Free tool 09 / Copy
See which of 6 checks is losing you the click, before you hit send.
Paste a headline, subject line or hook. Get six scores and the words costing you the click, in 10 seconds.
- Pinpoint the exact words earning or costing you points
- Compare your line against six weighted, explainable checks
- Identify your single highest-value fix, ranked first
Read this before you trust the number
A formula cannot tell you whether a hook is true, whether it fits this audience, or whether it will actually get clicked. This tool scores craft mechanics only. Every rule below is a heuristic drawn from conventional direct-response practice, not from measured performance data on your list — and a hook that deliberately breaks several of these rules can outperform one that satisfies all of them. Treat the score as a second opinion. Treat the annotated hook below it as the useful part. And treat the "ideal length" ranges for what they are — widely repeated conventions that vary by source and audience, not a finding anyone has measured against your readers.
The useful part
Your hook, annotated
Every word that earned or cost you points, marked up in place. This is what the score is made of — read this before you read the number.
The marked-up version of your hook appears here as you type.
- Vague word — dotted
- Jargon — wavy
- Cliché / clickbait / caps — struck through
- Tension, promise or specificity win — solid
- Clarity flag — dashed
Or load an example
Three genuinely strong hooks and three genuinely weak ones, so the score spread is visible.
Compare variants
This session's hooks
The last 8 hooks graded in this tab, newest first. Kept in memory only — reload the page and this list is gone. Click one to load it back into the grader.
Nothing graded yet this session.
The order to fix it in
Your three highest-value changes
Ranked by points recoverable, and named against the actual words in your hook — not generic advice.
Grade a hook and its top three fixes appear here, ranked.
Methodology
How this is scored
Six dimensions summing to 100. Every point is computed from the text you typed — nothing is hard-coded except the word lists and the channel target ranges, both published in full below.
-
1
Length fit checks the channel's target window (18 pts)
Each channel has a different conventional character range — 28–50 for an email subject, 20–40 for an ad headline, and so on. Inside the range scores full marks; outside it, the score tapers rather than falling off a cliff, because a line four characters over a soft convention is not a failure.
-
2
Specificity rewards facts and punishes filler (22 pts)
Digits, currency, percentages, concrete time markers and capitalised mid-sentence words each add points, because they are the parts of a line that could not be true of any other claim. A fixed list of vague words —
things,results,growth,powerfuland others — subtracts points, because they could describe almost anything. -
3
Tension and promise measure the pull, not just the polish (18 + 14 pts)
Tension rewards question form, contradiction and reversal words (
but,without,why), open loops and a named obstacle — the mechanics of a line that makes a reader want the next sentence. Promise rewards verbs of change and gain (cut,save,stop,get) that tell the reader something happens to them, not just that a topic exists. -
4
Clarity penalises anything that costs a reader effort (18 pts)
Deductions apply for average word length above roughly six characters, individual words over twelve characters, a fixed jargon list (
synergy,leverage,ecosystem,actionableand others), a simple passive-voice pattern, and more than two commas or semicolons strung together. -
5
Craft penalties deduct for the tells experienced readers spot instantly (10 pts, starts full)
ALL CAPS words longer than a four-letter acronym, more than one exclamation mark, more than one emoji, clickbait constructions ("you won't believe…"), and a fixed cliché list —
game-changer,unlock,10x,seamless,the ultimate guideand others — each deduct from a starting 10.
What this model is built on
- Tension, promise and specificity as craft moves — Robert Bly's The Copywriter's Handbook and Eugene Schwartz's Breakthrough Advertising are the direct-response references this framing is drawn from: a line that names a stated obstacle, a mechanism, and a specific outcome. Neither book is a controlled study of what gets clicked — they are widely used craft doctrine, and this tool scores adherence to that doctrine, not proven performance.
- Jargon and passive-voice penalties — the plain-language guidance published at plainlanguage.gov (US General Services Administration) is the basis for the clarity checks: prefer short, active, concrete sentences over filler and hedging.
- Channel length ranges — consolidated from the ranges commonly repeated by email and ad platforms (subject-line display truncation on desktop and mobile clients, ad headline character limits). These are conventions, not measurements — see the honesty note above the tool.
Nothing here has been validated against open rates, click-through rates or conversion data for any specific list or audience — this tool does not have access to that data, and no free tool honestly could. It also does not attempt sentiment analysis; the "sentiment-free" line in the stats strip is a deliberate admission, not a computed score. Every list of words used for scoring — vague, jargon, cliché, clickbait, tension, promise — is published in full in this page's source, so you or a model reading it can check exactly what triggered a deduction.
Questions people ask about writing a headline
What makes a good hook?
A good hook does three things at once: it is specific (a number, a name, a timeframe — something only true of this claim), it creates tension (a question, a contradiction, an obstacle the reader recognises), and it promises something happens to the reader, not just a topic. Vague, safe or purely descriptive lines fail one or more of those tests, however polished the sentence sounds.
How long should a headline be?
There is no single correct length — the widely repeated ranges, roughly 28–50 characters for an email subject or 20–40 for an ad headline, are conventions from practitioners and email-client display limits, not findings from controlled tests. They exist because truncation and scanning behaviour differ by channel. Treat them as a target window to hit, not a rule that overrides a genuinely sharp line that runs a little long or short.
Do power words actually work?
Words like "amazing", "powerful" or "results" feel persuasive to write but carry no information — they could describe almost any claim, which is exactly why readers learn to skip past them. The words that actually work are concrete ones: a digit, a named obstacle, a stated timeframe. "Power words" is really a euphemism for vague words dressed up with enthusiasm, and most readers can tell the difference even if they cannot name it.
Why do my headlines sound generic?
Usually because they describe the topic instead of creating tension around it. "Five Tips for Better Marketing" states a subject; it promises nothing and resolves nothing. Generic headlines are also often built from safe, reusable words — solutions, growth, quality — that fit every business and therefore signal none. Swap the topic word for the specific outcome, obstacle or number that is actually true of this piece.
Should a headline be a question?
Sometimes, not always. A question earns attention by opening a loop the reader wants closed, which is why it scores well on tension in this tool. But a question anyone would answer "no" to, or one whose answer is obvious, wastes the slot. It works best when the honest answer is genuinely uncertain or mildly uncomfortable — not as a reflex format applied to every line you write.
Common follow-ups
Does anything I type here get sent anywhere?
No. The grader is a single JavaScript file running in your browser. There is no server call, no analytics event carrying your hook, no account, and no storage beyond the current tab. If you copy the result link, your hook is encoded in the URL — treat that link the way you would treat the line itself.
Why did a hook I actually like score low?
Because the score only measures the six mechanics on this page — it has no idea whether your claim is true, whether it fits your audience's inside jokes, or whether it will land. A hook can deliberately break the rules and still be the right hook. Read the annotation, decide whether each flagged word is actually a problem for your reader, and override the score when your judgement disagrees with it.
Can I just stuff in numbers and the word "why" to game the score?
A little, yes — that is a known limit of any rule-based scorer, and this one says so rather than pretending otherwise. But stuffing has a ceiling: extra clauses cost clarity points, repeated punctuation and hype words cost craft points, and a line that reads like it was optimised for a formula reads that way to a human too. The annotation makes stuffing visible, which is the actual defence.
Why does the same word sometimes count for and against me?
A few words genuinely do double duty. "Stop" is both a tension marker (a reversal) and a promise verb (something the reader gets to do), and both are scored — that is not a bug, it is the same word doing two jobs in the sentence. The annotation always shows which single reading was used for the highlight, even when the underlying score credited more than one dimension.
Next
A sharp hook still needs the right question underneath it.
The best headline in the world cannot rescue a page that answers a question nobody is asking. The Question Map finds the ones your buyers actually type.