Structure Guide

Three-act structure: where should the major beats fall?

Beat percentages are useful diagnostic ranges, not laws. The more important question is whether each turning point changes the protagonist’s objective, strategy, knowledge, or available choices.

Typical broad shape

Act I: roughly the first 20–30%

Act II: roughly the middle 45–60%

Act III: roughly the final 20–30%

Beat placement ranges at a glance

BeatCommon rangeDramatic jobWarning sign
Inciting disturbance0–15%Destabilizes the existing life and introduces the central trouble.The story can continue unchanged if ignored.
First act break20–30%Commits the protagonist to the main arena, objective, or consequence.The protagonist remains a spectator after the turn.
Midpoint45–60%Changes understanding, agency, stakes, or strategy.It is only a larger version of earlier activity.
Second act break / crisis70–85%Exposes the failure of the current approach and creates the final choice.It merely delays the same plan.
Climax80–98%Forces decisive action against the central conflict.A new device or helper solves what the protagonist could not.
ResolutionFinal 2–10%Shows the consequence and new equilibrium.The story ends as soon as the mechanism is solved.

These ranges overlap because stories distribute work differently. A mystery may introduce the crime immediately but withhold the detective’s true commitment. A romance may establish attraction early while the relationship’s irreversible turn arrives later. An epic novel may spend longer establishing multiple lines than a ninety-minute thriller. Placement matters when it changes the audience’s experience—not because a spreadsheet demands an exact percentage.

Act I: establish and commit

The first act establishes the story’s normal pressures, delivers the disturbance, and makes the central conflict unavoidable or worth pursuing. It should provide enough context to understand why the protagonist’s commitment matters, but not require every piece of history before the story can move.

Diagnostic: At the act break, can you state what the protagonist is now trying to accomplish and what makes withdrawal costly? If not, the opening may be full of setup but short on commitment.

Act II: adapt and escalate

The middle tests the first strategy, compounds consequences, develops relationships, and changes the protagonist’s understanding. The midpoint should divide the act functionally: before it, the protagonist may be learning the arena; after it, choices should become more deliberate, costly, or desperate.

Diagnostic: Compare the plan immediately before and after the midpoint. If it is unchanged, ask what new truth, loss, victory, or exposure could make repetition impossible.

Act III: choose and prove

The final act compresses options and brings the external conflict together with the character’s deepest unresolved pressure. The climax is strongest when the protagonist must use what has been learned and accept a cost the opening version of the character would have avoided.

Diagnostic: Could the same climactic solution have been attempted in Chapter 2? If yes, the middle may not be contributing enough knowledge, capacity, relationship, or change.

When an unconventional shape works

Some novels use four parts, five acts, braided narratives, episodic structures, or deliberately delayed commitments. Those structures still need orientation, development, turns, and payoff. Use three-act language as a lens for dramatic function, not a mandate for visible labels.

An intentional slow opening works when tension, curiosity, and consequence are already accumulating. A late climax works when the ending still has room to register its cost.

How to calculate beat positions without worshipping the numbers

For a chapter outline, divide the chapter number by the total chapter count. Chapter 8 in a 32-chapter novel begins around 25 percent. For a screenplay or timed format, divide the page or minute by the estimated total. Use the result to spot large imbalances, then return to dramatic function.

A beat at 34 percent is not automatically late. It becomes a problem if the preceding third feels provisional, repeats setup, or withholds the real objective. A climax at 75 percent may be early if the remaining quarter lacks a new dramatic question; it may be appropriate if the aftermath contains a consequential second resolution.

Placement problemWhat the reader may feelBetter question than “move it”
Late first act breakThe real story has not started.Which setup can occur after commitment under greater pressure?
Early midpointThe second half repeats or runs out of escalation.Is this actually a first major reversal, with a deeper reorientation still missing?
Late crisisThe climax feels compressed or convenient.What decision needs more room to generate the final plan?
Early climaxThe ending feels overlong.Does the aftermath contain a distinct unresolved conflict, or can it be compressed?

A worked example

In a 40-chapter political fantasy, a diplomat discovers an assassination plot in Chapter 3 (7.5%). She reports it but tries to remain neutral. In Chapter 10 (25%), her evidence is framed as treason and she must flee with the intended victim: the first act break commits her. In Chapter 21 (52.5%), she learns the victim ordered an earlier atrocity, changing protection into a moral dilemma. In Chapter 31 (77.5%), her attempt to expose both factions causes a coup and destroys her credibility. The climax in Chapters 36–38 requires her to release evidence that also condemns her family.

The percentages are ordinary, but the sequence works because every turn changes the story’s meaning and available strategy. Moving a weak beat to the “correct” page would not repair it if nothing changes afterward.

For a complete audit, continue to story structure analysis, the 15-point outline checklist, or the guide to fixing a slow second act.

Last substantially updated October 28, 2026.

See The Shape

Compare beat placement with what the story is actually doing.

Outline Analyzer maps structure and evaluates the dramatic work around the turns, rather than grading percentages alone.