What a content refresh actually is
A content refresh is the deliberate re-audit and rewrite of an existing page to match the current best version of itself. It is not a one-line edit to update the published date. It is a planned exercise: pull the rank, scan the SERP, identify what has changed since the page was written, and update the body, the schema, the internal links and the FAQs to reflect that change.
The unit of refresh is one URL. The trigger is usually one of five things: the page has gone stale, the rankings have slipped, the SERP intent has shifted, the competitor entity coverage has moved, or the annual refresh schedule has come around. Each trigger maps to a slightly different refresh scope, but the workflow underneath is the same.
The aim is to keep the page's existing link equity and age signal while bringing the body up to current best practice. The page's URL stays. The internal-link inbounds stay. The schema stays (with the dateModified updated). The body and the parts of the schema that should change get the rewrite.
Why refresh almost always beats new
Three reasons refreshing an existing page tends to outperform writing a new article on the same topic.
- The existing URL has link equity. Inbound links, both internal and external, take months to accumulate and are non-trivially expensive to earn. A new URL on the same topic starts with zero. The refresh keeps the equity; the new article discards it.
- The existing URL has age signal and rank history. Google's freshness systems weight a steadily-updated page differently to a brand-new page on the same topic. The "page that has been here for two years and was just updated" reads as a more durable asset than "page that appeared yesterday".
- Refreshing avoids accidental cannibalisation. Writing a new article on a topic an existing page already covers is how keyword cannibalisation happens. Two pages compete for the same intent, Google picks one (usually the wrong one), and both pages rank worse than a single consolidated page would have. See how to fix keyword cannibalisation for the diagnostic.
The exception is when the existing page has a structural problem (wrong intent, wrong content type, badly fragmented information) that cannot be fixed with a body rewrite. In those cases the right move is sometimes to write a new page and 301-redirect the old URL to it. See 301 vs 302 redirects for the redirect logic. Most pages do not need that; most pages need a refresh.
When to trigger a refresh
Five triggers are worth automating into the content map.
Trigger 1: The annual schedule
Every evergreen page has a refresh date 12 months after publication. The refresh date moves forward 12 months each time the page is refreshed. Pages that pass the refresh-worthiness check on that date get refreshed; pages that fail go into the pruning queue.
Trigger 2: A ranking drop
A page that has dropped more than five positions over four weeks for its primary keyword is showing signs of decay. The decay is often an intent shift in the SERP rather than a problem with the page; the refresh check confirms which.
Trigger 3: A SERP intent shift
Sometimes the dominant format of a SERP changes (the ten-blue-links page that used to be commercial intent has flipped to informational, or vice versa) and an existing page on the topic suddenly mismatches the new dominant pattern. The fix is to refresh the page to match the new dominant format.
Trigger 4: New competitor entity coverage
A competitor pages a major update that adds entities yours does not cover. Yours starts to look thin relative to the new SERP. A refresh closes the entity gap.
Trigger 5: Stale statistics or screenshots
Pages that reference dated statistics, software screenshots, or pricing get unreliable fast. A refresh of just the citation set, screenshots and dates often lifts the page back to current.
The six-step refresh workflow
The whole workflow takes two to three hours per page once you have run it a few times.
Step 1: Confirm the page is worth refreshing
Three checks. Does the page currently rank in the top 30 for its target keyword? Has it picked up impressions in GSC over the last 90 days? Does it still match the current dominant intent of the SERP? Yes to all three means refresh. Two out of three is a judgment call. One or zero means the page is a candidate for pruning, not refresh. See content pruning explained for that path.
Step 2: Pull the baseline
Record the current rank, the keywords the page already ranks for (from GSC), the impressions and clicks over the last 90 days, and any conversion data from GA4. This is the baseline you measure the refresh against four and twelve weeks later.
Step 3: Re-scan the SERP
Open the top ten ranking URLs for the primary keyword. Compare the dominant page format, the entity coverage, the headings, and the average word count against the page being refreshed. Note the gaps. The gaps are the refresh scope.
Step 4: Update the body
Rewrite the outdated sections, close the entity gaps, refresh the statistics and screenshots, add new FAQs that match new related-search questions, and update the schema's dateModified. Keep what was working: the headings that match SERP language, the entities the original covered, the parts the existing rank confirms Google liked.
Step 5: Update the internal-link graph
Add inbound links to the refreshed page from any siblings or cross-pillar pages published since the original went live. Update the outbound links on the page to current targets (the destination pages may have changed URLs or been refreshed themselves). See internal linking strategy for the broader pattern. Verify the pillar page still links to the refreshed page in its cluster index.
Step 6: Track the post-refresh response
Re-pull the rank and traffic four weeks after the refresh, then again at twelve weeks. The four-week pull is the early signal; the twelve-week pull is the consolidated read. Log the result against the baseline so the refresh either gets credited or the workflow gets adjusted.
What to update on the page
Eight things are worth checking on every refresh:
- The title tag and meta description, against current SERP CTAs. See meta description and title tag best practice.
- The H1 and the dek paragraph, against the current dominant intent.
- The TL;DR, against any updated working positions on the topic.
- The body H2 sections, against the must-cover entities from the new SERP scan.
- The FAQ block, against the current People Also Ask and related searches.
- The schema, especially
dateModified,author(if changed), and any HowTo or FAQPage entries that should be updated. - The internal-link graph: inbound links from new siblings, outbound links to current targets, the canonical tag if relevant.
- The CTA block, against any new service-page URLs or current campaign offers.
What not to touch unless there is a real problem: the URL (changing URLs discards link equity), the canonical (changing canonicals can flip indexation), and the publication date in the visible byline (Google reads dateModified from the schema, not from the visible date alone).
Common mistakes
- Confirming the page is worth refreshing before starting the work.
- Recording a rank and traffic baseline before the refresh.
- Updating the body to close real entity gaps, not for the sake of editing.
- Keeping the URL, the canonical and the inbound link graph intact.
- Updating the dateModified in the schema, not just the visible byline date.
- Tracking the post-refresh result at four and twelve weeks.
- "Refreshing" by changing the byline date with no content delta. Google reads the actual change.
- Removing entity coverage the original page had. The page loses the topical depth that earned its current rank.
- Changing the URL because the new title sounds nicer. The link equity vanishes.
- Refreshing pages that have never ranked. Those are pruning candidates, not refresh candidates.
- Writing a new article on the same topic instead of refreshing the existing one. Hello cannibalisation.
- Skipping the SERP re-scan. The whole point of the refresh is to match the current SERP.
Tools and a checklist
- Google Search Console. The Performance report is where you see the page's current keywords, impressions and CTR. The free way to find refresh candidates: any page where impressions are climbing but CTR or rank is not.
- Google Analytics 4. For the conversion data tied to the page. A page that ranks well but converts poorly is a CRO problem, not a refresh problem.
- An entity-extraction tool. InLinks, Surfer, Frase, Clearscope, or a structured LLM prompt. Used to extract the entity gaps between the page being refreshed and the current top three SERP results.
- A rank tracker. Spreadsheet or a $30-a-month tool. Tracks the primary keyword rank weekly so the refresh effect is visible.
- A refresh log. One row per refreshed page, columns for baseline rank, refresh date, four-week rank, twelve-week rank, traffic change. This is the feedback loop that improves the workflow over time.
For the wider on-page work the refresh feeds into, see content depth and word count myths. For the technical-side checks that should accompany a refresh, see canonical tags explained and how to run a technical SEO audit. For our website audit service, we run the refresh-candidate identification at scale.
Perth and WA context
Three patterns from running content refreshes for Perth and WA clients.
Trade and service pages decay on suburb-name changes. When the ABS or the local council renames a suburb boundary, or when a new estate gets named (Alkimos, Eglinton and Yanchep all expanded in the 2020s), service pages that targeted the old names start to underperform. The fix is to refresh the suburb mentions and add the new estate names. See SEO Joondalup and the local keyword research chapter for the modifier framework.
Mining and resources content decays on commodity cycles. A WA mining services page written during a iron ore boom and not refreshed reads as outdated when the cycle turns. Refreshes that update commodity references, project names and named workforce statistics keep these pages credible. See mining SEO and SEO Karratha for the regional pattern.
Healthcare and legal pages need credentialled reviewer sign-off on every refresh. For YMYL topics, the refresh is also a re-verification of the medical or legal claims on the page. The refresh workflow should include a named reviewer (registered practitioner or admitted lawyer) sign-off line. See healthcare SEO and legal SEO for the category-specific patterns, and E-E-A-T explained for the wider framework.
For the strategic frame, see the Content Strategy pillar. For the brief template the refresh should be re-checked against, see how to write a content brief. For the pruning path that some refresh candidates end up on, see content pruning explained. For the wider authority signal refreshes feed into, see topical authority explained.