Executive summary
A common piece of SEO advice says Google prefers sites that behave like publishers: post news, run an RSS feed, update often, and rankings follow. Google's documentation supports part of this and contradicts the rest.
The supported part is about speed. Google says RSS and Atom feeds help it "keep your content fresher in its index", that sitemaps and feeds are how it learns about changes, and that it runs "query deserves freshness" systems that favour newer pages for some searches. The contradicted part is the idea that activity itself is rewarded. Google ignores <changefreq> in sitemaps, uses <lastmod> only when it is verifiably accurate, lists changing dates without changing content as a sign of search-engine-first content, and has said through its search advocate that publishing frequency is not a ranking factor.
The working model: feeds and frequent publishing decide how quickly Google sees what you changed. The query decides whether being new matters. A site that understands the difference publishes on the cadence its topic needs. A site that doesn't produces a lot of pages Google has no reason to rank.
The claim, split into its parts
| Claim | What Google documents | Verdict |
|---|
| Google prefers news | Freshness systems favour newer results "for queries where it would be expected" | True for some queries, not for sites |
| RSS feeds help rankings | Feeds help discovery and keep the index current; submission "does not guarantee the indexing" | Helps crawling, not ranking |
| Frequent updates rank higher | <changefreq> is ignored; date changes without substance are flagged | False as stated |
| Fresh dates help | Visible and structured dates should match and reflect real updates | Only when the update is real |
Freshness belongs to the query
Google's ranking systems guide is direct about where freshness comes from: "We have various 'query deserves freshness' systems designed to show fresher content for queries where it would be expected." The word doing the work is expected. Its examples are a new film, where recent reviews push out older production news, and an earthquake, where a generic search tilts towards news after a recent quake.
The 2011 announcement that introduced the modern version of this system was more detailed. It described three kinds of query with freshness needs: hot topics, recurring events and topics that change often, such as camera reviews. It also gave the counter-example. A fast tomato sauce recipe from a few years ago is still useful. A week-old 49ers score is not. In Google's words, "different searches have different freshness needs."
So the question is never whether Google prefers fresh sites. It is whether the queries you want to rank for are ones where searchers expect new information. A page about how to choose a roofing material answers an evergreen question. A page about this year's home renovation rebates answers a freshness question. Both can rank, but only the second one loses ground for being old.
What RSS and Atom feeds actually do
Google's 2014 Search Central post on sitemaps and feeds still describes the model well. XML sitemaps "describe the whole set of URLs within a site, while RSS/Atom feeds describe recent changes." Sitemaps are large and fetched less often. Feeds are small and fetched more often. Google's recommendation was to run both, because feeds "will provide all updates on your site, helping Google to keep your content fresher in its index." It also recommended WebSub, which pushes feed updates to subscribers instead of waiting for them to check.
The same post carries the limit: "submitting sitemaps or feeds does not guarantee the indexing of those URLs." A feed tells Google that something changed. It doesn't make the change rank. Current documentation says the same in shorter form. You can submit a CMS-generated RSS or Atom feed as a sitemap, but "this feed only provides information on recent URLs."
One more piece has gone away. Sitemap advice from that era told site owners to ping Google whenever a sitemap changed. Google retired the ping endpoint in 2023, saying most unauthenticated submissions it received were spam. Submit the sitemap or feed in Search Console or list it in robots.txt, and keep <lastmod> accurate.
For a site that publishes often, a feed is cheap, useful plumbing. For a site that changes twice a year, it does very little. This site runs one Atom feed, at /feed.xml, covering the journal and the newsroom. Its dates come from each post's own published and modified fields, so a redeploy doesn't make every entry look new.
Why "update often" fails as a rule
Three statements from Google close off the idea that activity is rewarded for its own sake.
<changefreq> is ignored. Google's sitemap documentation says so plainly: "Google ignores <priority> and <changefreq> values." Declaring a page as changing daily has no effect.
<lastmod> is used only when it is honest. Google uses it "if it's consistently and verifiably accurate", and it "should reflect the date and time of the last significant update to the page." A site that sets every <lastmod> to today, or bumps it on every deploy, teaches Google to ignore the field. That takes away the one signal that could get a real update recrawled faster.
Cosmetic date changes are flagged as manipulation. Google's helpful content guidance asks: "Are you changing the date of pages to make them seem fresh when the content has not substantially changed?" That question sits in the list of signs that content was made for search engines first.
Publishing frequency has the same problem. John Mueller, Google's search advocate, has been widely reported as saying that frequency is not something the ranking algorithms look at. It can affect how often Google checks a site, which is a crawling effect, not a ranking one. We could only find his remarks quoted in trade press, not in his original posts, so treat the wording as reported.
Where frequency does change outcomes
Frequency still matters. It just works through crawling and through what gets published.
Crawl demand follows change. Google's crawl documentation describes how often it wants to revisit a page in terms of popularity and staleness. A site that changes meaningfully and often gets checked often. A site that rarely changes gets checked rarely, which means a seasonal page or price update published there can take a while to show. Cadence buys faster pickup, not better position.
More honest pages answer more queries. A site that publishes a well-researched piece every week will, over a year, cover more questions than one that publishes four times. That compounding comes from covering more queries. Activity alone doesn't produce it.
Volume without review is a liability. Google's spam policies treat scaled content, meaning many pages made mainly to rank, as abuse no matter how it is produced. Our own newsroom started as an AI-assisted digest three times a week, with every post going through human review. We cut it to weekly because the review gate mattered more than the cadence. On a domain that has already pruned templated pages once, 156 lightly reviewed posts a year was a bigger risk than publishing less.
Should a business site act like a news site?
Google no longer requires publishers to apply for Google News. Sites that follow its news content policies are considered automatically for its news surfaces. News sitemaps exist for this, limited to articles from the last two days and up to 1,000 news entries per sitemap.
That is open to anyone. It is useful to almost no service business. Top Stories and the News tab favour publishers with original reporting and an editorial track record on fast-moving topics. A roofing company's announcement that it hired a new estimator is not going to appear there, and dressing it up as news doesn't change that.
What does carry over is the newsroom habit: dated, attributed, specific posts on things that genuinely changed in your field. A new provincial rebate, a by-law change, a supplier price increase, a product line you dropped and why. Those are freshness-sensitive queries where a small site can be the most current source. They also give AI assistants something concrete and dated to cite. See What Actually Works for AI Citations.
Counter-argument, taken seriously
The obvious objection: sites that publish more often do rank better, and anyone who has watched a busy competitor pull ahead has seen it. Google's statements about frequency describe the algorithm, not the outcome.
The correlation is real. Its causes are mostly the ones above. A busy site covers more queries, gets crawled more, earns more links, because new material gives people something to link to, and builds topic depth that helps every page on it. A site that adds the same number of pages without the research gets the crawling and none of the rest. When frequency correlates with rankings, frequency is not the factor doing the work.
The second objection is that Google has been less than candid about ranking signals before, so its statements shouldn't be taken at face value. Fair, and some of this journal's own pieces argue exactly that, for example on AI Overviews and clicks. But the advice here doesn't depend on trusting Google's account. Accurate dates, real updates and a feed are cheap and correct whether or not freshness is weighted the way Google says. Fake freshness is the only option that depends on Google being wrong about itself.
What everyone is missing
Freshness is a property of the query, not the site. A site doesn't become "fresh". Individual pages answer questions that either reward recency or don't.
An inaccurate <lastmod> costs real updates. If every date is always new, Google stops trusting the dates, and genuine changes wait for normal recrawling.
Feeds matter more for crawling than for readers now. Far fewer people read RSS than a decade ago, but Google still documents feeds as its fast lane for recent URLs.
The newsroom model works for businesses when the news is about the field, not the company. Industry changes earn queries. Company announcements mostly don't.
Practical takeaways
- Sort your target queries by freshness need. Evergreen, recurring, frequently changing and breaking. Only the last three need a refresh plan. See Seasonal Content Wins Impressions on a Calendar for the recurring kind.
- Publish a feed if you publish often. RSS or Atom, submitted in Search Console. Include
<pubDate> or <updated> set to when the content actually changed. - Make every date honest. Visible date,
dateModified and <lastmod> move together, and only when a reader would notice the change. - Choose a cadence you can review. A weekly post someone checked beats three a day nobody did. Cut the cadence before you cut the review.
- Write field news, not company news. Rebates, regulations, prices, standards. Dated, sourced and specific.
- Watch crawl and index timing, not just rankings. Search Console's Crawl stats and URL Inspection show whether updates are being picked up. See Technical SEO.