How to read the analysis
What the studio reader is noticing — and what it isn't.
Folio describes your manuscript. It never grades you. Every observation in the margin is a measurement against the age band you chose — change the band and the same text reads differently, because the reader changed, not the writing.
Where a metric rests on published research, the source is named below. Where it's a working convention of the engine, we say that too. Draft notices changes across the book; Listen supports read-aloud rehearsal; Language opens the vocabulary, sound, and sentence evidence.
Structure, context, and examples
The structure under the manuscript title is the physical working dummy: its spread count and default page composition. Changing structure creates a separate copy and keeps the source in Manuscripts; Folio never automatically rewrites or reflows prose.
Reading context is analysis only. Audience and experience change which evidence Folio foregrounds, but never alter words, pages, formatting, or export structure. Explore opens examples read-only and always provides a route back to the writer's saved manuscript.
Length
Word count, measured against the publishing conventions for your band (SCBWI guidance; Mary Kole's length guidelines):
| Band | Ages | Words |
|---|---|---|
| Board | 0–3 | up to 100 |
| Early picture | 0–3 | up to 400 |
| Picture | 3–7 | up to 600 |
| Early reader | 5–9 | 1,000–2,500 |
| Chapter | 7–10 | 5,000–10,000 |
Sentence counts use real segmentation — "Now run along," said Mrs. Rabbit. is one sentence, despite the period in Mrs. — and average sentence length tracks the breath of your prose read aloud.
Targets are conventions, not laws. But editors feel them, and the youngest readers feel them most.
Vocabulary
Familiar words (Tier 1)
The share of your words a young reader is likely to already know — measured against the Dale–Chall familiar-word list (words roughly 80% of fourth-graders recognize in print) plus the Dolch and Fry sight words. This is the broad "how much of this is already known" number. A note on calibration: the list is normed at the top of our age range, so it's most discriminating for early-reader and chapter texts and naturally sits near 100% for board books.
Sight-word coverage
A narrower measure: the share of your words on the Dolch and Fry lists specifically — the high-frequency words beginning readers learn to recognize whole, without sounding out. High coverage means an early reader moves through your text on recognition; lower coverage shifts the work to decoding.
Unique words & type-token ratio
Unique words divided by total words. In a board book, a low ratio is usually craft — repetition is how the youngest readers participate. In a chapter opening, a higher ratio is normal. There is no good score; there is only the question of whether the repetition you see is the repetition you meant.
Reach words
Words likely to sit just past your target reader's current mastery — flagged because they fall outside the familiar Tier-1 core (the Dale–Chall list plus sight words). The writing canvas stays unmarked by default; the Language lens attributes each one to the spread where it first appears when you choose to inspect vocabulary.
Some reach words are the point. The zone just past mastery is where vocabulary grows — educators call it the zone of proximal development (Vygotsky) — and shared reading is how children cross it. A handful per spread invites growth. A pile-up on one spread turns reading into work.
One honest limit: a character's name and an inflected form like "bunnies" both read as reach words, and the tool doesn't yet separate a rich literary word ("mischief") from a genuinely rare one ("sirrah"). Trust your eye over the count.
Phonology
Decodability
A 0–1 estimate of the sound-out load your words carry. It weights how early each speech sound is typically mastered (consonant norms from Crowe & McLeod, 2020 — the research base SLPs already use; strictly, norms for saying sounds, which we use as a stand-in for decoding them) and how complex each syllable's shape is. The exact weighting is an engine convention, documented in the source.
Read it with its scope in mind: it matters most when the child is the reader — the early-reader band. Board and picture books are read aloud by an adult, so a low score there isn't a warning. And two honest limits: the score counts sight words ("the", "this") as if they were sounded out, even though emerging readers learn them whole — so high-repetition text can score lower than it should — and it doesn't account for word length. For read-aloud bands, the phoneme inventory below is the more useful surface.
Syllables
Average syllables per word, plus the shapes: CV ("go") and
CVC ("cat") are the friendly ones; consonant clusters
(CCVC, "stop") arrive later in development.
Phoneme inventory
Every speech sound your manuscript uses, grouped by the age most children produce it accurately — later-acquired sounds (r, th) tint warmer. Click any phoneme to light up every word containing it, in place. Useful for spotting a sound your text leans on, or for matching a text to a child's current inventory.
Guessed pronunciations
Folio's pronouncing dictionary covers the core early-reading vocabulary. Anything outside it — invented names, rare words — gets an estimated pronunciation, and this list names those words. It's the engine being honest: phoneme and meter readings for these words rest on guesses. If your manuscript stars a Snorfblat, weigh the phonology numbers accordingly.
Prosody
Dominant meter is the stress pattern your lines lean toward: iambic (da-DUM), trochaic (DUM-da), anapestic (da-da-DUM), dactylic (DUM-da-da). Meter consistency is how much of the text fits that pattern, and the rhyme scheme letters (AABB, ABAB) mark which line endings rhyme.
Prose reads "—" or "mixed", and that is not a defect — most picture books aren't metered. Verse with a deliberately loose beat reads "mixed" too, and so can verse built on words the dictionary had to guess. Lines are read the way you broke them: the meter starts fresh on each line, and an unstressed pickup syllable before a line's first beat is understood, not penalized.
If you write metered verse, watch consistency while you revise — a dip often points at the exact line where the beat stumbles.
Per spread
The pacing strip. Each bar is one spread's word count against a soft ceiling — the manuscript target spread across your non-wordless spreads, with headroom. The ceiling is a documented heuristic, not a rule: picture books distribute words unevenly on purpose. The strip exists so the unevenness is a choice you can see.
Warnings
Folio currently warns about two things: total word count over your band's maximum, and under its minimum where one exists. "Info" severity means noted, not fix it.
What Folio doesn't do
It doesn't score you, rank you, or compare you to anyone. It will never write or rewrite your text — not a word. And your manuscript never leaves your browser: analysis runs locally, drafts save to your device, and there's no account and no server.