The brief said 800 words. The draft feels like 400. WordShift counts words, characters, sentences, and a reading-time guess in the browser. Use it before you submit, before you tweet a thread that will be sliced, before you paste into a CMS that lies about counts. Do not use it as a style editor, a plagiarism checker, or a Chicago-style sentence authority. Reading time is a heuristic (often ~200 wpm). Your audience may be slower. Hyphenated compounds and CJK text still start arguments. For case changes, CaseShift. For filler paragraphs, LoremShift.
Paste the text. See words, characters with and without spaces, sentences, reading time. That is the job. CMS counts often exclude short words or markdown syntax. If your number disagrees with Google Docs, check whether you pasted markdown, footnotes, or a table. Count the same string both places before you accuse either tool.
When not to use it: SEO keyword density rituals, and grading student essays for originality. Also skip code; camelCase tokens will inflate 'words' in a way that does not help. For diffs of two drafts, DiffShift. For a hard character cap (SMS, meta description), trust characters, not words. Code and camelCase tokens inflate 'words' in a way that does not help a prose brief.
Privacy is local. Unpublished drafts stay in the tab. That is the reason not to use a web counter that stores pastes. Close the tab. The count is not stored as a document. CJK and hyphenated compounds still start arguments; know which CMS rules you were given. Count the same plain paragraph in Docs and here before you accuse either tool of conspiracy.
A guide draft pastes in at 1,240 words, 7,400 characters with spaces, 52 sentences, ~6 minutes at 200 wpm. The CMS says 1,198 because it stripped the markdown headings you included. You accept 1,240 as the editorial count. A meta description is 168 characters — over a 155 cap — you trim in MetaShift's mental model, not here. The draft never uploaded to a counter. The CMS stripping markdown headings explains 1,198 versus 1,240, not a secret algorithm.
WordShift counts in the browser. Drafts never upload. That is the point for unpublished writing. Close the tab when you are done; nothing is filed. Unpublished drafts stay in the tab; that is the reason not to use a counter that stores pastes.
Full policy: Shift Privacy Policy. Questions: [email protected].
Footnotes, text boxes, and whether you pasted markdown. Count a plain paragraph in both. If those match, the missing words are in elements you did not paste. Footnotes and text boxes you did not paste are the usual missing words when Docs disagrees.
A common adult silent-reading guess. Technical docs read slower. Do not print '6 min read' as a laboratory measurement. Technical docs read slower than a 200 wpm guess; do not print '6 min read' as a laboratory number.
SMS and some SEO tools count differently. Know which cap you were given. WordShift shows both so you can stop guessing. SMS and some SEO tools disagree about spaces; WordShift shows both so you can stop guessing the cap.
Paste only the body. A webpage counter that scrapes your URL is a different, worse product. Paste only the article body; counting a rendered webpage with nav is how you inflate a brief.
Related tools: CaseShift · LoremShift · MarkdownShift
Guides · All Shift tools · About · Contact