About Word Count Kit

Why it exists

Counting words is a solved problem, and yet the tools that do it tend to be covered in advertising, ask you to paste an unpublished draft into someone else's server, and answer only one of the questions you actually have.

Word Count Kit handles the whole set of questions — length, structure, reading time, speaking time, distance from a target — and handles them without your text ever leaving the page.

The second half of that sentence is the part that took the work. Doing every calculation in the browser means the counting rules, the grapheme handling, the frequency analysis and the timing arithmetic all had to be written out rather than delegated to a service, and it means a very large document has to stay responsive on a mid-range phone rather than on a server. Both of those are solvable; neither is free, which is why most tools in this category do not bother.

What it will not do

It will not generate text, rewrite your sentences, expand a paragraph to reach a word count, or complete an assignment. This is a fixed product boundary rather than a feature that has not been built yet.

The reason is simple: a tool that measures your work is safe to have open next to your work. A tool that writes it is a different thing with different consequences, and mixing the two would make the first one unusable in the places it is most needed.

How the counting rules were chosen

Every rule is documented on the tool page that uses it, and every rule has an automated test behind it. Where a genuine ambiguity exists — whether a hyphenated compound is one word or two, whether an emoji is one character or four — the choice is stated openly rather than left for you to reverse-engineer from a number that does not match your word processor.

The core rule is deliberately plain: a word is a run of characters with no spaces in it that contains at least one letter or digit. That definition survives contact with the awkward cases, which is the only real test of a counting rule. An email address counts once. A URL counts once. A date written 12/03 counts once. A dash standing alone between two spaces counts as nothing at all, because it contains no letter and no digit.

Characters are counted as grapheme clusters — what a reader would call a character — rather than as the units a program stores internally. A flag emoji, an accented letter typed as two code points, and a family emoji made of five joined pieces each count as one. That is what makes the number match the one a platform with a character limit will show you, and it is the single most common reason two character counters disagree.

None of these choices is the objectively correct one, because there is no objectively correct one. They are defensible, consistent, tested and written down, which is the most any counter can honestly offer.

Where the estimates come from

Reading time uses 200 words per minute as the default for ordinary silent reading of ordinary prose, with slower and faster presets on either side and a free field for your own rate. Speaking time defaults to 130 words per minute, which is close to the pace of an unhurried presentation; conversational speech runs faster and read-aloud formal delivery runs slower.

Page counts assume a 12-point serif face, one-inch margins and no headings or figures: roughly 250 words to a double-spaced page, 333 at 1.5 spacing and 500 single-spaced. Every one of those assumptions is stated on the page that uses it, and every one of them can be replaced with a figure you measured from your own document.

These are averages, and the honest thing to say about an average is what it does not describe. It does not describe a reader working through unfamiliar technical material, a nervous speaker at double their rehearsed pace, or a department with its own formatting rules. The figures are planning tools. Treat them as the middle of a range and you will not be caught out.

How it is paid for

The free version may carry advertising in three fixed positions: below the tool, between sections of explanatory text, and near the footer. Never in, above or over the editor. A paid version may exist later; if it does, it will add convenience features and remove advertising, and it will not take away anything the free version does today.

That placement rule is enforced in the code rather than in a style guide. The component that renders the editor has no way to render an advertisement, so an ad cannot drift into the writing surface through a later change. Advertising also cannot read what you type: it runs in its own frames with no access to the editor, and the text is never written anywhere an external script could reach it.

Analytics, where used, records which tool was opened and which features were used. It never records the text, never records a sample of it, and never records its content in any form — only, at most, a coarse size band so that performance problems on very large documents can be found.

Who publishes this site

Word Count Kit is published by Word Wise, an independent one-person project rather than a company with a marketing department. There is no venture funding behind it, no data business attached to it, and no plan that depends on collecting anything from the people who use it.

That is worth stating plainly because it explains the shape of the product. A site that made its money from the text passing through it would have built this very differently, and would have had a reason to keep your draft. This one has no such reason, so it does not keep it.

Corrections, bug reports and complaints all go to the same inbox and are read by the person who wrote the code. The contact page explains what makes a report useful.