Message Pressure Check

Updated .

Paste a message you are about to send. This tool counts its words, sentences and questions, compares your use of “you” with “I” and “we”, and lists every match against a published set of 55 pressure phrases such as “act now”, “last chance” and “you have to”. Each match comes with a plain pattern for rewriting it. The result describes the wording and is never a score of how persuasive the message is.

This tool needs JavaScript. Work through it by hand: read the message against the listed phrases further down this page, count each match, and apply the rewrite pattern for its group. Count the words, sentences and question marks yourself. The method and a worked example below let you reach the same answer on paper.

Message Pressure Check

The box starts with an illustrative message written for this page. Replace it with your own.

The message

Read in your browser only. The text is never sent anywhere and is not stored.

What the tool found

Paste a message to see its counts

  • Words n/a
  • Sentences n/a
  • Average sentence length n/a
  • Longest sentence n/a
  • Questions asked n/a
  • “You” words n/a
  • “I” and “we” words n/a
  • Listed wording that offers a way to say no n/a

Listed pressure phrases found

Nothing to show yet.

Rewrite patterns for the groups found

Nothing to show yet.

About the thresholds. The 35-word long-sentence note and the phrase list below are fixed choices made for this tool. They are this tool’s own rules of thumb, not research findings and not benchmarks.

Everything you type stays in your browser. Nothing is sent anywhere, and nothing is stored, so reloading the page clears it.

How it works

The tool makes four plain passes over the text. Each is a count or a match, so you can repeat any of them by hand.

  1. Words, sentences and questions. A word is a run of letters or digits, so “risk-free” and “don’t” each count once. A sentence ends at a full stop, question mark, exclamation mark or line break. Abbreviations such as “e.g.” can split a sentence early, so read the sentence figures as approximate. Questions are the number of question marks.
  2. Sentence length. The average is words divided by sentences. The longest is the word count of the single longest sentence. If it passes 35 words the note says so, because a reader may need two passes.
  3. “You” against “I” and “we”. One side counts you, your, yours, yourself and their contractions. The other counts I, me, my, mine, myself, we, us, our, ours, ourselves and their contractions. The two numbers are shown side by side with no target ratio, because no research-backed ideal is reported.
  4. Phrase matches. Matching ignores capital letters and reads curly apostrophes as straight ones. A phrase has to appear as whole words, so “hurry” does not match inside a longer word. Every occurrence counts. Wording that offers a way to say no is counted from a second short list (“up to you”, “your call”, “feel free to say no”, “fine to say no”, “okay to say no”, “if you'd rather not”, “if it doesn't fit”, “if it does not fit”, “no worries if”, “if now isn't a good time”).

How the verdict is worded. It reports a count of listed phrases, for example “3 listed pressure phrases found” or “no listed pressure phrases found”. It never says a message is good, bad or persuasive.

The full phrase list (55 phrases in 7 groups)

The rewrite pattern for each group

Where the list comes from. The list is this tool’s own, written for personal and professional messages. It borrows the idea of urgency and scarcity wording from published work on interface design. Mathur and colleagues (2019) crawled about 11,000 shopping websites and catalogued countdown timers, limited-time messages and low-stock messages, and in their data some timers reset while the offer was still available. The Federal Trade Commission’s staff report (September 2022) names countdown timers on offers that are not actually time-limited and claims that an item is almost sold out when supply is ample. Both studied web interfaces. Neither tested this phrase list, and neither reports that any phrase in it changes how a reader responds. Effect of these phrases on readers: Not reported.

Limits. A listed phrase can be honest. “Last chance” is accurate when registration really closes tonight, and the pattern then is to say the date and the reason. A message with no listed phrase can still pressure through tone, timing or the relationship between you and the reader. The thresholds and the list are this tool’s own rules of thumb, not research findings or benchmarks.

Worked example

The box above starts with an illustrative message written for this page, not a real one. It has 51 words in 6 sentences, an average of 8.5 words per sentence and a longest sentence of 15 words. It asks 1 question. It has 3 “you” words and 1 “I” or “we” word (the “me” in “trust me”), and no listed wording that offers a way to say no.

It matches 8 listed pressure phrases: “last chance”, “spots are filling up” and “ends tonight” (urgency and scarcity wording), “you have to” (demand wording), “everyone knows” and “trust me” (assertions in place of evidence), “guaranteed” (a certainty claim) and “don’t miss out” (fear-of-missing-out wording).

Before (illustrative)

Hi Sam, this is your last chance to join the workshop. Spots are filling up and the price ends tonight, so you have to decide now.

After (illustrative)

Hi Sam, the workshop is on 14 November and registration closes on 7 November, when the venue count is fixed. The price rises that day because the venue deposit is due. If it does not fit your plans, tell me and I will send the next date.

The rewrite states the real deadline and its reason, names the cause of the price change, and leaves a way to say no. Run through the tool, it shows no listed pressure phrases and one piece of wording that offers a way to say no. The dates are invented for this example.

Frequently asked questions

Does my message leave my browser?

No. The counts and matches are computed in your browser. The text is not sent anywhere and is not stored, so reloading the page clears it.

Does a result of no listed phrases mean the message is fine?

It means none of the listed phrases appeared. A message can still pressure through tone, timing or the relationship between you and the reader, so use the result as one check among several, alongside the disclosure test.

What if a flagged phrase is true?

Then say why. If registration really closes tonight, give the date, the time and the reason, so the reader can check it. The pattern asks you to make the deadline something the reader can verify, not to delete true information.

Why these phrases?

The list is this tool’s own. It draws on the kinds of urgency, scarcity and false-claim wording described in the Federal Trade Commission staff report and in Mathur et al. (2019), extended to written messages. It has not been validated against reader responses, and the tool does not claim it has.

Is there an ideal number of you-words or sentence length?

No ideal is reported in the research this page cites. The tool shows the counts so you can notice your own pattern. The only fixed threshold is a note when the longest sentence passes 35 words, which is this tool’s own rule of thumb.

Can I add my own phrases?

No. The list is fixed so the same message gives the same result every time, and so the page can show the whole list. If you notice wording that pressures you when you receive it, treat that as a reason to rewrite it, listed or not.