Content Freshness

Content freshness is how current a page's information actually is — and how clearly that currency shows to search engines. Its weight is query-dependent: for time-sensitive searches, a stale page can't compete no matter how many links it has, while for stable topics, freshness barely registers. Spending update effort without knowing which kind of query you serve is how teams burn quarters rewriting pages that were fine.

What Google documents vs what the industry made up

The documented part: Google has confirmed freshness systems since the 2011 "Freshness Update" built on the earlier Query Deserves Freshness (QDF) concept — the engine detects when a query wants recent results (news, recurring events, frequently updated topics) and boosts recency for those queries specifically. Google also documents how it determines page dates and explicitly recommends against artificially freshening dates without substantive changes.

The made-up part: that changing the "last updated" date is itself a ranking tactic, that republishing old posts with today's date reliably lifts them, or that every page needs an update every N months. Google determines dates from multiple signals — visible dates, structured data, sitemap lastmod, when content actually changed on crawl — and a date bump with no content change is a signal contradiction, not a boost. Sometimes date-bumped pages jump for a few days and settle right back; that round trip is not a strategy.

Freshness weight by query type

One nuance worth calling out: AI search raised the floor. Retrieval systems behind AI Overviews and LLM assistants lean toward recently maintained sources for anything time-flavored — we looked at this in how Perplexity, ChatGPT, and Google AI pick their sources.

How to check it on your own site

1. Find decay first. In Search Console, compare the last 6 months against the previous 6 at the page level. Pages with steadily sliding clicks on stable-demand queries are your freshness candidates.
2. Classify each candidate's query type using the table above. A decaying definition page probably has a different problem; a decaying "best tools" page almost certainly has a freshness problem.
3. Read the SERP. Search the target query and note the dates on the top results. If everything ranking is from the last year and your page shows 2022, that's the answer.
4. Audit your date signals for consistency: visible date, dateModified in Article schema, and sitemap lastmod should agree. Crawl with Screaming Frog and extract them to spot mismatches at scale.
5. Update substance, then signals. Fix facts, refresh examples, replace dead references, add what changed since publication — then let the modified date reflect a change that actually happened.

Common audit mistakes

• Date-bumping without content changes. The classic. At best it does nothing durable; at worst you've taught Google your dates are noise. Fix: no date change without a substantive edit, ever.
• "2026" in the title, 2023 in the body. Users notice, and it shreds trust the moment they hit an outdated screenshot. Fix: the title's promise is a checklist for the body.
• Refreshing by calendar instead of by query class. Rewriting stable evergreen pages quarterly adds churn risk for zero gain. Fix: cadence follows the table above, not a uniform schedule.
• Hiding dates entirely. Some sites strip all dates hoping content reads as timeless. Users and search engines both handle a visible honest date better than a suspicious absence. Fix: show real dates.
• Counting a refresh as done at publish. If clicks don't recover in 4–8 weeks, the update didn't address why the page was losing. Fix: annotate update dates and re-check GSC on a schedule.

Frequently asked questions

Does changing the publish date improve rankings?

Not by itself. Freshness systems respond to fresh content, and Google cross-checks dates against what actually changed. Pair a date change with a real update or skip both.

How much has to change for an update to count?

There's no published threshold, so use the honest test: did the answer a searcher gets materially improve? Corrected facts, current examples, new developments — yes. Swapped synonyms and a new hero image — no.

Should I use dateModified or datePublished in schema?

Both, accurately. datePublished stays fixed; dateModified moves when the content genuinely does. They serve different jobs and Google reads both.

Does freshness matter for pages that rank in AI answers?

Yes, on time-flavored topics. Generative engines penalize staleness at retrieval time — an outdated page often doesn't get pulled into the answer at all, which is harsher than sliding a few positions.

Related reading

Freshness is one half of a pair — the other is knowing which pages legitimately don't need it, which is the domain of evergreen content. To score your own pages, try the content freshness score check, and for a title-tag angle see our case study on freshness signals in titles.