We Pointed Our Word Counter at Our Own Blog
A word counter is usually described as a thing that tells you how many words you have written, which is like describing a thermometer as a thing with numbers on it. The count is the least useful output. To show what the rest of it is for, we ran every post on this blog — 22 of them, 24,003 words — through our own counter. The results were not flattering, which is what makes them worth publishing.
Your Crutch Words Are Not the Ones on the List
Advice about repetition almost always names the same handful of culprits — just, very, actually, however. We checked ours against that list and the list was mostly wrong about us. A sensible threshold for a non-technical word is roughly once per 300 words; past that, it is worth asking whether the word is doing work.
| Word | Uses | Per 300 | Verdict |
|---|---|---|---|
| every | 93 | 1.16 | Over the threshold |
| same | 81 | 1.01 | Over the threshold |
| actually | 51 | 0.64 | Worth watching |
| because | 50 | 0.63 | Fine — it earns its place |
| just | 22 | 0.27 | Not our problem |
| very | 4 | 0.05 | Not our problem |
| however | 0 | 0.00 | Never used it once |
That is the whole argument for measuring rather than reading advice. Technical terms are exempt — a piece about sudoku will say sudoku a great deal, and swapping in synonyms would make it worse. But “every” at 93 uses is a tic, and we did not know we had it. Turn on exclude common words in the frequency table, set a minimum length of four, and your own version of this list appears in a second.
Sentence Length Is the Number That Actually Bites
Across 1,084 sentences we average 22.1 words each. The usual recommendation is 15 to 20, so we are running long — and one post, our piece on how long golden hour lasts, averages 39.7 words per sentence, roughly double the guideline. Reading it back with that number in hand, the diagnosis is obvious: too many sentences carrying two ideas joined by a dash.
Being straight about a limitation: our counter does not compute a Flesch readability score. But average sentence length is one of only two terms in the Flesch formula — the other is syllables per word — so words divided by sentences already tells you most of why a piece reads as heavy. If that ratio is above 25, you do not need a score to know what to do. Find your three longest sentences and split each into two.
The counter also reports reading and speaking time — 225 and 130 words per minute respectively — which is the fastest way to sanity-check a conference talk or a script against the slot you have been given. We wrote up where those two rates come from and when they mislead separately, and there are seven repeatable edits that pull a long draft back down without gutting it.
Words, Characters, and Knowing Which One Binds
Word count answers “have I covered this?” Character count answers “will this fit?” Confusing them is how a meta description gets truncated mid-clause in the search results. If two tools have ever given you different totals for the same text, the culprit is usually hyphens, contractions or URLs — what actually counts as a word covers the rules.
Rather than making you pick a tool, the counter shows both and runs a live characters-left readout against the limits that actually bind: X posts at 280, bios and SMS at 160, meta titles around 60, meta descriptions around 155, Instagram captions at 2,200 and Google Ads headlines at 30. There is a fuller character limit cheat sheet covering every platform if you are writing a full content package.
For search work specifically, the keyword density checker reports which terms and phrases dominate a draft and at what rate — useful for confirming your primary term is present without the draft reading like it was written for a crawler. And if you are writing to a fixed brief, the essay word counter is set up around admissions and exam limits rather than blog targets.
About That Goal-Setting Statistic
Writing advice loves a number: you are supposedly some specific percentage more likely to hit a goal if you write it down. The figure gets attributed to research by Gail Matthews at Dominican University, and it is worth knowing that the quoted percentage varies between articles — we have seen the same study credited with 33% and 42% in the same piece of writing.
Read the study page itself and there is a pleasing irony waiting: it opens by dismantling the even more famous claim that 3% of a Yale graduating class wrote down their goals and later out-earned everyone tenfold. That study, Matthews notes, was never conducted. It is an urban myth. A page whose first move is to debunk a goal-setting statistic is a strange place to source one from.
Setting a word target before a session is still a good habit, and a counter is what makes it checkable rather than vague. It just does not need a percentage bolted onto it to be worth doing. Which is, in miniature, the argument of this entire piece: measure your own writing instead of taking someone’s word for it.
Mistakes Worth Skipping
Closing the tool once you see the number
The word total is the least informative thing on the screen. The frequency table, the unique-word count and the sentence count are where the editing decisions come from. If you only wanted the number, your writing app already had it.
Running the frequency check mid-draft
Counting while you are still generating shifts attention from the argument to the scoreboard. Set a target before you start, ignore the tool until you stop, then use everything it gives you during the edit. Frequency analysis is an editorial pass, not a writing companion.
Trusting a generic crutch-word list
Our own audit is the argument against this. The words most commonly named — just, very, however — barely register in our writing, while "every" and "same" run well over the threshold. A list someone else wrote describes their habits, not yours. Measuring is the only way to find your own.
Chasing a word count for SEO
Length is not a ranking factor. It correlates with thoroughness, and thoroughness is what actually helps, so use a counter to check you have covered the topic — not to pad toward a number someone published as a target.
Using word count where characters are the constraint
A meta description is capped in characters, not words, and so are ad headlines, SMS and social bios. Counting the wrong unit means finding out when the platform truncates you.
Common Questions
What does a word counter actually give you beyond the word count?
Ours reports characters with and without spaces, sentences, paragraphs, unique words, estimated reading time and speaking time, plus a full word-frequency table you can sort, filter by minimum length or starting letters, restrict by case, and strip common stop words from. You can paste text or upload a PDF or Word file, and it all runs in your browser — nothing you paste is uploaded anywhere. The frequency table is the part most people never open and the part that actually changes your writing.
How many times can I use a word before it becomes a problem?
A reasonable working threshold for a non-technical word is about once per 300 words — past that, question it. Applied to our own blog, that flagged “every” at 1.16 per 300 and “same” at 1.01, while “just” sat at 0.27 and “very” at 0.05. Technical terms are exempt: an article about sudoku is going to say sudoku a lot, and replacing those with synonyms would make it worse, not better. The threshold is a prompt to look, not a rule to obey.
Does the tool calculate a Flesch readability score?
No, and it is worth being straight about that rather than implying otherwise. What it does give you is the sentence count alongside the word count, and average sentence length is one of only two inputs to the Flesch formula — the other being syllables per word. So dividing words by sentences gets you most of the way to understanding why a piece reads as heavy. Our own blog averages 22.1 words per sentence, which is above the 15–20 range usually recommended, and that is actionable without a score attached to it.
What is the difference between a word counter and a character counter?
They answer different questions. Word count is about depth and coverage — essays, articles, reports, anything with an editorial or academic minimum. Character count is about fitting a fixed space: X posts at 280, SMS and bios at 160, meta titles around 60, meta descriptions around 155, Google Ads headlines at 30. Our counter shows both at once and includes a live characters-left readout against those platform limits, so you do not need to decide which tool to open.
Is it true that writing down your goals makes you far more likely to achieve them?
The honest answer is that the specific percentages circulating for this are not reliable. The Dominican University research by Gail Matthews on written goals, commitment and accountability is real, but the widely-quoted figures attached to it vary from article to article, and the study page itself opens by debunking an even more famous goal-setting statistic as an urban myth. Setting a word target before a session is still a sound working habit — just do not expect a specific percentage to be riding on it.
Words, characters, sentences, reading time and a sortable frequency table with stop words stripped out. Paste text or drop in a PDF — free, no sign-up, and nothing you paste leaves your browser.
Open the word counter →Curious how far this kind of counting goes? We once counted every letter in all 172,835 words of the ENABLE lexicon.