Headline analyzer
What is Word Balance in a Headline?
Word balance is a 0–100 sub-score equal to the mean of 4 vocabulary bucket scores: common words (target 20–30% of tokens), uncommon words (10–20%), emotional words (10–15%), and power words (target 1–3 count). Each bucket scores 100 when inside its target band and decays when outside it; the 4 scores are averaged to produce the final word-balance score.
Seeklyn shows each word-balance bucket with its percentage and a colour-coded chip for every token — visible in the Word Balance panel at /dashboard/headline.
What does word balance measure?
Word balance measures how well a headline mixes 4 vocabulary categories — common, uncommon, emotional, and power words — against research-backed target ranges. A headline with all common words reads as bland; one crammed with power words reads as clickbait. The 0–100 score rewards headlines that hit all 4 ranges simultaneously.
Seeklyn labels every token in the headline with its category (power > emotional > common > uncommon, in priority order) so writers can see exactly which bucket each word fills. The category chips match the bucket percentages precisely — no word is counted twice.
What are the 4 vocabulary buckets?
4 buckets cover every headline token. Common words are high-frequency English words that keep the headline readable; the target is 20–30% of tokens. Uncommon words add surprise and distinctiveness; the target is 10–20%. Emotional words carry a strong valence score (LabMT hedonometer rating below 4.0 or above 6.0); the target is 10–15%. Power words are persuasive trigger words (e.g. "proven", "ultimate", "best"); the target is 1–3 in absolute count, not percentage.
| Bucket | Target | Scoring method |
|---|---|---|
| Common | 20–30% of tokens | Score 100 in band; −4 pts per percentage point outside |
| Uncommon | 10–20% of tokens | Score 100 in band; −4 pts per percentage point outside |
| Emotional | 10–15% of tokens | Score 100 in band; −4 pts per percentage point outside |
| Power | 1–3 words (count) | 100 for 1–3; 0 for none; 30 for >3 (clickbait) |
How is the word-balance score calculated?
The word-balance score is the arithmetic mean of the 4 bucket scores, rounded to the nearest integer. Ratio buckets (common, uncommon, emotional) score 100 in band and lose 4 points per percentage point outside, floored at 0. The power bucket scores 100 for 1–3 words, 0 for none, and 30 for more than 3 — a partial score that penalises overuse without zeroing the whole dimension.
Why does word balance carry the highest CTR weight?
Word balance carries 34% of the CTR score — the largest single weight — because vocabulary mix is the broadest predictor of headline engagement. A headline that scores well on type and length but uses only flat, common words still reads as dull. By measuring all 4 buckets at once, word balance captures the richness of the writing in a way that single-dimension checks (length alone, sentiment alone) cannot.
Frequently asked questions
What happens if a headline has 0 power words?
The power bucket scores 0, which pulls the word-balance mean down. Adding 1 power word flips the bucket score from 0 to 100, raising the word-balance score by up to 25 points (one quarter of the mean).
What counts as an emotional word?
Seeklyn uses the LabMT hedonometer valence list. Any word with a valence below 4.0 (unhappy end) or above 6.0 (happy end) qualifies as emotional. Words in the neutral band (4.0–6.0) do not count toward the emotional bucket.
Can a word be in more than 1 bucket?
No. Seeklyn assigns each token to exactly 1 bucket using the priority order: power first, then emotional, then common, then uncommon. A word that qualifies as both power and emotional is counted only as power.