Organization 7 min read Updated November 2026

Save the Cat Beat Sheet Explained

The official beat sheet has 15 beats positioned by percentage, not page number. What it is, where the page numbers come from, and what it suits.

Save the Cat is a story structure created by Blake Snyder, built around fifteen beats running from Opening Image to Final Image. Its official description:

Save the Cat beat sheet: outline a novel in 15 beats

Deckle's Outliner showing every chapter and scene of a novel as a row with its synopsis, status and word count, ready to restructure. The book laid out in the Outliner

"The Save the Cat! Beat Sheet gives writers a clear roadmap of 15 story beats from Opening Image to Final Image, ensuring structure, pacing, and character transformation stay aligned in any medium."

Unlike most story frameworks, it is precise about where each beat falls, and the official positions are given as percentages of the whole.

The Fifteen Beats

Grouped into three acts, with the positions published on the official site. The acts are Syd Field's, from the three act structure, though Break into Three comes at 80% here rather than his 75%:

Act One

  • Opening Image, 0 to 1%
  • Theme Stated, 5%
  • Set-Up, 1 to 10%
  • Catalyst, 10%
  • Debate, 10 to 20%

Act Two

  • Break into Two, 20%
  • B Story, 22%
  • Fun and Games, 20 to 50%
  • Midpoint, 50%
  • Bad Guys Close In, 50 to 75%
  • All Is Lost, 75%
  • Dark Night of the Soul, 75 to 80%

Act Three

  • Break into Three, 80%
  • Finale, 80 to 99%
  • Final Image, 99 to 100%

Percentages, not page numbers. This matters. The page numbers you will see quoted, Catalyst on page 12, Break into Two on page 25, Midpoint on page 55, and so on, come from Snyder's 2005 book and assume a 110-page screenplay. The official site publishes percentages, which is what makes the structure portable to a novel at all.

For a 90,000 word novel, the Midpoint at 50% is around word 45,000, and All Is Lost at 75% is around word 67,500. That arithmetic is the whole reason the percentage form is more useful than the page form.

Start With the Hero, Not the Beats

Jessica Brody's advice is to know three things about the protagonist before plotting: the flaw (what holds them back), the want (the external goal they chase) and the need (what they must learn by the end). In her words, she would "start with a character first, at least those three things." The beats are then the external events designed to force that change, which is why a beat sheet built without them tends to read as a list of things that happen rather than a story.

Single-Scene and Multi-Scene Beats

The fifteen beats are not fifteen scenes. Brody, who coined the terms, splits them into two kinds:

  • Single-scene beats happen in one scene: Catalyst, Break into Two, Midpoint, All Is Lost and Break into Three.
  • Multi-scene beats run across many scenes and chapters, between the single-scene beats: Set-Up, Debate, Fun and Games, Bad Guys Close In, Dark Night of the Soul and the Finale.

Fun and Games alone covers 20 to 50% of the book. Brody keeps a long beat moving by giving it a goal, and each scene inside it a smaller goal of its own. The Finale, the last 20% of the novel, has its own structure, the Five-Point Finale: Gathering the Team, Executing the Plan, the High Tower Surprise, Dig Deep Down, and Execution of the New Plan.

Two Things People Confuse

The novel version is a different book by a different author. "Save the Cat! Writes a Novel is a book by Jessica Brody which adapts Blake Snyder's storytelling framework for fiction." Snyder wrote for screenwriters; the fiction adaptation is Brody's.

The ten "genres" are not beats. Snyder's categories, Buddy Love, Dude With A Problem, Fool Triumphant, Golden Fleece, Institutionalized, Monster In The House, Out Of The Bottle, Rites of Passage, Superhero and Whydunit, are a separate taxonomy of story type. They are routinely mixed up with the fifteen beats. They are not marketing genres either: "Monster In The House" is a structural category, not a shelf.

What It Is Good For, and Not

Good for: diagnosing a saggy middle, since the framework is most specific exactly where drafts most often fail; commercial fiction with a clear external plot; and giving a vocabulary to a structural problem you can feel but not name.

Less good for: literary fiction organised around something other than escalating external stakes, ensemble novels with no single protagonist, and any book where the percentages would force events rather than describe them.

The honest caution is about precision creating false confidence. A framework that tells you the Midpoint belongs at exactly 50% feels more rigorous than one that says "somewhere in the middle", but the precision comes from screenwriting, where a two-hour running time makes minute counts meaningful. A novel has no equivalent constraint, and a scene does not become correct by landing on the right word count.

Use the percentages as a diagnostic rather than a target: if your midpoint arrives at 70%, that is worth investigating. It does not mean it is wrong.

Note also that Save the Cat is a live commercial framework with an active trademark. The beat names and structure are freely described on the official site; the per-beat explanations belong to the books.

What Deckle Does

There is no Save the Cat template in Deckle, and no beat sheet. What exists is structure you can shape to it.

Folders and documents let you build the three acts and fifteen beats as a skeleton before writing, then write into it. On the Outliner's board, a single-scene beat fits as one scene card, and a multi-scene beat as a folder that you fill with its scenes as you plan them.

The Codex's Character template already has sections for Motivation & Goals, Flaw / Inner Conflict and Arc, which is where the want, the flaw and the need go before the first beat is written.

Labels are the practical beat tag: mark each scene with the beat it serves and filter by it, which shows immediately whether you have four scenes doing Fun and Games and none doing Debate.

Word counts per document are what make the percentages usable. The Outliner shows every document's length, and a folder's card shows the total of everything inside it, so with the acts as folders you can see what share of the draft each act takes and which scene is sitting near the 50% mark. That is the diagnostic use described above, and it needs no template.

Templates let you save a beat structure you have built once and reuse it, without the app choosing a framework for you.

For the other dominant framework, see the hero's journey structure. For whether to plan at all, see outlining vs discovery writing.

Common Mistakes

  • Quoting page numbers as official. The site publishes percentages; the page numbers assume a 110-page screenplay.
  • Confusing the ten genres with the fifteen beats. They are separate taxonomies.
  • Crediting the novel adaptation to Snyder. That book is Jessica Brody's.
  • Treating percentages as targets. They are diagnostic, and a novel has no running time.
  • Applying it to a book it does not suit. Ensemble and literary fiction often resist it.
  • Assuming precision means validity. The specificity comes from screenwriting constraints.

FAQ

How many beats are in Save the Cat?

Fifteen, from Opening Image to Final Image.

Where should the Midpoint fall?

The official positioning is 50%, which in a 90,000 word novel is around word 45,000.

Are the page numbers official?

The site gives percentages. The page numbers come from the 2005 book and assume a screenplay.

Is it only for screenplays?

Snyder wrote it for screenwriters. The fiction adaptation is a separate book by Jessica Brody.

Do I have to hit every beat?

No. Treat a missing or displaced beat as something to look at, not something to correct automatically.

Is each beat one scene?

No. Single-scene beats such as the Catalyst and the Midpoint are one scene; multi-scene beats such as Fun and Games and Bad Guys Close In run across many.

Sources

Try it against your own manuscript

Open the full app with a guided tour, import your Scrivener project, and export a chapter before deciding anything.

$197 once, every update included

Windows, Mac and Linux. 30-day money-back guarantee.