A client asks "how much will this brief cost?" — and in many firms and translation practices, the answer still depends on eyeballing a document and estimating its length. It's a simple calculation in theory, but one that generates real disagreement in practice: page count is not a reliable proxy for the amount of actual content, and estimating it wrong means billing it wrong.
This guide explains why page count is a misleading unit for pricing written work, how word count and character count solve the problem, and how to automate that count instead of estimating it by eye.
Why page count is a bad billing unit
The most common mistake is treating "page" and "amount of content" as the same thing. A document can span 3 physical pages and contain roughly 1,500 words of dense text (small font, narrow margins, single spacing) — or span 5 physical pages and contain only 600 words (generous white space, large headings, tables, images). Two documents with the same page count can represent wildly different amounts of actual drafting or translation work.
That's precisely why the professional translation industry settled on a different unit decades ago: the word. Rates are quoted per source word, not per page, because word count is a direct, unambiguous measure of the content itself, independent of font, margins, or line spacing.
| Context | Common billing basis |
|---|---|
| Professional translation (industry standard) | Per source word |
| Legal drafting and paralegal services | Per word, or an informal "standard page" of ~250 words, depending on the firm |
| Copyediting and proofreading | Per word, or per 1,000 words |
| Localization into character-based languages (e.g., Japanese, Chinese) | Per character, since "word" isn't a well-defined unit in those languages |
| Court or agency filing limits (word or page caps) | Varies by jurisdiction and court rule — always check the specific local rules |
The practical implication: if your fee agreement says "$X per page" without defining what counts as a page — font size, margins, line spacing — there's an ambiguity that can generate billing disputes with the client. The safer approach is for the fee agreement to define the price explicitly in terms of word count (with character count as a secondary reference), and for the final count to be produced the same, verifiable way every time.
Why counting by PDF or Word page number is imprecise
Word shows "X pages," and a PDF has a fixed page count. But that number doesn't reflect the actual density of text, because of:
- Font and spacing: a document in 14pt font with double spacing has far less text per page than one in 11pt single spacing — even with identical content;
- Margins: generous margins reduce the usable text area per page;
- Non-text elements: headers, footers, tables, letterhead and signature blocks take up page space without contributing proportionally to the word count;
- Scanned PDFs: a PDF produced by a scanner has no selectable text layer — counting pages says nothing about the actual amount of content, and extracting text requires OCR first.
That's why billing "by the page" tends to be imprecise and open to dispute — billing by word count, with the count done in an auditable way, is the technically sound method.
How to count words and characters quickly
RoseLab's word and character counter accepts three input formats:
- A Word .docx file — the count reads the text directly from the document;
- A PDF file — extracts the selectable text from the PDF (doesn't work on scanned PDFs without OCR);
- Pasted text — useful for checking a specific excerpt or a text that hasn't been saved as a file yet.
Step by step:
- Open the word and character counter;
- Upload the Word/PDF file, or paste the text into the box;
- Choose whether you want the result by word count (the default and primary metric) or by a specific character-count basis, if your fee agreement uses one;
- The result shows total word count, character count with and without spaces, and — if you selected a character-count basis — how many billing units that represents.
The entire count happens in the browser — the document is never sent to a server, which matters when the text is a draft contract or a filing that hasn't been submitted yet.
Applying word count to a fee calculation
Once a per-word (or per-character) rate is defined in the fee agreement, calculating the fee for a piece of written work becomes direct:
Fee = word count × rate per word
Example: a legal brief with 2,800 words, at a rate of $0.12 per word → 2,800 × $0.12 = $336.00.
For markets that bill by character instead (common in translation and localization work involving character-based languages, or when a contract explicitly defines a fixed character-count unit):
Fee = (character count without spaces ÷ unit size) × rate per unit
Having this number generated by a neutral tool — instead of estimated by eye — reduces the chance of disagreement with the client, and in the event of a fee dispute, the count can be reproduced and checked by any party, since the method is repeatable.
Other common uses of word and character counting
- Certified and professional translation: translators typically quote and bill per source word — the industry-standard basis worldwide;
- Copyediting and proofreading: editors often use word count (or a per-1,000-word rate) as the budgeting unit for revision work;
- Competitions, grant applications, and academic submissions with word limits for essays, personal statements, and abstracts — counting before submitting avoids disqualification for exceeding the limit;
- Estimating reading or drafting time on projects with many documents.
Frequently asked questions
Is a page a reliable way to bill written work? Not on its own. A page is a unit of layout, not of content — the same text can span different numbers of pages depending on formatting, while the word count changes only if the text itself changes. Word count is a more objective proxy for the actual amount of work.
What billing basis is standard in professional translation? Per source word is the near-universal industry standard for translation. Some markets and language pairs (especially those without clear word boundaries, like Japanese or Chinese) bill by character instead.
Does the tool count words in a scanned PDF? Not directly — a scanned PDF is an image, with no text layer. You need to run OCR (optical character recognition) first to extract the text before it can be counted.
Can I change the counting basis after already running the count? Yes — the result recalculates instantly when you switch between word count and character count, without needing to reload the document.
Does the count include footnotes and headers? It depends on how those elements are structured in the original file — the tool extracts text from the document body as read by the Word/PDF parsing library. For cases that require absolute precision in documents with many footnotes and special elements, do a visual check of the result.
Full workflow: from draft to final file
Counting words tends to be one step in a larger legal document preparation workflow. After calculating the fee based on the count:
- Clean the Word file of metadata and review comments before sending the draft to the client;
- Convert it to PDF or finalize the document for filing;
- Record the hash of the final file delivered to the client, creating a verifiable record of the exact version that was billed and delivered.
Ready to put it into practice?
Free, no sign-up — and your files never leave your computer.
Count words and characters now — free