How brand feature stories run on local news sites.

Published by NewsUSA · Updated November 9, 2026

Read this knowing who wrote it. NewsUSA publishes this page and supplies feature stories of this kind, and some of its own stories are in the sample. The study is about the stories, not who supplied them: it names no news site and no supplier, and it does not compare suppliers.

What we found

• Length. The median story ran 585 words. The middle half ran 468 to 712 words, and 111 of 208 stories (53%) fell between 450 and 700.
• Photos. 201 of 208 stories ran with at least one photo, and 193 had a lead photo above the story. 89 had no photo inside the story text itself.
• Topics. Home and garden (41) and health (41) led, then personal finance (24), food and drink (18) and education and careers (17).
• Publishing day. 109 of 218 stories first went up on a Sunday or a Monday. Only 5 first went up on a Saturday.
• Seasonal timing. Holiday-season stories were already running by September 21. 8 of the 9 were first dated October 8 to 28, 8 to 11 weeks before Christmas.
• Headlines. The median headline ran 10 words. 65 of 218 contained a number, 54 spoke to the reader with "you" or "your", and 15 asked a question.
• One feed, many sites. 208 of 218 stories ran on 13 or all 14 of the sites, which take these stories from one shared feed.

How long the stories run

Most stories sat between 400 and 800 words. Food pieces ran shortest: 7 of the 12 shortest stories were food pieces. Explainers ran longest: 10 of the 14 longest were about health, money or technology. Counts are of 208 stories whose text we could read.

For a story you are writing, aim for roughly 450 to 700 words, the range the feature story template uses. The feature story checker flags a draft that is far outside it.

Photos

Almost every story ran with a photo: 201 of 208. The usual layout was one photo at the top of the story (193 stories), with more photos inside the text in 119. Food stories carried the most, 2.3 photos inside the text on average against 1.2 for all stories. Send a clear photo that shows the subject of the story, with a plain caption and the rights to use it.

Topics

Practical, everyday subjects dominate. 16 stories fit none of the topics below. Each story has one topic, set by keyword rules and then checked by an AI model, so treat the counts as approximate.

When stories go up

109 of 218 stories first went up on a Sunday or a Monday, and very few on a Saturday. The date is the publication date each site stamped on the story, in its own time zone; when a story ran on several sites, we used the earliest.

Seasonal stories run early

Stories tied to a season went up well ahead of it. The one Halloween story was first dated October 4, 27 days before Halloween. The one Thanksgiving story was first dated October 14, 6 weeks before Thanksgiving. Holiday-season stories were running by September 21.

The sections we read list only their most recent stories, so earlier seasonal stories may have rolled off before we looked. Read "first dated" as "already running by." To plan your own dates, use the 2027 feature story calendar; the story angle finder suggests an approve-by date for each dated moment.

Headlines

• The median headline ran 10 words; the middle half ran 8 to 12.
• 65 of 218 headlines contained a number, often a list ("5 ways", "4 tips").
• 54 spoke to the reader with "you" or "your", and 27 contained "how to".
• Only 15 asked a question.

How we did the study

• Where. 14 US local news sites that each have a section of syndicated brand feature stories. Thirteen belong to one newspaper group; the fourteenth carries the same feed. We read that section only, not the rest of each site.
• When. We collected the stories on November 6, 2026 and kept those first published August 1 to October 31, 2026: 218 stories and 2,907 appearances, one for each story on each site.
• How. An automated script made ordinary page requests at a slow, fixed rate, with a user agent that names this site. It followed each site's robots.txt, did not log in, and did not read article text a site keeps for subscribers.
• Length and photos. Counted on the copies whose text was readable without a subscription: 208 stories. Five more were a single image with no text and five had no readable copy; they are left out of these counts.
• Topics. One topic per story, from keyword rules over the headline and opening, then a review of every headline by an AI model (Claude), which changed 55. Treat topic counts as approximate.
• Duplicates. One story is one headline. The same story on several sites counts once. Translated and state-specific versions with different headlines count as separate stories.

Limits

• This is a sample, not a census. The sections list only recent stories, so busy topics reach back days and quiet ones months. Do not read the month counts as a trend: we found 34 in August, 82 in September, 102 in October, mostly because older stories had rolled off. The same goes for the topic counts, because the older months come mostly from slower sections.
• The 14 sites share one feed, so the study shows what that feed carried, not 14 separate editorial choices. It may not describe sites that pick stories one at a time.
• Dates are as each site stamped them. A site may restamp a story.
• Length and photo counts come from two sites' layouts; the same story can look different elsewhere.

Use this data

The figures, the CSV and the charts may be reused with a link to this page. Download the figures (.csv).

Cite this page

Questions

Related guides