The Scene Goal–Conflict–Disaster Method
How to structure a fiction scene with a named Goal–Conflict–Disaster pattern, including detective and romance examples and a revision checklist.
A scene built on the Goal–Conflict–Disaster method gives a character a concrete want, puts a force in the way, and ends on a turn that ruins or complicates the original plan. That three-beat spine is how a page of talk or action becomes plot instead of activity.
Dwight Swain popularized a version of this in Techniques of the Selling Writer. You do not need his whole system. You need the habit: enter with a try, meet a no, leave with a new problem.
Goal: a want you could film
A usable scene goal is visible. The reader should be able to tell, mid-page, whether the character is getting closer or farther.
Good goals:
- Get the witness to name the second man in the dock.
- Get through dinner without mentioning the pregnancy test.
- Get the spare key back before her sister arrives.
Weak goals:
- Have a conversation.
- Show that she is fierce.
- Worldbuild the city.
If you catch yourself writing a weak goal, ask: what would count as winning in the next twenty minutes of story time? That answer is the goal.
Goals can be inner, but they still need an outer test. “Prove I am not my father” becomes “do not take the bribe when it is offered in this office.” The inner want rides on the outer try.
This is plot at scene scale. Plot is not a separate weather system. It is wants colliding.
Conflict: a no that can last
Conflict is not volume. It is incompatible aims, limited resources, or a truth that cannot be spoken yet.
Types of opposition that work in a single scene:
- Another will. He wants secrecy; she wants the story.
- A constraint. The train leaves in four minutes; the confession needs ten.
- A social rule. You cannot accuse your host at his own table without becoming the villain of the room.
- An inner veto. She could ask, and her pride will not let the question out.
Stack one primary opposition. Three villains in one kitchen is noise. One sharp no, tested twice, is drama.
Conflict should be allowed to win. If you know the hero will get the answer because the plot “needs” it, the scene goes slack. Let the goal be in genuine danger. That is how you raise the stakes inside a single unit rather than only at act turns.
Disaster: the options change
Disaster is a nickname for a turn. The attempt ends in a state the character did not budget for.
Four useful disaster types:
- Outright failure. The witness walks. The kiss does not happen. The door stays locked.
- Success with a price. She gets the name and is now a target.
- Wrong success. He gets a yes from the person who cannot actually help.
- New information that kills the plan. The letter is a forgery. The alibi was for a different night.
The disaster should move the plot forward. If the character can shrug and try the same tactic tomorrow, you wrote inconvenience, not disaster.
After a disaster, many stories need a sequel beat—reaction, dilemma, decision—so the next goal is chosen. That pairing is the scene sequel method. Goal–Conflict–Disaster is the left hand. Sequel is the right.
Worked example: detective questioning
Goal: Detective Ruiz wants Len, the night guard, to admit a second person was on the loading dock at 2:14 a.m.
Conflict: Len wants to keep his job and his self-image as someone who “didn’t see anything.” He is not clever. He is loyal and tired. He can stonewall.
Attempts:
- Ruiz starts small: coffee, the broken machine, a joke about the shift. Rapport is a tactic, not kindness.
- He offers two timestamps. Len supplies a tidy story.
- Ruiz lets silence sit. Len fills it with extra detail—always a tell.
- Ruiz mentions a camera angle he does not have. Len’s gaze flicks to a real blind corner.
Disaster: Len does not confess. He names a wrong person—a junior temp who was not even on the schedule. Ruiz cannot use the statement in court. Worse, the real protector (a supervisor) now knows Len was questioned. The investigation did not stall. It got more dangerous and more specific. Next scene cannot be “ask Len again, louder.” It has to be the supervisor, or Len in trouble, or both.
That is how you connect scenes into a plot: the disaster writes the to-do list.
Worked example: romantic dinner gone wrong
Goal: Priya wants one clean evening—no talk of the missed gallery opening—so she can decide whether she still wants the marriage.
Conflict: Julian wants the evening to prove that nothing is wrong. His goal is incompatible with hers. He will charm, refill, reframe.
Attempts:
- She chooses a public restaurant so neither of them can detonate. Constraint as strategy.
- He tells a funny work story. She laughs on purpose. Both are performing.
- She asks a small question: “Was Maya there?” He says no too fast.
- She does not show the photo. She waits. He fills the wait with a second lie, smaller, about traffic.
Disaster: The waiter brings complimentary dessert “from Maya at the bar,” who is not at the bar. Julian did not arrange it. Someone else is in the marriage’s atmosphere. Priya does not throw the plate. She thanks the waiter. The disaster is social and irreversible: the private problem is now public-shaped. She cannot go home and pretend the night was a reset. The next scene is about who Maya still is to them, not about whether the risotto was good.
Notice the disaster is not a screaming match. It is a change of options.
How to write the middle without padding
Between goal and disaster, show tactics. A character who wants something tries more than one move.
- Ask.
- Bargain.
- Threaten.
- Charm.
- Withhold.
- Offer a lesser truth.
Two or three tactics are enough for most scenes. If you need seven, you may have two scenes glued together.
Keep time continuous when you can. Jumping across “an hour later” inside a supposed scene often means you needed summary plus a new Goal–Conflict–Disaster unit.
Revision checklist
- Can a stranger name the goal in one sentence?
- Could the opposition actually win?
- Are there at least two attempts?
- Does the last beat change the available plans?
- Does the next chapter require this turn?
- Did you confuse atmosphere with conflict?
If the scene is pretty and still fails, cut toward the try. Pretty is not a beat.
Where writers get stuck
They write a goal the character already has in the bag. They write conflict as personality (he is “gruff”). They write disaster as a cliffhanger that does not change knowledge or options—“and then the phone rang” with no consequence.
Fix by raising the cost of failure inside this hour. Not the fate of the kingdom. The fate of this attempt.
Stacking with the rest of the craft
Use Goal–Conflict–Disaster to build the pursuit. Use sequel to let the character choose. Use stakes work so the no hurts. Use scene-to-scene connection so disasters accumulate instead of resetting.
You can draft the three sentences in a notebook, then write. You can also reverse-engineer a limp chapter: highlight the want, the no, the turn. If one highlight is missing, you found the repair.
The method is small on purpose. Small methods get used.
When the three beats are named and you still want more texture on the page, PlotnAI can help you thicken attempts and sensory pressure without replacing the disaster you already chose.