What Does a Headline Score Actually Measure?

A headline score measures the shape of a headline, not whether it is any good. The Headline Analyzer on this site adds up five sub-scores, each worth up to 20 points: how many words the headline has, how many characters, how many words from a fixed list of "power" words, how many from a fixed list of emotional words, and the balance between common and uncommon words. The total, out of 100, is fully reproducible: the same headline always gets the same score. What it cannot measure is whether the headline is true, whether it matches the page, or whether your readers will care.

This guide takes the five parts one at a time, with real headlines run through the Analyzer, so you can read a score and know what moved it.

The five sub-scores

1. Word count (0 to 20)

Six to ten words earn the full 20. Four, five, eleven or twelve words earn 14. Three words, or thirteen to fourteen, earn 8. Anything else (one or two words, or fifteen and up) earns 3. Words are counted the way the tool splits them: runs of letters and digits, with apostrophes and hyphens kept inside a word, so "don't" and "step-by-step" are one word each.

2. Character length (0 to 20)

Forty to sixty characters earn 20. Thirty to thirty-nine, or sixty-one to seventy, earn 14. Twenty to twenty-nine, or seventy-one to eighty, earn 8. Anything else earns 3. The count includes spaces and punctuation, after trimming the ends.

3. Power words (0 to 20)

The tool has a fixed list of persuasion-style words: "how", "why", "you", "your", "guide", "quick", "proven", "free", "best", "today", "step", "tips" and about forty others. No matches scores 0; one scores 12; two score 16; three or more score 20. Every occurrence counts, so a headline that says "how" twice has two matches.

4. Emotional words (0 to 20)

A second fixed list holds feeling words: "surprising", "brilliant", "shocking", "inspiring", "irresistible" and about thirty-five more. It is scored the same way: 0, 12, 16 or 20. Most plain, useful headlines have none, which is one reason the scores below rarely pass 80.

5. Balance (0 to 20)

A third list holds common function words: "the", "a", "and", "to", "is", "your", "it", "not" and so on. The tool divides the number of common words by the total. A share of 15% to 40% earns 20; 10% to 50% earns 12; anything else earns 5. The idea is that a headline with no common words reads like a keyword list, and one made almost entirely of them says nothing specific.

The five add up to the score, which is then graded: 75 and over is "excellent", 60 to 74 "good", 40 to 59 "fair", under 40 "poor".

Eight headlines, scored

Headline	Words	Chars	Length	Chars score	Power	Emotion	Balance	Total
Repotting	1	9	3	3	0	0	5	11 poor
Thermostatic Radiator Valves Explained	4	38	14	14	0	0	5	33 poor
How to Fix a Dripping Kitchen Tap	7	33	20	14	12	0	20	66 good
Seven Signs Your Houseplant Needs a Bigger Pot	8	46	20	20	12	0	20	72 good
Why Your Sourdough Starter Is Not Rising and How to Fix It	12	58	14	20	20	0	12	66 good
Your Quick, Proven Guide to Fixing a Dripping Tap Today	10	55	20	20	20	0	20	80 excellent
Council Tax Bands in England: How Your Band Was Set and How to Challenge It	15	75	3	8	20	0	20	51 fair
It Is Not What You Think It Is	8	30	20	14	12	0	5	51 fair

Reading the results

Where the score is useful

"Repotting" scores 11 because a one-word headline tells a reader almost nothing, and the Analyzer is right to flag it. "Thermostatic Radiator Valves Explained" scores 33, mostly because it has no common words at all: the balance score drops to 5. Adding a few connecting words, as in "How Thermostatic Radiator Valves Work, and When to Fit One", makes it read like a sentence and takes it to 72, with balance up from 5 to 20.

The council tax headline shows the length checks earning their place. At 15 words and 75 characters it scores 3 and 8 on the first two parts. It is clear, but long. Trimming it to "How Your Council Tax Band Was Set, and How to Challenge It" keeps the meaning and scores 66: words and characters rise from 3 and 8 to 14 and 20, though balance slips from 20 to 12 because a larger share of the words are now common ones.

Where it is not

The highest score in the table, 80, goes to "Your Quick, Proven Guide to Fixing a Dripping Tap Today". It earns full power marks from five list words ("your", "quick", "proven", "guide", "today"). It is not a better headline than "How to Fix a Dripping Kitchen Tap" (66). It is vaguer about the tap, and "proven" is a claim the page would need to support. The extra 14 points come entirely from vocabulary the list happens to contain.

"It Is Not What You Think It Is" scores 51, the same as the clear council tax headline, despite saying nothing at all about its subject. The length and character parts reward its shape, and "you" is on the power list. The balance score (5, with 88% common words) is the only part that notices anything is wrong.

The lesson is simple: a low score usually points at a real problem (too short, too long, all keywords, no connecting words). A high score does not prove a good headline. The guide on when a higher score reads worse goes further into what happens when you write for the power-word list.

How to use it well

1. Write the headline for the reader first, saying plainly what the page gives them.
2. Run it through the Analyzer and look at the lowest sub-score, not the total.
3. Fix that part only if the fix makes the headline clearer. A low length score is worth acting on. A zero for emotional words usually is not; most informative headlines have none.
4. Never add a word just because it is on a list. "Proven", "ultimate", "best" and "shocking" are claims. If the page cannot back them up, they cost trust.
5. Check the width separately. The Analyzer counts characters, not pixels. For a title that will appear in search results, measure it in the Headline Pixel Counter, because capitals and wide letters change where it is cut.

What the score does not measure

• Truth. The tool cannot tell whether a headline's promise is kept by the page.
• Relevance. It does not know what the page is about, so it cannot tell whether the headline matches it.
• Audience. A headline that suits a trade audience may score lower than one written for a general reader, and be the right choice anyway.
• Clicks, shares or rankings. The score is a heuristic built from fixed word lists and fixed ranges. It is not trained on any click data and predicts nothing about performance. If you want to know how headlines perform for your audience, your own analytics are the only honest source.
• Meaning of the words. It matches spellings, not senses. "You" counts as a power word wherever it appears, while "gluten-free" does not count as "free" because the tool keeps hyphenated words together as one word.
• Width. Characters are counted, but the space they take is not; see the Pixel Counter for that.

Used this way, a headline score is a fast second opinion about shape: is it too short, too long, all keywords, or all filler? That is genuinely useful. Anything beyond that is your judgement, and should stay that way.