A content lead opens a blog post from March 2022, changes the published date to today, swaps one screenshot, and hits update. Two weeks later the ranking has not moved. Down the hall, a colleague rewrites a different two-year-old page, adds a section answering a question the top results all skip, and watches it climb from position 14 to 6 over the next month. Same update button, opposite outcomes. The difference is not that Google rewards freshness. It is that Google rewards freshness for some queries and ignores it for others, and it can tell a genuine revision from a new timestamp.
Content freshness is one of the most over-promised ideas in SEO. The pitch is seductive: keep publishing, keep updating, and Google will keep you near the top. The reality is narrower and more useful, and it starts with what Google actually means by the word.
What Google means by freshness #
Google’s Search Central documentation is unusually direct here. It says: “We have various ‘query deserves freshness’ systems designed to show fresher content for queries where it would be expected.” The operative phrase is “where it would be expected.” Freshness is not a site-wide bonus you earn by editing often. It is a query-level judgment.
Consider the two examples the documentation itself offers. Someone searching for a movie that just came out probably wants current reviews, not articles written while the film was still in production. Someone searching for “earthquake” normally gets preparation resources, but if one just struck, recent news moves up instead. In both cases the page did not get better overnight. The query changed what the searcher needed.
The idea is older than the documentation. In a 2007 interview, Google’s then head of search described a “query deserves freshness” approach that tries to spot when a topic is hot, judging by whether publishers and searchers are suddenly paying more attention to it than usual. Nearly two decades later the framing is the same: freshness is a response to intent, not a reward for recency.
Which queries actually deserve freshness #
If freshness is query-dependent, the first job is sorting your queries. A rough split covers most cases:
- Freshness-hungry: breaking news, ongoing events, “best X for 2026” shopping queries, sports results, software versions, and anything with a year or “latest” baked into it. Here, an old page loses ground fast.
- Freshness-neutral: definitions, how-to steps for stable processes, historical facts, foundational concepts. A search for how to tie a bowline does not deserve freshness, and a 2018 page can still win it.
- Freshness-mixed: a large middle where a current page helps but the query is not time-critical. Being up to date is an edge, not the deciding factor.
This is where “old” and “stale” pull apart, and the distinction is the whole game. If the query is “what is a canonical tag,” a 2019 page that is still accurate can outrank a shiny 2026 rewrite, because the answer has not changed. If the query is “Core Web Vitals thresholds,” a page frozen in 2021 is quietly wrong, and it will slip. Not because it is old. Because it is stale, and staleness is a relevance problem wearing a freshness costume.
What updating a page does, and what it doesn’t #
When you edit and republish, you prompt Google to recrawl and re-evaluate the page. If the content genuinely improved, more complete, more accurate, better matched to what the query wants, that re-evaluation can help you. If the only thing that changed is the date, it generally does nothing at all.
Google says this in plain language. Its guidance on people-first content asks two pointed diagnostic questions: “Are you changing the date of pages to make them seem fresh when the content has not substantially changed?” and “Are you adding a lot of new content or removing a lot of older content primarily because you believe it will help your search rankings overall by somehow making your site seem ‘fresh?'” Google answers its own second question in three words: “(No, it won’t).” The timestamp is not the lever. The substance behind it is.
A two-question filter before you touch a page #
Before you refresh anything, run it through two questions, in order. The order matters, because the first can make the second irrelevant.
First: does this query deserve freshness? If not, recency is not your lever, and you should be looking at depth, relevance, and links instead. Second: is my change substantive? If not, you are polishing a date field and calling it work. The four combinations map cleanly to what you can expect:
| Query deserves freshness? | Change is substantive? | Likely outcome |
|---|---|---|
| Yes | Yes | Best case. A recrawl plus a real improvement is where updates pay off. |
| Yes | No | Little to nothing. Competitors making real updates take the position. |
| No | Yes | Possible gain, but from better content, not from being recent. |
| No | No | No reason to expect movement. Spend the hour elsewhere. |
Most wasted refresh effort lives in the second and fourth rows: real work on a date field, or a bulk pass over evergreen pages that never needed one.
What a substantive refresh actually looks like #
On a single page, a real refresh means correcting figures that have gone out of date, adding a section that answers a question the current top results ignore, cutting advice that no longer holds, and tightening the match between the page and the query. It is editorial work, not a metadata edit.
At the scale of a whole site or a content cluster, the move is triage, not a blanket sweep. Prioritize pages that are both freshness-sensitive and slipping, and leave the stable ones alone. A “update everything” campaign spends its budget on evergreen pages that were fine, and to Google’s systems a wave of edits made with no real change can look like exactly the “seem fresh” behavior the documentation warns against. A page ranking steadily for a stable definition rarely needs an annual rewrite. A “best CRM for small business” page competing on a query with a year in it needs current entries, current pricing, and current screenshots, or it reads as abandoned.
How long before anything moves #
Two clocks run after you publish. The first is crawling: following a meaningful edit, Google generally recrawls within days to a few weeks, sooner for high-authority or frequently-crawled sites, later for small or rarely-updated ones. The second clock is evaluation, and it runs slower. Even after a recrawl, ranking changes can lag, and site-level quality signals can take months to settle. Google has consistently framed content improvement as something that can take time rather than an overnight switch. Expect weeks, not hours. Patience here is not passivity; it is the correct baseline.
Frequently asked questions #
Does adding today’s date to an old post help rankings? Generally not, if the content itself did not change. Google explicitly treats a date-only change as an attempt to seem fresh, and says plainly that it will not help.
How often should I update content? There is no universal interval. Update a page when it is inaccurate, incomplete, or competing on a freshness-sensitive query, and leave it alone when none of those apply. The calendar is the wrong trigger.
Can updating a page hurt rankings? It can. Cutting useful content, drifting away from the query’s intent, or breaking internal structure during a “refresh” are all real ways to lose ground. An update is a risk to manage, not a guaranteed gain.
Does Google prefer fresh content over authoritative content? Only for queries that deserve freshness. For stable topics, established authority and relevance usually win, and a newer page has no automatic advantage.
Freshness is misunderstood because the same action, hitting update, produces wildly different results depending on the query and the substance behind the edit. The honest boundary is this: recency is a lever only when the query asks for it and the change earns it. Everywhere else, the pages that win are the ones that answer the question better, and a date is not an answer. Before your next content refresh, sort your pages by whether the query even rewards freshness. That single filter will save more hours than any bulk update ever gains you.