Tone Analysis

Tone analyzer: what "tone" means here

"Tone" can mean two different things. In writing and marketing it is the register of a text: formal, businesslike, academic, conversational. In review mining and social listening it means sentiment: positive, negative or neutral. This tool does the first job. It measures how formal and how subjective a text is and lists the words behind the verdict. It does not sort reviews into happy and unhappy — brand-mention monitoring needs a sentiment analysis service.

Analysis is dictionary-based, with no neural network involved, so results are explainable and identical on every run. Paste text (10 to 50,000 characters) or give a page URL and the tool extracts the main text for you. Pick the language the text is written in: markers are looked up in that language's dictionary.

How to read the tone check results

Formality, 0–100%

The score is driven by markers. Formal connectors and bureaucratic phrasing ("consequently", "in accordance with", "pursuant", "utilize", "facilitate") and academic vocabulary ("methodology", "empirical", "it should be noted") push it up. Informal words ("gonna", "yeah", "stuff", "basically", "awesome") pull it down. Legalese such as "herein" or "aforementioned" weighs three times as much as an ordinary marker. Average word length adds or removes up to 10 points, and contractions ("don't", "we're") above 2% of words cost another 15.

When a text has few markers, formality comes out low even if the writing is clean and neutral. A low score means "few signs of formal register", not "full of slang".

Subjectivity, 0–100%

This counts hedges and evaluative words: "I think", "in my opinion", "probably", "unfortunately", "obviously", "amazing", "terrible". The bar is low — one such phrase per 50 words already reads as 100%. Opinionated blog posts land anywhere between 20 and 60%; terms of service or technical documentation should sit close to zero.

Primary tone

Detected markers

The words found in each category are the most useful part of the report. Strip formal markers to make a text plainer; remove evaluative words to make it more objective; cut hype words to make it sound less like an ad. Markers are matched as substrings, so short ones like "ok" or "like" can also be counted inside "book" or "likely". If the list looks odd, that is why.

Which tone fits which text

Tone is not a ranking factor in itself. It affects whether people read to the end and trust what they read. Two problems show up most in commercial copy: needlessly formal phrasing ("we facilitate the implementation of") that scores high on formality and low on clarity, and hype without evidence ("the ultimate", "game-changing") that inflates both the marketing and the subjectivity scores.

Spotting stiff phrasing and what to use instead

Take this sentence: "Prior to deployment, the team will utilize the framework in accordance with the policy." The checker flags "prior to", "utilize" and "in accordance with" as formal markers and "framework" as academic vocabulary. None of them is wrong, but together they make a simple instruction sound like a contract. Plain-English alternatives:

Rewritten: "Before launch, the team follows the policy and uses the framework." Formality drops, meaning stays. In legal copy fixed formulas can be justified because they act as terms of art; elsewhere, cut whatever doesn't add precision. The same logic applies to hype words on the marketing scale: replace "revolutionary" and "seamless" with what the product actually does.

The tone profile of AI-written drafts

Drafts from language models have a recognisable profile: formality higher than you'd expect for a blog, and subjectivity near zero. Connectors like "furthermore", "moreover" and "consequently" sit in the formal-marker dictionary and push formality up, while the model's habit of avoiding any personal stance flattens subjectivity. If a blog post comes back "Professional" with 0% subjectivity, add the author's view, experience and evidence-backed opinions. For removing stock transitions in bulk, run the humanizer and then re-check the tone.

Tone adjustment: what it can and can't do

"Analysis + Adjustment" nudges a text toward a more formal target (Formal, Professional) or a more casual one (Casual, Friendly) by swapping words from a dictionary: "get → obtain", "buy → purchase", "need → require" one way, and "utilize → use", "facilitate → help with", "commence → start" the other. It changes no more than about 5% of words and never restructures sentences.

Treat the adjustment as a pointer, not a rewrite. For deeper reworking of templated or machine-written copy there is the text humanizer; for sentence length and word difficulty, use the readability checker.

A practical workflow

1. Run 2–3 reference texts that you and your audience like and note their formality and subjectivity — that is your benchmark.
2. Check the new draft or live page by URL and compare.
3. Work through the marker list: rewrite stiff phrasing, cut unsupported claims or back them with data.
4. Re-check; if you restructured paragraphs, run the coherence checker as well.