write, try to see whether you’re still aiming in the right direction. If you’re doing the division and nesting as above, you should be doing this at every level you’re at: if you’re writing the first part of the hometown arc, you should know both where the story is likely to end, and where the hometown arc is supposed to end. I think this gives much better results than not planning for the future, while still allowing a lot of freedom in terms of what you’re writing.

Even if you’re doing complete blackboxes for later arcs or chapters, knowing what they’ll be in general terms allows you to do some foreshadowing, and allows the later stuff to creep its way into the earlier stuff so that you’ll have prepared the audience. This kind of foreshadowing should be casual, usually not infodumped, just there in the background, because you know that you’ll probably get to it later.

Know Your Characters

You don’t want to be deciding how many siblings a character has fifty chapters in. You don’t want to be figuring out what college they went to when they run into a stranger. You want to know them, inside and out. Now, this mostly applies to main characters, since they’re the ones that are on-screen all the time, and the ones most likely to be hurt if there needs to be a retcon. You should know your characters and their histories because characters are at the core of the work, and their histories are one of the primary things that inform them. Even if you don’t end up using a lot of it, it helps to make them deeper and richer. When writing Worth the Candle, this was very easy for Juniper, because he borrows quite a bit from my own life, but I tried to do this with the other main cast as well. Being able to draw on things that were written down in a long life outline means that you don’t have to worry that you’re contradicting yourself, and it makes pulling formative or relevant things out of their backstory when appropriate feel organic.

The same goes for your world, but to a somewhat lesser extent. Know the customs, know the laws, know the religions, and know how your characters relate to all of this. Make a map, even if it sucks, and at least give each of the major kingdoms a name and a single sentence description. Figure out what the magic is and how it works, and then what people would know about it. Know your tech level (which is easier if you pin it to a historical date, but if you’re not going to do that, then figure out which areas of technology are at which levels). All of this is by way of reducing overhead when actually writing. A lot of people succumb to ‘worldbuilder’s disease’, endlessly working on their world instead of writing, but what I’m advocating is an approach that gets you as much bang for your buck as possible. Figure out the demographics of the hometown, and the province, and the nation, but don’t do that for neighboring towns, or provinces, or nations. Think about what’s most relevant to the characters, and what’s going to be relevant downstream, and figure that out, nothing else (unless you want to).

Themes Create Themselves

… if characters are consistent. This is a bit of advice stolen from Wildbow, though I think I take a bit weaker of a stance, which is that if you have a character with some particular personality or history, you will naturally end up writing about that personality or history a lot if you keep increasing the word count. Even if you made no solid attempt at themes, you’d have them anyway, because you’d just keep writing about the same thing in different ways, and the characters would keep inserting ‘their’ theme into the plot.

With that said, I think there’s a lot that can come from knowing what themes those characters will create, and allowing there to be intentional mirrors or companions to it. If you know that the main character is recovering from the brutal loss of his son, then you should look at the other characters and see how they’ll relate to that, and if possible, make characters that can hold a mirror up to the main character. I don’t think you necessarily need to be incredibly blatant about his, but if you know what theme your consistent character is going to end up revolving around, you should figure out the ways in which other characters will be drawn into the magnetic attraction of the theme, or be repelled by it. Characters develop their own theme, but it’s better if there’s some thematic connection between characters, because that makes the theme stronger and the work more focused.

Similarly, themes should ideally be part of plot. If I’m writing a story about a character with a dead son, then my theme is loss, or the pointless cruelty of the world, or a father’s love, or something like that. Having that as a running theme when the plot is about a heist from a casino is dissonant (depending on execution, I guess, because of course I’m not thinking about how to tie those two things together). Character themes should work in harmony with plot themes. Got a dead son? The plot should be about death, like stopping a killer, or curing a disease, or someone else curing a disease and science going too far. Or maybe the plot is about a surrogate son, or protecting the ones we love, or the dark sides of fatherhood.

Point being, themes grow organically, but we can also apply pesticides and fertilizers.

Give Yourself Room to Grow

Try not to pin absolutely everything down. This is the opposite side of knowing your characters: give them a little breathing room so that you have some ability to wiggle. Sometimes, in the course of writing, you’ll want some ability to insert a friend from the academy, or an uncle with a particular set of skills, or a rare magical creature. This keeps you from getting penned in, and helps to keep things fresh if you’re finding that writing within the same stagnant set of facts is getting old.

It’s for this same reason that I suggest taking a focused approach to worldbuilding that zeros in on the things that are going to be directly relevant to what you’re working on. You probably want to know the names of the neighboring kingdoms, and maybe have a single sentence description of what makes them unique, but if the characters aren’t going to go there until book three, you probably don’t want to pin down a bunch of worldbuilding until you’re closer to actually writing.

One of my big takeaways from writing Worth the Candle was this: you are not going to have all your great ideas in the weeks or months that you do your initial planning. You’re going to have great ideas along the way, and needs that must be met, and you’ll want to be able to have some ‘give’ to the world and characters so that you’re not forced to file every new good idea away for some future project.

Sprawl Is Not the Enemy

… bloat and filler are. Web serials are known for their sprawl, for touching on lots of things, going to lots of places, and having a lot of characters. They’re known for their high word count. I don’t think that any of this is a bad thing. What’s bad is having words that contribute nothing to the story, reveal nothing about the characters, or which aren’t Cool. Words should fulfill a purpose of some kind, and if they serve no purpose, or not enough of one, they should be cut. The question I always ask myself is ‘does this justify itself’, and if the answer is no, I will try to do the painful work of trimming it. Some things justify themselves through pacing, the need for a reader to have a break from the tension and conflict, and some justify themselves by being connective tissue between things, but I think overall, you should still be careful to make sure bits like that can justify themselves on a number of different fronts. To me, the ideal scene is one that advances the plot, reveals something about the characters, provides exposition, interrogates an idea or theme, and is Cool. Most scenes cannot meet that lofty ideal.

Give Yourself Outs

Similarly to giving yourself room to grow, you should also give yourself room to bail out on things that aren’t working for whatever reason. In the extreme case, this means dumping a character that’s not working for the plot, but there are a lot of lesser cases where you might decide that something which seemed like a good idea two years ago is actually going to be hard to do well, or perhaps has too much overlap with what you’ve already written, or maybe … I dunno, you just don’t want to anymore. When you start your story, you only have an outline for what it’s going to be, and you need to give yourself the ability to divert from that outline. This might be easy or difficult depending on the story you’re trying to tell, but knowing that you can go a different direction from the one you planned means that you have a better ability to handle the good new ideas as they come up, or the reevaluation of old ideas, or anything else.

A part of this is also just giving yourself permission to change things, which I know for some people can be a bit of a psychological block. The outline is there for you to have good foreshadowing and development, but it’s not supposed to be a crutch that you’re leaning on. (Though do make sure that if you bail on a plan, you make sure that you’re also not making downstream problems, not unless you can fix them.)

Write With Confidence

This is a hard one, because along the way, people will be telling you that you suck, or that you’re doing it wrong, and sometimes the person telling you that will be yourself. Writing with confidence means writing as though you know exactly what you’re doing and are simply in the middle of part twenty-seven of a ninety part plan.

There are a few reasons to do this. I think part of this is just a good outward facing attitude: I read writers who apologize in advance for something that’s rushed or bad, and 90% of the time it’s just insecurity that I wouldn’t have noticed without the apology. Readers don’t need to know what you’re making up as you go along, what you think is executed poorly, where you felt constrained by continuity, or what’s just ugly connective tissue. This is, I admit, largely a matter of opinion, and more about the public relations and community management aspects of writing, but it’s something I feel fairly strongly about. It’s fine to discuss how you feel the writing is going, but I think it should be done in the confident way of a master craftsman who sees that the table he’s building will need to be sanded down a bit before the varnish can go on, rather than someone who is building a table for the first time and totally fucked it up.

Whatever you do, definitely don’t let any of that hesitation or anxiety about the writing creep into the text itself. There are various ways that I’ve seen this happen, some of them more overt than others, but it’s always a turn-off.

Be Consistent

I don’t mean that you need a consistent daily output of words, though you should. I don’t mean that you should have a consistent release schedule, though you should do that too. What I mean is that you should know what kind of writing you’re putting out, and stick with the same feeling as much as possible. If you need to change genres or themes or contexts, telegraph it first, then ease into it gently rather than swerving suddenly. Sometimes you’ll find that you’ve mined out everything you wanted to mine out, or sometimes you’ll need to move in another direction to keep things fresh, but to the extent you have readers, those readers are there because of what you were already putting out, and new readers will need to make it through the early stuff anyhow. Know what you’re ‘selling’ and do your best to stick with your brand.

And if you find yourself selling something that you weren’t selling when you started, that’s fine, just make sure that when people come back to your web serial week after week, they more or less know the kind of thing they’ll get. I think this kind of consistency is crucial for audience growth, even more than consistency of word output. That said, even within a given work, there’s a lot of room to play around, so long as you’re not fundamentally going in a different direction.

Conclusion

I think there’s a lot to like about writing web serials, but also a lot of ways that I’ve seen it go wrong. The worst part about writing something very long is that you can feel stuck in what you’re writing, or like you’ve painted yourself into a corner, and a lot of this advice is given with the hope that it helps you avoid that. Some of it is contradictory, I know, but that’s because there’s always a balance to be struck along the way.

Addendum: Examples from Worth the Candle

This is a set of quick examples of each of these principles from writing Worth the Candle, spoilers obviously follow, and this probably won’t make sense if you haven’t read Worth the Candle.

• Work in Chunks: This was the primary way that Worth the Candle was written, typically with arcs set aside in their own chunks and dealt with in turn. Later in the serial’s lifecycle, I batched chapters up for release so that they would contain some kind of mini-arc that was a coherent part of a larger arc. Early on, I tried to think of each “book” as containing two large arcs with similar thematic overlap, but this kind of fell to the wayside.
• Know Where You’re Going: I more or less knew what the ending to Worth the Candle would be from the first day of writing, and as a consequence, the first mention of Fel Seed is in chapter three, and the first mention of Arthur is in chapter four. I really think this helps the work feel like it has a better narrative focus than it perhaps actually does, and there’s a lot of stuff that gets to feel like foreshadowing because I had it in mind from the start, even if I didn’t necessarily know all the exact details. It also allows things to build up a kind of mythical quality to them, which has pros and cons: the big pro is that everyone is hyped for it, but the big con is that it can be hard to live up to the hype.
• Know Your Characters: Juniper was easy, since he had my own life story with some alterations, but I tried to be fairly through with the others. Before Amaryllis was introduced I had worked out the keystone moments of her life and some of her big influences, things like being born to a much older father, the deaths of her parents, being put under the wing of a matriarch, being sent off to school, and then having a brief time in court politics before a dramatic fall from grace. Knowing a quick life story like that makes an enormous difference and helps to avoid some Early Installment Weirdness. The same goes for the rest of the main cast, and for the world of Aerb itself, whose major magics get mentioned early and often, then explained later in a way that gives the whole thing a lot more “free” cohesion and depth.
• Themes Create Themselves: I think there’s such a thing as narrative pareidolia, finding thematic or narrative meaning in things which are random or abstract, or simply not planned. Most of the major themes of Worth the Candle are planned, especially the stuff relating to grief and depression, but along the way I got to look at other characters and see how they dealt with that stuff, because obviously if one character is talking about it an examining it, the others will as well. I think my favorite example of themes creating themselves through consistent characterization was when Grak had his soul manipulated and liked it, because it gave him meaning and purpose he was otherwise lacking, offering him a way of dealing with the intolerable. This allows something that was just part of a ploy by an enemy have a richer connection to the theme of suicide. That wasn’t something that I planned, but it came about by looking at Grak’s view of the world. Characters, just by being themselves will naturally enrich the text like that.
• Give Yourself Room to Grow: Aerb is probably a bad test case, because it’s a huge world with almost a hundred different magic systems, but there are a few places where I deliberately left myself options. For one, I never named all of the exclusion zones, which would have allowed me to sneakily add in a new one if the idea was good enough. For two, I never named all of the mortal species, which meant that I could have added another, though they would have had to be relatively low-impact. Third, there being a whole lot of entads meant that I could introduce new and unique things for a fight if I was feeling that things were getting too samey for me. Overall, it made writing a lot more fun, though ideally the reader didn’t have any idea what was planned from the start and what wasn’t. The one time I can recall getting accused of adding something in post facto was with dragons, which I’d thought about in some detail from the start of the first book (because the major inspiration is Dungeons & Dragons) but didn’t introduce until fairly late. The lesson there probably being that if you have something high-impact, you should drop hints about it early, even if you’re going to hold off on the extended infodump.
• Sprawl Is Not the Enemy: Generally speaking, there’s not much that I would cut from Worth the Candle, and I really do think that the sprawl is to its benefit. You get to see so many things from so many angles and you get so much of each character that it can’t help but feel richer and deeper. By the time you get to the end, you’ve spent over a million words in Juniper’s head, so it’s natural that he feels more like a complex person than if it had only been for a single novel. Worth the Candle definitely embraced the sprawl, and there were a few times that I decided to go ahead with things that I could definitely have left out, just because taking a wander is one of the benefits of the serial medium. Both Blue-in-the-Bottle and Doris Finch could have been axed if I’d been concerned about sprawl, but they’re two of my favorite plotlines, and in my opinion more than justify themselves.
• Give Yourself Outs: There were a few places in Worth the Candle that I gave myself a chance to bail out, but it only happened for real a few times, even if I sometimes did some test writing to see whether I could make it work another way. The biggest out I probably gave myself was with the Maddie stuff, which I wasn’t sure I was going to do, and wasn’t sure was going to work: there might be some alternate world in which that got skipped over, if I wasn’t feeling like I could do it well, or I wasn’t feeling like the work would benefit from it. There were a few other things that were ‘revelations’ that I reserved the right to go right on by, which in some ways made me more confident in them, because I didn’t feel like I had to write them. The biggest of these was probably the dream-skewered, who are ‘fake’, but because I gave myself an out, I was able to try writing them as real, a collection of people who thought they were from Earth. Giving yourself outs feels great and really improves the writing process.
• Write With Confidence: Not much to say here, especially because so much of it isn’t in the text itself. Worth the Candle is a bit special in that the metafiction elements mean that it often comments on itself, but I did my best to try to have those not undercut the work itself, and a lot of the places where something is pointed out as stupid or bad worldbuilding, this is meant to be the case (as with the Ell). When responding to comments, I try to be understanding of criticism, not too defensive, and professional. I do think that helps the work itself, to the extent people pay attention to you as an author and what you’re saying outside the text.
• Be Consistent: Not much to say here either. Consistent characterization helps a ton, as does having an end in sight, even if you have to wiggle. Knowing the thematic core of your work also helps a ton. Worth the Candle changes quite a bit over the course of its long runtime, but most of those changes are gradual and planned.