McClatchy Headline Style Guide

Apple’s editors appear to favor “What to know” when selecting stories for the Featured slot — possibly because it signals a comprehensive briefing article. But readers who encounter a non-featured “What to know” article in a regular feed may find the format less compelling. The formula works as a targeting signal for editorial curation, not as an organic traffic driver.

“Here’s what we know” and “Here’s what happened” are explanatory frames that Apple editors consistently select for breaking and developing stories. The formula signals authoritative, organized coverage — something editors want foregrounded. The effect is topic-conditional: it works for Crime and Business because those are topics where Apple News featuring is achievable. Sports content earns 0% featuring regardless of formula used.

Apple News’s editorial curation prioritizes local news and civic content over sports. The sports finding is not formula-dependent — no headline formula unlocks featuring for sports content. The platform’s own section taxonomy treats Sports as a distinct feed that rarely enters the main Featured rotation. Wire/national content (Nation & World) similarly underperforms because Apple News favors local publishers writing about their own markets.

Apple News displays article previews with a truncated headline in the feed. Too-short headlines don’t convey enough context to earn a click. Too-long headlines get cut off visually and may signal low-quality writing to the algorithm. The 90–120 char range is where T1 McClatchy articles consistently perform above median. This is a different optimal range than SmartNews (70–90 chars) — the two platforms have different display formats and different algorithm signals.

The question format creates a tension: Apple’s human editors appear to like questions as featured stories (perhaps because a clear question implies a satisfying answer). But the Apple News algorithm does not reward question format for non-featured articles — organic distribution of question headlines underperforms the baseline. This is one of the clearest Featured vs. organic divergences in the dataset. If the goal is organic views, avoid questions. If the goal is specifically to pursue a Featured slot, questions are a viable targeting tool. When using question format, prefer “How” and “Why” over “Will/Can” (speculative) and “Who” (identification) — explanatory question framing signals comprehensive coverage and is favored by Apple editors over speculative or identification framings.

Quote ledes overall outperform the baseline for Apple News featuring. But within that group, who is being quoted matters. Institutional/authority voices (law enforcement, officials, elected leaders) signal breaking news and civic importance — two things Apple’s editors consistently reward with Featured placement. Expert/scientific quotes signal novelty and credibility. First-person subject quotes (“I was terrified”) read as feature writing, which Apple editors select less often for the Featured slot. This is a directional finding — subtypes have small n and p-values are uncorrected for multiple comparisons. Treat as strong guidance, not a confirmed rule.

This is a time-series observation across all T1 Apple News articles. The mechanism isn’t confirmed, but the pattern is consistent: as the Apple News environment matures, list-structured and numeric headlines are gaining traction while question format loses it. This directional shift suggests leaning into number leads (“5 things…”, “3 reasons…”) and pulling back on question-format articles that aren’t specifically targeting a Featured slot.

General/Discovery content (nature, wildlife, science) has the highest individual-article ceiling of any content type in the McClatchy T1 dataset — a single snake species article reached 53K views. The common thread across top performers in this vertical is a discovery frame: something new was found, something rare was seen, something never documented before was observed. The headline should foreground the novelty (“Never-before-seen”, “Rare footage”, “Scientists discover”) rather than burying it in the middle of the headline. This framing also works for SmartNews — apply it consistently across both platforms for this content type.

Apple News uses section metadata to route articles to topic-specific feeds. “Main” is not a real section — it’s the default fallback when no section is assigned. Articles with only a Main tag are treated as unclassified and deprioritized. This is a one-time configuration fix per article, not a headline decision — but it has a larger impact on distribution than any headline formula.

The content produced by Sara Vallone’s trendhunter team (Mind-Body, Everyday Living, Experience) has not received any Apple News Featured placements in the Jan–Feb 2026 data window (vs. a 1.2% baseline across all ANP articles). This may reflect the content type, the topics, or the section tagging — the root cause is not yet established. Until the cause is identified, do not build strategy around Featured placement for this content. Focus on organic Apple News views and SmartNews distribution, where formula choices have a clearer effect.

SmartNews uses an algorithmic feed with no human editorial curation layer. The algorithm ranks articles by predicted engagement and completion rate. Question-format headlines perform significantly below baseline across all topics and publication types — this is one of the most statistically robust findings in the analysis (p=3.4e-6 on a dataset of 918 articles). The penalty is consistent enough that it should be treated as a hard rule, not a soft guideline. This is the opposite of the Apple News Featured dynamic, where editors occasionally favor questions.

“What to know” combines two weak signals for SmartNews: it reads as a summary/briefing format (lower engagement prediction) and it doesn’t front-load a specific concrete claim or fact. SmartNews rewards headlines that clearly signal what the reader will learn and why it’s worth clicking. “What to know” signals “general overview” rather than “specific revelation.” The effect is statistically comparable to the question penalty — both are hard avoidance rules for SmartNews.

SmartNews displays article headlines in a more compact format than Apple News. The shorter optimal length (70–90 vs. 90–120 for Apple News) reflects differences in how each platform renders article previews. A headline that feels right-sized for Apple News may feel bloated in a SmartNews feed. When writing one headline for both platforms, prioritize the Apple News length and trim for SmartNews — or write two versions (see Platform Headline Pairs in the Editorial Playbooks).

SmartNews rewards clarity and specificity. A direct declarative headline tells the reader exactly what happened without requiring them to complete a mental question or fill in a gap. This format also avoids the question and WTK penalties. It’s not a growth formula (it doesn’t significantly lift above baseline) but it’s the safest option when you don’t have a strong formula signal for a particular story.

Most headline formulas create a platform trade-off: what works for Apple News hurts SmartNews, and vice versa. “Here’s what to know” and “What to know” are dramatic examples of this tension. “Here’s” (“Here’s what happened,” “Here’s what this means”) is the exception: it performs above baseline on SmartNews (directionally) while also being the strongest confirmed formula for Apple News Crime/Business featuring. When operational constraints mean publishing one headline to both platforms, “Here’s” is the safest choice. The p=0.038 finding for SmartNews does not survive Bonferroni correction at k=5 tests, so treat it as directional guidance, not a confirmed finding.

Number leads (“5 things…,” “3 reasons…,” “10 ways…”) signal structured, scannable content — which SmartNews rewards because readers are more likely to engage with content that promises a clear, bounded payoff. The trend data shows number leads improving over time on both Apple News and SmartNews, making them a relatively safe investment in both channels. The effect is stronger when the number reflects the actual structure of the article (a genuine list) rather than being applied artificially to a narrative piece.

Nature/wildlife/science content has the highest individual-article ceiling in the dataset (53K+ views). The same discovery-frame headlines that work for Apple News also translate to SmartNews because the framing is driven by content novelty (a genuinely rare finding or observation), not by platform-specific algorithm preferences. The one SmartNews-specific addition: avoid question format even for science stories (“Is this the rarest animal ever found?” hurts on SmartNews). Use declarative discovery framing instead (“Scientists find never-before-seen species in [location]”).

Push notifications from news brands compete with dozens of other apps for the same lock screen. Attribution language (“Sheriff says…,” “Witness told reporters…”) signals that the story has a specific, sourced claim — something a reader can’t get from a social post or a headline from a competitor. It also signals credibility: the outlet has a named source, not just a rumor or rewrite. This effect is consistent across news brand notifications (Miami Herald, KC Star, Charlotte Observer, Sacramento Bee) and is the most reliable positive signal in the notification dataset.

Outcome language tells readers the story has reached a definitive state. Crime stories with pending or uncertain status generate less urgency than stories where something has been resolved — an arrest was made, charges were filed, a verdict was reached. Readers have learned to treat breaking crime coverage with skepticism until confirmed outcomes appear. Outcome words signal that the story has cleared that threshold. The combination of outcome word + attribution (“[Name] convicted, prosecutor says”) stacks both signals and produces the highest CTR in the dataset.

Push notifications compete on attention in a context where the reader hasn’t chosen to engage — they’ve been interrupted. A question headline in a notification creates friction: the reader has to hold the question in mind and decide whether the answer is worth tapping. A declarative headline with specific information removes that friction. The question format also has the lowest perceived credibility signal of any formula — it reads more like clickbait than breaking news. This rule applies universally to notifications; the Apple News Featured exception (where editors favor questions) does not transfer to the notification surface.

Entertainment readers follow people, not topics. The possessive structure (“Taylor Swift’s tour cancellation,” “Beyoncé’s statement,” “Tom Hanks’ diagnosis”) signals a personal development for someone the reader tracks. The named entity provides identification; the possessive signals intimacy and personal stakes. This is a different engagement model than news brand notifications, which reward credibility and outcome signals. Numbers in entertainment headlines (“5 things about [celebrity]’s drama”) underperform because they signal listicle format in a context where readers want the specific personal development, not a curated list about it.

The top-performing news brand notification series in the dataset followed a single developing story — a disappearance with a celebrity connection — across multiple updates. Each notification built on the last rather than re-establishing the full context. Readers who engaged with earlier installments already have the context; the notification should deliver the new development (“Guthrie family attorney speaks out,” “Second suspect named”) rather than re-summarizing. This structure rewards loyal notification subscribers while maintaining enough context to be comprehensible to first-time engagers.

Push notifications compete for attention on a locked screen where the full text may or may not be visible. Too-short notifications (≤70 chars) tend to omit the concrete payoff that earns a tap — they signal that something happened without explaining why it matters. Too-long notifications (≥110 chars) risk being cut off mid-sentence on most lock screens, which is worse than being short. The 70–89 char range threads this needle: long enough to include a subject, verb, object, and a single context cue (source or outcome), short enough to display fully. This is a directional finding — treat as a starting target, not a hard rule, and track CTR by length bucket in the ongoing notifications log.