# Headline craft > Five times as many people read the headline as read the body copy. A headline does three jobs at once: it tells the reader what they're about to get, gives them a reason to care, and signals what kind Parent: [Writing and Documentation](https://llms-explorer.com/tree/writing-and-documentation/) · 13 facets · 44 facts · page: https://llms-explorer.com/tree/headline-craft/ ## Overview - Five times as many people read the headline as read the body copy. A headline does three jobs at once: it tells the reader what they're about to get, gives them a reason to care, and signals what kind of piece it is. — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#overview) ## 1. The five-times rule and the 80-cent budget - Ogilvy's empirical claim: five times more people read the headline than the body. The practical implication: the headline is where 80% of the work happens. Joanna Wiebe's discipline: never write fewer than 10 headline variants before picking one. Ogilvy demanded 16. — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#1-the-five-times-rule-and-the-80-cent-budget) ## 3. The "you-frame" rule - A headline pointed at the reader outperforms a headline pointed at the writer, the product, or an abstract topic. — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#3-the-you-frame-rule) - Topic frame (weak): "The future of distributed databases" — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#3-the-you-frame-rule) - You-frame (strong): "What your replication setup is costing you (and how to fix it)" — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#3-the-you-frame-rule) ## 4. Numbers in headlines: odd, specific, and not too round - BuzzSumo's analysis of 100M headlines: numerals increase CTR by ~36%. — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#4-numbers-in-headlines-odd-specific-and-not-too-round) - B2C top performers: 10, 5, 15, 7, 20, 6 — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#4-numbers-in-headlines-odd-specific-and-not-too-round) - Odd numbers outperform even numbers for curiosity-driven content. — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#4-numbers-in-headlines-odd-specific-and-not-too-round) - Specific numbers outperform round numbers. "$1,247 saved" beats "$1,200 saved." — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#4-numbers-in-headlines-odd-specific-and-not-too-round) ## 5. Curiosity-gap vs declarative headlines - Declarative headlines state the conclusion in the headline itself. Good for hard news, technical posts. — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#5-curiosity-gap-vs-declarative-headlines) - Curiosity-gap headlines withhold the conclusion to drive the click. The failure mode: clickbait - a gap so wide the body can't close it. — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#5-curiosity-gap-vs-declarative-headlines) ## 7. News vs feature vs op-ed vs SEO headline registers - News headline: Active verb, present tense. "Atlas adds vector search for Voyage embeddings." — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#7-news-vs-feature-vs-op-ed-vs-seo-headline-registers) - Feature headline: Label or question form OK. "The quiet revolution in how databases handle vectors." — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#7-news-vs-feature-vs-op-ed-vs-seo-headline-registers) - Op-ed headline: Argument-forward. "We're thinking about vector search wrong." — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#7-news-vs-feature-vs-op-ed-vs-seo-headline-registers) - SEO headline: Keyword-front-loaded. "Vector search benchmarks: 12 databases compared (2026)." — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#7-news-vs-feature-vs-op-ed-vs-seo-headline-registers) ## 8. The deck / subhead pairing - A deck does three things: — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#8-the-deck-subhead-pairing) - Expands what the headline could not fit. — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#8-the-deck-subhead-pairing) - Answers the question the headline raised. — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#8-the-deck-subhead-pairing) - Sizing: Heds are 6–10 words. Decks are 10–25 words. — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#8-the-deck-subhead-pairing) ## Anti-Patterns - The "we" headline. "We released a new feature today." The reader doesn't care what you did. — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#anti-patterns) - The label headline. "Distributed Systems Notes" names a folder, doesn't earn a click. — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#anti-patterns) - The pun headline. Ogilvy banned them in 1963 and the data still says he's right. — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#anti-patterns) - The bait-and-switch. Headline promises X, body delivers Y. — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#anti-patterns) - The under-specified curiosity gap. "What we learned" without a hint of what you learned. — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#anti-patterns) ## References - Ogilvy, David. Ogilvy on Advertising (1983). — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#references) - BuzzSumo headline studies. — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#references) - Wiebe, Joanna. Copyhackers headline formulas. — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#references) - PMC, "When curiosity gaps backfire" (2024) — [source](https://llms-explorer.com/sources/mdb-context-hub/headline-craft/#references) ## Where this helps - Deciding what to promise a reader before writing a piece — the headline sets the expectation for the whole piece, whether written first or rewritten last. — [source](https://llms-explorer.com/tree/headline-craft/) *(AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.)* - Choosing the right register for a specific publication context — news, feature, op-ed, or SEO — each of which follows a different structural convention. — [source](https://llms-explorer.com/tree/headline-craft/) *(AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.)* - Diagnosing why a piece isn't getting clicks or opens despite strong body content, since the headline is usually the first thing to audit before the content itself. — [source](https://llms-explorer.com/tree/headline-craft/) *(AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.)* - Writing email subject lines, landing-page headers, or social captions, where the same you-frame and number-specificity principles apply as to article headlines. — [source](https://llms-explorer.com/tree/headline-craft/) *(AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.)* ## How to apply this - Draft at least ten headline variants before selecting one, following Joanna Wiebe's discipline, rather than settling on the first phrasing that comes to mind. — [source](https://llms-explorer.com/tree/headline-craft/) *(AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.)* - Rewrite topic-frame headlines like "The future of X" into you-frame headlines like "What X is costing you" so the reader, not the writer or the abstract topic, is the subject of the sentence. — [source](https://llms-explorer.com/tree/headline-craft/) *(AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.)* - Use specific, odd numbers instead of round ones when a number belongs in the headline — "$1,247 saved" reads as more credible than "$1,200 saved." — [source](https://llms-explorer.com/tree/headline-craft/) *(AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.)* - Match the headline register to the piece: active-verb present-tense for news, label-or-question form for features, argument-forward for op-eds, keyword-front-loaded for SEO. — [source](https://llms-explorer.com/tree/headline-craft/) *(AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.)* ## Common mistakes - Writing the headline last and skipping the you-frame rewrite pass, so the piece ships with a topic-frame or "we" headline that nobody outside the team cares about. — [source](https://llms-explorer.com/tree/headline-craft/) *(AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.)* - Trusting instinct over volume — picking the first or second headline drafted instead of writing the ten variants Wiebe's process calls for. — [source](https://llms-explorer.com/tree/headline-craft/) *(AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.)* - Letting the deck restate the headline instead of doing its own job — expanding what the headline couldn't fit and answering the question the headline raised. — [source](https://llms-explorer.com/tree/headline-craft/) *(AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.)* - Sizing headlines and decks by feel instead of the working range — headlines that run well past six to ten words, or decks that don't stay in the ten-to-25-word band, tend to lose the reader partway through. — [source](https://llms-explorer.com/tree/headline-craft/) *(AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.)* ## Known issues - The five-times and 36%-CTR statistics come from specific studies (Ogilvy's advertising-era observation, BuzzSumo's 100M-headline analysis) — they describe averages across large corpora, not a guarantee for any single headline. — [source](https://llms-explorer.com/tree/headline-craft/) *(AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.)* - Curiosity-gap headlines are a double-edged tool: under-specify the gap and it reads as clickbait, over-specify it and the gap disappears along with the reason to click. — [source](https://llms-explorer.com/tree/headline-craft/) *(AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.)* - The you-frame rule and number-specificity findings are best documented for direct-response and marketing copy; the pack doesn't establish how strongly they transfer to registers like news or academic writing, where objectivity and technical precision constrain phrasing differently. — [source](https://llms-explorer.com/tree/headline-craft/) *(AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.)* - Register conventions shift with the publication and platform — a headline correctly built for SEO front-loading may read as awkward or robotic in a context that expects a feature or op-ed register. — [source](https://llms-explorer.com/tree/headline-craft/) *(AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.)* ## Context files - [Headline craft](https://llms-explorer.com/downloads/sources/mdb-context-hub/headline-craft.md)