Skip to content

15 Writing Metrics Every Writer Should Track

The short answer

Track a few metrics tied to what the piece must do. Length (words, characters, reading time) tells you whether it fits the brief or the platform. Readability (sentence length, syllables per word, grade level) tells you how hard it is to read. Style signals (hard sentences, passive voice, adverbs, repetition, rhythm) point to the sentences worth a second look.

Use them as flags, not scores to chase. Below are 15 metrics, each with what it measures, exactly how WritePadPro calculates it, a sensible target and what to do when the number looks off. Our How We Count page documents the full counting rules.

Length metrics

1. Word count

WritePadPro counts a word as any run of characters between spaces, tabs or line breaks, the same rule word processors use. So “well-known”, “don’t”, “42” and a URL each count as one word. Chinese, Japanese, Thai, Lao, Khmer and Burmese, which are written without spaces, are split into words with the browser’s built-in word segmenter.

Target: whatever your brief, assignment or platform sets. For web articles there is no magic length; match the depth the reader’s question needs. Act: set a word goal in the WritePadPro editor workspace and watch the ring fill.

2. Character count

Characters are counted as the characters you see. An emoji, an emoji with a skin tone, a flag or an accented letter counts once, even when it is stored as several code units: “Apple 🍎 café 👍🏽 x” is 16 characters, though JavaScript’s length property reports 20. The no-spaces count removes every kind of whitespace, including line breaks and non-breaking spaces.

Target: the platform’s limit. Platforms often apply their own rules (X, for example, counts every link as 23 characters), so leave a margin. Act: use the editor’s character limit bar, with presets for 60, 160, 280, 500, 2,200 and 3,000 characters.

3. Reading time

Words ÷ 238 words per minute, the average silent reading speed for adults reading non-fiction reported in a 2019 review of reading-rate studies by Marc Brysbaert. Results are shown in minutes and seconds.

WordsReading (238 wpm)Speaking (150 wpm)
5002 min 6 sec3 min 20 sec
1,0004 min 12 sec6 min 40 sec
1,5006 min 18 sec10 min 0 sec
2,0008 min 24 sec13 min 20 sec

Target: none in itself; it sets expectations. Real readers vary a lot, and skimmers read faster while technical text slows everyone down. Act: if a “quick guide” reads in 12 minutes, it isn’t quick. Cut or split it.

4. Speaking time

Words ÷ 150 words per minute, a comfortable conversational pace. The editor also shows handwriting time at 13 words per minute. The Speaking Time Calculator lets you set any pace from 60 to 300 wpm.

Target: your time slot, minus a buffer for pauses and slides. Act: time yourself reading one page aloud and use your own pace from then on.

Structure metrics

5. Sentence count

A sentence ends at . ! ? or … followed by a space or the end of the text, and also at a paragraph break, a list item or a heading, so an unpunctuated heading isn’t merged with the next sentence. The counter doesn’t split after decimals (3.14), web addresses, initials (J. K. Rowling), titles (Dr., Mr.), “e.g.”, “No. 5” or “etc.” followed by a lowercase word. “Dr. Smith arrived at 3.30 p.m. on Monday.” counts as one sentence.

Act: sentence count matters mostly as the input to the next metric.

6. Average sentence length

Words ÷ sentences. It is the single strongest lever in most readability formulas.

Target: for web and business writing, an average of roughly 15 to 20 words is a practical working range (our suggestion, not a standard), with real variety around it. Act: the Readability Checker lists every sentence over 25 words; split those first.

7. Paragraph count and length

In the editor, each block you create with Enter counts as a paragraph. The standalone tools count blocks separated by a blank line, the usual rule for plain text.

Target: short paragraphs for screens. The federal plain language guidance on digital.gov suggests no more than about 150 words per paragraph. Act: the editor’s Long paragraphs highlight flags any paragraph over 150 words.

Readability metrics

8. Syllables per word

Syllables are counted with an English rule set: vowel groups, adjusted for silent “e”, endings such as -ed, -es and -le, vowel pairs said as two syllables (me-di-a) and a list of irregular common words. Numbers have no syllables. It is an approximation, because English spelling doesn’t map exactly to sound.

Why it matters: in Flesch Reading Ease, every extra 0.1 syllables per word costs 8.46 points; in Flesch–Kincaid, it adds 1.18 grade levels. By comparison, one extra word per sentence costs only 1.015 points and adds 0.39 grades. Long words move the scores fast. Act: replace long words only where a shorter one means the same thing.

9. Flesch Reading Ease

206.835 − 1.015 × (words per sentence) − 84.6 × (syllables per word), shown on a 0–100 scale where higher is easier. 60–69 is “standard”, 70–79 “fairly easy”, below 50 “fairly difficult” or harder.

Target: 60 or higher for general audiences is a sensible start. Act: work on sentence length first, then word length.

10. Grade level

Five formulas estimate a US school grade: Flesch–Kincaid, Gunning Fog, SMOG, Coleman–Liau and the Automated Readability Index. The editor’s “6 scores” panel shows all of them with their average; SMOG appears once the text has three sentences.

Target: about grade 7 to 9 for public web content, lower for instructions. Act: see our reading level guide for the formulas, why they disagree and how to lower the grade.

Style metrics

11. Hard and very hard sentences

The editor grades each sentence on its own with the Automated Readability Index. A sentence of 14 or more words is marked hard at grade 10 to 13 and very hard at grade 14 or above. Any sentence of 30 or more words is very hard regardless.

Target: a few hard sentences are normal in serious writing; very hard ones deserve a rewrite, and instructions should have none. Act: press Alt + Down arrow (Option + Down arrow on a Mac) in the editor to jump from one highlighted sentence to the next.

12. Sentence-length variety (Rhythm)

The editor’s Rhythm score (0–100) needs at least three sentences. Seventy percent of it comes from how much sentence lengths vary (the standard deviation divided by the mean): too little variation scores low, and very high variation scores slightly lower again. The other 30 percent penalizes runs of neighboring sentences within two words of each other’s length. Labels run from Monotonous through A little uniform and Good variety to Lively rhythm (80 and above).

Act: if the score is low, add a short sentence after a long one, or join two short ones.

13. Lexical diversity

The editor’s “Unique” figure is the type-token ratio: distinct words ÷ total words. The Vocabulary Analyzer adds Root TTR (distinct words ÷ √total words), Corrected TTR (distinct ÷ √(2 × total)), words used only once, lexical density (the share of words that aren’t common function words) and a word-length chart.

Caution: the plain ratio falls as a text gets longer, because common words keep repeating. Compare only samples of similar length, ideally 200 words or more. Act: a low figure in a short text usually means one word is overused; the Repeated words list in the editor’s Analyze tab shows which.

14. Passive voice

The editor flags a form of “be” (am, is, are, was, were, be, been, being), optionally followed by a word like “not” or “often”, then a past participle: a regular -ed form or one of about a hundred irregular ones such as “written” or “taken”. Words that usually work as adjectives, such as “tired” or “excited”, count only when “by” follows. It reports a count, not a percentage.

Target: no fixed limit. If you want a rate, divide the count by the number of sentences and watch the trend. Act: rewrite when the actor matters; keep the passive when the actor is unknown or beside the point.

15. Adverbs and qualifiers

The Adverbs highlight marks -ly words, minus a long list of -ly words that aren’t adverbs (only, family, early, likely, friendly and others). The Qualifiers highlight, off by default, marks softeners such as very, really, quite, just, basically, “sort of” and “I think”.

Target: look for clusters rather than counting every one. Act: replace an adverb plus a weak verb with a stronger verb (“walked slowly” → “ambled”), and keep qualifiers that carry real meaning, such as “probably” in a forecast.

Which metrics matter for which job

WritingWatch closelyGlance at
Essay or assignmentWord count, pages, sentence lengthGrade level
Blog post or articleReading time, Flesch Reading Ease, hard sentences, paragraph lengthPassive voice, lexical diversity
Social post or bioCharacter count against the platform limitQualifiers
Speech, video or podcast scriptSpeaking time, sentence lengthRhythm
Instructions or help articleGrade level, very hard sentences, passive voiceReading time
FictionRhythm, repeated words, adverbsGrade level

How to use metrics without gaming them

  • Measure finished drafts, not every keystroke. Scores on a half-written page mostly measure what is missing.
  • Use enough text. Readability and diversity figures are unstable below about a hundred words.
  • Compare like with like. Track the same metric, in the same tool, across drafts of the same kind of writing.
  • Fix the sentence, not the number. Chopping sentences into fragments lowers the grade and wrecks the rhythm. If a change makes the text worse to read, undo it.
  • Remember what isn’t measured. No metric checks accuracy, logic, structure or whether you answered the reader’s question.

All 15 metrics are calculated in your browser. The editor shows most of them live in its sidebar and highlights; the Text Statistics tool shows a dashboard for any pasted text. For a plain-English tour of each number, see Text Statistics Explained.

Frequently Asked Questions

Which writing metric matters most?

It depends on the job. For limits and briefs, word or character count matters most. For how easy a text is to read, average sentence length has the biggest single effect on readability scores, followed by word length in syllables. Pick two or three metrics that match the goal of the piece.

How does WritePadPro calculate reading time?

It divides the word count by 238 words per minute, the average adult silent reading speed for non-fiction reported in a 2019 review of reading-rate research, and shows the result in minutes and seconds. A 1,000-word article reads in about 4 minutes 12 seconds.

Why does the WritePadPro character count differ from other tools?

WritePadPro counts characters as you see them, so an emoji, a flag or an accented letter made of several code points counts once. Tools that count code units or code points report higher numbers for the same text, and social platforms such as X apply their own weighting.

What is a good type-token ratio?

There is no universal good value, because the ratio falls as texts get longer. Compare texts of similar length, ideally 200 words or more, or use the Root or Corrected TTR in the Vocabulary Analyzer, which are designed to be less sensitive to length.

Does WritePadPro show passive voice as a percentage?

No. The editor shows the number of passive constructions and highlights each one. To get a rate, divide that count by the number of sentences, and use it to compare drafts rather than as a fixed limit.

Try These Tools

Related Articles