News Wins the Day. Evergreen Wins the Year.

In 2024 we analyzed the college sports coverage of a national media brand — one of the strongest names in its category, with domain authority most publishers would trade for. The pattern we found is the pattern we now see everywhere: the brand was winning the news. Top Stories placements, headline traffic, a steady drumbeat of articles. And beneath every headline, the evergreen queries — the team page, the stats page, the schedule, the roster, the comparison, the record — were going to sites with a fraction of its authority, because those sites had built the deeper page.

This report is about that gap. It is measurable, it is widening as referral channels shrink, and the instinctive response to it — cut evergreen, chase the day — is the wrong one.

1. The two clocks

News and evergreen content run on different clocks, and the difference is larger than most editorial calendars assume.

The news clock. Parse.ly’s analysis of the top 5,000 posts from each publisher on its network defined an article’s lifespan as the time it took to receive 90% of its 30-day pageviews. The median was 2.6 days. Articles with heavy social traffic lived roughly 14 hours longer. Poynter’s earlier reporting on Parse.ly data found that for most publishers, less than 10% of pageviews came from content older than three days, and for nearly half of them, less than 5%; topic-focused sites did better than breaking-news sites.

The search clock. Ahrefs’ 2025 study of 1.3 million keywords found 72.9% of pages ranking in Google’s top ten were more than three years old, up from 59% in 2017, and the average page in position one was five years old, up from two. Of one million newly published pages, 1.74% reached the top ten within a year (6.11% among English pages with real content) — down from 5.7% in 2017.

A news article is designed for the first clock. It earns its traffic in three days and then, in search terms, it is an old page with a dated headline competing against pages that have accumulated five years of links. An evergreen page is designed for the second clock: it starts slowly and, if it is the deepest page on its subject, it compounds.

2. Every news channel is shrinking at once

The news clock would be a manageable trade-off if news distribution were stable. It is not. Chartbeat data published by the Reuters Institute in January 2026 shows every major referral channel news publishers rely on in decline simultaneously.

Source: Reuters Institute, Journalism, Media, and Technology Trends and Predictions 2026, citing Chartbeat data. Periods differ by platform.

Similarweb’s estimate for organic search visits to news publishers tells the same story from a different vantage: from a peak above 2.3 billion to under 1.7 billion by May 2025, with the share of news searches ending without a click rising from 56% to almost 69% in a year. Our state of organic search report covers the mechanics; the relevant point here is that news content is the category most exposed, because a headline is precisely the kind of thing an AI Overview can restate in a sentence.

3. The gap, measured on one brand

The brand we analyzed in 2024 illustrates what this looks like inside a single site. Around one prominent college basketball program, Google’s index held roughly 28,000 distinct keywords — stats, players, coaches, schedules, records, comparisons, history. The brand’s pages ranked for a small number of them, at an average position of 21. Only about 2% of the keywords placed high enough to earn any traffic at all.

The pages that beat it were instructive. A competing sports site with less domain authority and fewer backlinks held two top-ten positions for the head stats query with a single evergreen page: static rather than loaded by script, fast, dense with team and player statistics, each statistic linked to a deeper page, tabs of related content linking back, a separate URL per season, a glossary of abbreviations. The brand’s equivalent page was dynamic, slow, and indexed for seven generic terms. Google could not see most of its content, and there was nothing beneath it to build a cluster.

Two further findings from that analysis generalize:

• Breadth was missing. More than 86% of the brand’s college sports traffic came from two sports; twenty-plus were covered by rivals. Nearly half of the top-division football programs had no dedicated coverage, and the hundreds of programs outside the top division — with billions in combined athletic revenue — had none.
• Demand was being ignored. Women’s programs accounted for roughly a third of search demand in our keyword sample and produced almost no traffic.

None of this is a criticism of the newsroom, which was doing what newsrooms do well. It is a description of a hole where the evergreen layer should be, and it is the same hole we find under most strong media brands.

4. Why news and evergreen need each other

The mistake is to treat news and evergreen as competing for the same budget. Built together, each fixes the other’s weakness.

	News	Evergreen (data-driven)
Durability	Short-term	Long-term
Author	Human, per article	Template, per data set
URL strategy	New URL per story	Permanent URL per entity
Traffic shape	Fast, then gone	Slow, then compounding
What it gives the other	Fresh links and attention to evergreen pages; a reason to revisit	Topical authority that lifts every new story; a place for readers to go next
Weakness alone	Requires constant publication to hold traffic	Higher upfront and maintenance cost

A ranking release, a data update, a milestone in the standings — each is a news event that points at an evergreen page, and each evergreen page is a stable address that accumulates the links those stories earn. Over time, the cluster’s authority is what lets the next story rank quickly. Ahrefs’ data on age and rankings is the mechanism: the pages that win are the ones that have been accumulating signals for years, and only a permanent page can do that.

5. Brands are cutting the wrong thing

The Reuters Institute’s survey of 280 media leaders recorded what publishers plan to publish less of in 2026. On a net basis — the share planning to increase minus the share planning to decrease — service journalism scored −42, general news −38, and evergreen content −32.

Source: Reuters Institute, Journalism, Media, and Technology Trends and Predictions 2026; net score = share planning to increase minus share planning to decrease, n=280.

The reasoning is that these are the categories AI summarizes most easily. That is true of thin evergreen: the explainer that restates what everyone knows, the listicle without data behind it, the service piece that answers a question in one paragraph an assistant can paraphrase. It is not true of evergreen built from data — the stats page, the ranked comparison, the entity profile with fifty sourced figures and a chart. Ahrefs found AI-cited pages are 25.7% fresher than organic results and the GEO benchmark found pages with added statistics gained 33% visibility in generative answers. The evergreen that survives is the evergreen that holds facts an answer has to cite.

Methodology and caveats

Lifespan and evergreen-share figures come from Parse.ly’s network analysis (top 5,000 posts per publisher over two months; lifespan defined as time to 90% of 30-day views) and Poynter’s reporting on Parse.ly data; both date from 2014–2015 and remain the most-cited public measurements. Page age and time-to-rank are from Ahrefs’ May 2025 study of 1 million random new URLs and 1.3 million keywords. Referral declines are Chartbeat data as published by the Reuters Institute; note the periods differ by platform. The single-brand analysis was conducted by us in mid-2024 using SEMrush and Ahrefs keyword and traffic estimates for one national sports media brand and its competitors; the brand is not named, the figures are estimates from third-party tools, and traffic attribution by URL pattern under-counts pages whose URLs did not include the sport. Conclusions about the pattern do not depend on the exact figures.

What to do about it

1. Give every recurring news subject a permanent home. A team, a company, a school, a product, a candidate: one entity page per subject, at a stable URL, that every story links to and that links to the depth beneath it.
2. Make the evergreen layer data-dense. Stats, records, rankings, comparisons, history — sourced and dated. This is the content assistants cite and rivals with less authority are currently using to beat you.
3. Render it statically and completely. If your evergreen page loads its facts with a script, search engines may not see them. The competitor’s static stats page beat a stronger brand’s dynamic one for exactly that reason.
4. Turn data updates into news. Each refresh is a story: new rankings, a record broken, a metric up or down. It gives the newsroom a hook and the evergreen page a link.
5. Cover the whole category, not the top of it. The long tail of teams, programs, products, and places is where authority is built cheaply and where the demand — often a third or more of it — is unserved.
6. Measure on the second clock. Judge evergreen investment over eighteen months, not eighteen hours. The pages that dominate search are years old for a reason.

Sources

1. Parse.ly — The data behind the average article lifespan — Median lifespan 2.6 days; top 5,000 posts per publisher, two-month window
2. Poynter — Should publishers be taking better advantage of evergreen content in their archives? (2014, Parse.ly data) — Share of pageviews from content older than three days
3. Ahrefs — How long does it take to rank in Google? (Stox, May 15, 2025) — 1 million new URLs; 1.3 million keywords
4. Reuters Institute — Journalism, Media, and Technology Trends and Predictions 2026 (Newman, January 2026) — Chartbeat referral data; survey of 280 media leaders
5. Press Gazette — How Google AI Overviews is fuelling zero-click searches for top publishers (July 2025, Similarweb data)
6. Ahrefs — Do AI assistants prefer to cite fresh content? (July 28, 2025)
7. Aggarwal et al. — GEO: Generative Engine Optimization (KDD 2024)
8. Media Factual — analysis of a national sports media brand's college coverage (June 2024, SEMrush and Ahrefs estimates) — Brand not named; figures are third-party estimates