Most character building techniques produce a dossier.
You end up with a document. Height, eye color, childhood pet, the name of the village they grew up in, a zodiac sign you assigned in about four seconds. It is a real artifact and it took real hours and none of it changes a single thing that happens on page forty.
That is the failure mode nobody warns you about. Not blank pages. Filled ones that do no work.
This piece is about the build phase specifically. Before the draft, when the character does not exist yet and you are assembling one out of nothing.
Once you are drafting and the character needs to shift under pressure, that is a different job, and the character development techniques guide is the companion for that half. Here we are talking about construction.
The test I apply to every technique below is the same one, over and over. Does the output of this exercise change what happens in a scene?
If a technique gives you a fact, it failed. If it gives you a collision, it worked.
A character is a machine for generating trouble
Before build order, the thing you are building.
Reading a character on the page, you are not receiving a personality. You are watching someone decide under pressure, and the deciding is what makes them legible. Everything you know about Amy Dunne in Gone Girl you know because she chose to do something extravagant and then explained the reasoning in a voice that made the reasoning sound reasonable.
So the useful working definition is narrow. A character is a set of pressures arranged so that acting on one of them costs you another.
That definition throws out a lot of what gets sold as character work. Favorite color costs nothing, and a tragic backstory costs nothing on its own, because a wound with no present-day price is just decoration.
What costs something is a person who wants two incompatible things and has to keep choosing between them in front of us.
Start with the want, not the wound
The wound-first approach is everywhere. Figure out what broke them at age nine, and the adult falls out of it.
It rarely does. Backstory is enormously seductive to write, it feels productive, and it produces a character who is fully explained and completely inert. You know why they are afraid of water and you still do not know what they will do on Tuesday.
Start instead with a want that has an object. Not "she wants to be respected." She wants her name on the survey that redraws the county line, and she wants it before her brother's name goes on it.
The difference is that the second one can be granted or denied by other people, which means other people are now in the story. An abstract want has no antagonist. A specific want comes with a list of everyone standing in front of it.
Get concrete enough that you can name the scene where the want gets tested. If you cannot picture the room, the want is still too vague. The character goals and objectives work is worth doing properly here, because everything downstream is hanging off it.
Then build the need that contradicts the want
Now find the thing they actually require, and make sure it is incompatible with what they are chasing.
This is the oldest engine in fiction and it still runs. In Pride and Prejudice, Elizabeth wants to be right about people, and what she needs is to be wrong about Darcy without losing her sense of her own judgment. She cannot have both, and the entire middle of the novel is her paying for that.
The contradiction is the fuel. If want and need point in the same direction, the character walks a straight line and nothing interesting can happen along it.
A practical way to build this: write the want as a sentence, then ask what getting it would cost them. If the answer is "nothing, they would just be happier," you have not found the need yet. Keep going until achieving the want would require giving up something the reader can see they cannot afford to lose.
Give them a competence, then find the pressure it fails under
A character who is good at nothing is boring. A character who is good at everything is worse.
Build one specific competence, narrow enough to describe in a phrase.
She reads land deeds. He can talk anyone into a room. She knows exactly how much a body can take before it stops.
Then, and this is the part that gets skipped, design the situation where that competence is precisely useless. The man who can talk anyone into a room, in front of someone who does not speak. The woman who reads land deeds, in a country that keeps no records.
That pairing is worth more to a plot than twenty pages of backstory, because it tells you what scenes to write. You are no longer wondering what happens next. You are steering the character toward the one room their skill cannot open.
Sherlock Holmes is the useful counter-case here.
Observation is the competence and it almost always works, which is the appeal, so Conan Doyle has to find pressure somewhere else. In "The Adventure of the Yellow Face" he lets Holmes reason confidently to a conclusion that is simply wrong, and the story ends with Holmes asking Watson to remind him of it next time he gets sure of himself.
A competence that never fails needs at least one story that says what happens when it does.
Flaws work the same way and get treated badly for the same reason. A flaw listed on a sheet is trivia. A flaw that determines the outcome of a specific scene is structure, which is the distinction the character traits and flaws breakdown gets into in more depth.
Attach a loyalty that has a price tag
Wants and flaws are internal. Loyalty is where the character connects to everyone else in the book, which is why leaving it out produces characters who feel like they are standing alone in a white room.
Pick one person or institution the character will not abandon. Then set a price on it, specifically, in terms of the want from two steps ago.
She will not testify against her brother. Testifying is the only way her name goes on the survey. Now every scene has a temperature, because the reader knows both halves and can feel the moment coming.
The loyalty does not have to be admirable. Loyalty to a bad institution is one of the most useful things you can hand a character, because it explains complicity without excusing it, and complicity is the most common human condition there is.
What the popular techniques are actually for
I have been hard on questionnaires and typologies and I want to be fair about them, because they are not worthless. They are just misfiled.
The 100-question character sheet is a consistency tool rather than a build tool. Once you have a character who works, filling out the sheet keeps their coffee order and their sister's name stable across four hundred pages, which is a real problem that real drafts have.
Using it to generate a character is like using a spellchecker to write the sentence.
Enneagram and MBTI typing have a similar shape. A type gives you a plausible behavioral floor, so when you are stuck on how someone would react in a low-stakes moment, the type produces an answer that will not contradict the rest of the book. It cannot give you a want with an object, because types describe tendencies and stories run on decisions.
Extensive backstory has one genuinely great use, and it is not the file. It is the two or three details you pull out of the file and put in a scene. The other forty pages are scaffolding that you take down after the building stands.
Here is how the common techniques sort out once you ask the only question that matters.
| Technique | What it produces | Does it change a scene? |
|---|---|---|
| Character questionnaire | Facts, cross-referenced | No. Keeps facts consistent once the character exists |
| Enneagram / MBTI typing | A default behavioral register | Rarely. Useful for small moments, not decisions |
| Extensive written backstory | Explanation for existing traits | Only the 2-3 details you actually stage |
| Physical description sheet | A visual, and sometimes a class signal | Sometimes, when the detail is contested by someone |
| Concrete want with an object | A list of people in the way | Yes. Generates opposition automatically |
| Want-need contradiction | A cost on every choice | Yes. This is the middle of your book |
| Competence plus its failure case | Scene assignments | Yes. Tells you which rooms to write |
| Priced loyalty | Stakes attached to other characters | Yes. Puts pressure on relationships |
| Voice test under pressure | Distinct dialogue and interiority | Yes, once the pressure is real |
The pattern is not subtle. Everything in the top half describes a person. Everything in the bottom half puts a person in a situation they have to get out of.
Voice comes last, and it comes from pressure
Writers try to build voice first because voice is the most visible part of a character, and it almost never works in that order.
Distinctive speech built before the want is just a verbal tic. A stutter, a catchphrase, a habit of answering questions with questions. Readers clock it in one scene and get bored of it in three.
Voice that lands is a byproduct of what someone is protecting. Once you know the want, the contradicting need, and the loyalty with the price on it, you know what the character deflects, what they will not say out loud, and where they get precise because precision is a defense. That is voice.
The build test is simple. Put the character in a scene where they must ask for something they are ashamed of wanting, and write only their dialogue.
If the shame is legible in the word choice, the voice is real. If it sounds like anyone, go back to the want, because the want is not specific enough yet.
Build in the right order, because the order is load-bearing
Each of these depends on the one before it, which is why the usual grab-bag of tips does not compound.
- Want, with a physical object and a deadline. Something a named person can grant or refuse.
- Need that contradicts the want. Getting the first must cost the second. If it does not, keep looking.
- Competence, narrow. One skill you can name in a phrase, plus the exact circumstance where it is worthless.
- Loyalty with a price. One person or institution, and the specific way it collides with step one.
- Two or three backstory details. Only the ones that explain a behavior the reader will see. Write more if you like, stage only these.
- Voice under pressure. Test it with the shame scene. Revise upward from there.
- Consistency pass. Now fill out the questionnaire, if you want one. It works fine at this stage.
Steps one through four are the build. Steps five through seven are finishing work.
The reason order matters is that later steps produce garbage when earlier ones are missing. Backstory written before you have a want will explain traits the character does not need, and voice written before the loyalty exists will be decoration.
It is much faster to do four things in sequence than nine things in a pile.
If you want structured prompts for running this, the character development exercises set has drills you can point at each step, and the broader character development hub covers where the build phase sits in the whole process.
Test the build before you draft
You do not need a finished character. You need one that survives contact with three scenes.
Write the character being denied their want by someone with a good reason. Not a villain, someone with a defensible position. If the scene has heat, the want is specific enough.
Then write them succeeding at their competence in a way that makes things worse. If you cannot construct that scene, the competence is too generic and needs narrowing.
Then write them choosing between the want and the loyalty, and let them choose wrong. A character who always chooses correctly is a mascot. The wrong choice is where readers start believing there is a person in there, and it is also where a build turns into a plot, which is roughly where developing complex characters picks up.
Three scenes, maybe two thousand words, most of which you will throw away. Cheaper than discovering in chapter eleven that your protagonist has no reason to leave the house.
The build phase ends when the character starts arguing
There is a moment where you plan a scene and the character will not do the thing you planned.
Writers describe this in mystical terms and it is not mystical. It happens when the constraints you built are specific enough to conflict with your outline, and your outline is what loses. The character has a want with an object, a loyalty with a price, and a skill that does not apply here, so the move you had in mind is unavailable to them.
That resistance is the whole goal of the build phase. Not a character you know everything about. A character with enough load-bearing constraints that the story has to route around them.
Everything in this piece is aimed at getting to that point before you write chapter one, because getting there in chapter eleven means rewriting ten chapters.
Frequently Asked Questions
What is the difference between character building and character development?
Building is construction before the draft, where you assemble wants, needs, competences, and loyalties for someone who does not exist yet. Development is what happens to that character across a draft, as pressure changes them and the reader watches it happen. Build first, develop during.
Do I need a character questionnaire at all?
Not to create a character. They are genuinely useful after the build, as a consistency reference so details stay stable across a long draft. Filling one out on a blank character mostly generates facts you never use.
How much backstory should I write before drafting?
As much as you enjoy writing, but only stage two or three details. The test is whether a detail explains a behavior the reader will actually see. If it explains something that never appears, it is a note to yourself, not part of the book.
How do I stop my characters from all sounding the same?
They sound the same because they want the same thing, or because none of them want anything specific. Voice is downstream of pressure. Give two characters incompatible wants and put them in the same room, and the difference shows up in the dialogue without you doing anything clever.
Can I use these techniques for side characters?
Yes, and it is faster than it sounds. A side character needs a want and one thing they will not do. That is often two sentences of work and it is enough to keep them from being furniture.
How long should the build phase take?
For a novel protagonist, a day or two of real thinking plus the three test scenes. If you are still building after two weeks, you are usually avoiding the draft. The build is deliberately unfinished, because the rest of the character comes from writing them.
