Flesch-Kincaid and Beyond: A Practical Guide to Readability Metrics

Readability metrics were invented to solve a practical problem: how do you objectively measure whether a document is appropriate for its intended audience? The metrics that emerged from this work are imperfect but useful — when you understand what they measure and what they do not.

The core idea behind readability metrics

All readability metrics work on the same basic observation: longer words and longer sentences are harder to read. Different metrics operationalise this observation in slightly different ways, producing scores that correlate with reading difficulty.

The scores are typically expressed as grade levels — a score of 8 means the text is appropriate for an 8th-grade reader in the US system. This makes them immediately interpretable: compare the document's grade level to the expected grade level of the audience.

Flesch-Kincaid Grade Level

The Flesch-Kincaid Grade Level formula:

Grade Level = (0.39 × average sentence length) + (11.8 × average syllables per word) − 15.59

It weights sentence length and word complexity into a single number. A score of 12 means the text is appropriate for a high school senior. A score of 16 corresponds to a college degree. Professional and business writing typically targets grades 8 to 12 for general audiences.

The formula is fast, widely implemented, and well-validated against comprehension studies. Its weakness is that it ignores syntactic complexity beyond sentence length — a sentence with five embedded clauses reads harder than one with a simple sequence of independent clauses, even at the same word count.

Flesch Reading Ease

The companion metric, Reading Ease, runs the same formula in reverse — higher scores mean easier text. Scale: 90-100 is very easy (5th grade), 60-70 is standard (8th to 9th grade), 0-30 is very difficult (professional).

Consumer-facing health information should target above 60. Academic papers routinely score below 30. Legal documents often score below 20.

SMOG Index

SMOG (Simple Measure of Gobbledygook) counts polysyllabic words — words with three or more syllables — in a sample of 30 sentences. The formula: 3 + √(count of polysyllabic words).

SMOG is better calibrated for documents with high conceptual density. It is the preferred metric for health communications and medical documents because it correlates more strongly with comprehension in populations with mixed reading abilities.

Gunning Fog Index

Gunning Fog = 0.4 × (average sentence length + percentage of words with three or more syllables)

A score above 12 indicates that the text is too complex for most readers. Business and general-purpose professional writing should target below 12.

Gunning Fog is sensitive to jargon — technical terms are typically polysyllabic and inflate the score significantly. This makes it useful for identifying whether a document is using more specialised language than the audience can handle.

Coleman-Liau Index

Coleman-Liau uses character counts rather than syllable counts, making it faster to compute in software. It tends to give slightly higher scores for technical writing than Flesch-Kincaid.

What metrics do not measure

All of these metrics measure proxies for readability, not readability itself. They do not capture:

**Syntactic complexity.** A sentence with multiple embedded clauses reads harder than a simple sentence of the same length. The metrics do not distinguish these.

**Conceptual density.** A sentence that introduces three new concepts reads harder than one that applies a familiar concept, regardless of word length.

**Organisation.** A well-organised document is easier to read than a disorganised one with the same sentence-level complexity. Metrics ignore structure.

**Assumed knowledge.** A sentence that is accessible to an expert is inaccessible to a non-expert at the same grade level. Metrics do not account for audience background.

These limitations mean readability metrics should be used as diagnostic tools, not as definitive assessments. A document that scores well on all metrics may still be difficult to read for its audience. A document that scores poorly may be appropriately complex for a specialist reader.

Practical use

Use readability metrics as a screening tool. A document that scores significantly above the expected audience level is a candidate for plain-language revision. A document that scores within range may still have specific problematic sections — run metrics on sections independently, not just on the whole document.

For documents with a regulated readability requirement — some jurisdictions require insurance policies, financial disclosures, and consumer contracts to meet specific grade levels — use the metric specified by the regulation and document the score.

For other documents, treat readability scores as a prompt for review rather than a pass/fail criterion. The question is not "does this score 8?" but "are the sections with the highest scores the sections where high complexity is necessary?"