Guides

Update vs Rewrite: A Decision Rule for SEO Articles

Read Time8 Min Read
Length1,879 Words
CategoryGuides
Update vs Rewrite: A Decision Rule for SEO Articles
AI Generated

The useful question is not whether an old article should be touched. It is whether the page still has a sound foundation. Across 719 editorial pieces, Fractl found the typical article reached peak traffic a median of four months after publishing, then lost about 76% of that peak within three months. Decay is normal. A full rebuild is not the default response to it. Update when the bones still hold. Rewrite only when they do not. If the core argument still matches search intent and most of the page is still accurate, a surgical update on the same URL is the right job. Artikle’s salvage test is a practical line: if more than 60% of the post is still accurate and well written, update it; if less than 40% is worth keeping, a rewrite is faster than patching. A thin page that can no longer compete with the format now ranking is a separate reason to rebuild, not part of that percentage test. A changed date, a few new sentences, or a new slug is not that decision. Small edits do not count as a refresh, and a new URL throws away the equity the work is meant to protect. The rest of this article is the fail-closed rule for choosing update, rewrite, or leave alone — and for refusing automation that would publish a rebuild without a human gate.

Summary
  • Default to a surgical update on the same URL when the argument, intent, and most of the page still hold.
  • Rewrite in place only when less than about 40% is worth keeping, the structure is broken, or the SERP format has shifted.
  • A date change or a tiny word-count tweak is not a refresh; meaningful expansion is what studies associate with gains.
  • Never change the slug to start over. A same-URL rewrite can still dip for a few days while Google re-evaluates the page.
  • Queue pages in roughly positions 4–20 that already earn impressions. Pages with no demand are a leave-alone or consolidate call, not a rewrite.

Update vs rewrite: what you keep

Update versus rewrite content compared on one desk, light marks versus a crossed-out outline

An update keeps the argument, the structure, and the URL, and replaces only what went stale. A rewrite rebuilds the piece in place because the foundation no longer matches the query. How much of the original is still worth keeping is what separates the two.

If more than 60% of the post is still accurate and well written, do a surgical update. If less than 40% is worth keeping, a rewrite is faster than editing. Refresh when the bones are solid and only examples or stats are stale. Rewrite when the structure is broken, positioning has shifted, or a thin post cannot compete with the guides now ranking.

Do not use a word-count jump alone as the test. Salvage is whether the core claim still answers the query.

Key Takeaway

The split — Keep the same URL and update when most of the page still answers the query; rewrite in place only when the foundation is broken.

Sources

When should you update a ranking page instead of rewriting it?

When to update a ranking page: the update path highlighted over rewrite or leave alone

With that split in place, update the ranking page when search intent and the core argument still hold and the salvage test passes. Do not rebuild a sound page just because it is old. A full rewrite is the right call only when the foundation is broken.

01
Update when the page still answers the query
Intent matches and the core argument is still sound. Refresh stale facts and examples on the same URL.
02
Rewrite only when the foundation fails
The angle no longer matches intent, the SERP format has shifted, or the post is too thin to compete with what now ranks.
03
Refuse a new URL for the same intent
A large word-count jump is not permission to rewrite a sound page, and it is not a reason to abandon the slug.

That restraint is what the gains come from. Workshop Digital compared 95 refreshed pages with untouched controls and saw organic sessions rise from 610 to 3,341, a 447.7% lift, while controls fell 20.2%. Across 41 blog refreshes, most were not complete rewrites, and sessions rose 545.1%.

Key Takeaway

The rule — Keep a ranking page and update it when the argument still fits the query. Rewrite in place only after the foundation breaks, and never mint a new URL for the same intent.

90-day session index (start = 100)

Sources

Should a Rewrite Stay on the Same URL?

Rewrite staying on the same URL while a new competing slug is crossed out

Yes. Keep the slug even when you rebuild the page. Changing the URL during a rewrite is the most common mistake: a new slug breaks internal links and drops the backlink equity this decision is meant to protect.

A same-URL rewrite can still dip 5 to 15 positions in the first 72 hours while Google re-evaluates the page. Expect the dip. Do not fix it by moving the URL or publishing a second page.

When the salvage test passes, the lower-risk path is the surgical update. A successful one typically moves a post 3 to 8 positions and lifts organic clicks 15% to 40%. Cosmetic edits can also fail to help. The risk is changing what Google already trusts, not touching a stale fact on a sound page.

Key Takeaway

Keep the slug — A rewrite stays on the original URL. A short ranking dip is normal; a new slug or a second page throws away the equity you are trying to keep.

How long a surgical update takes, and which positions are worth updating

Surgical update time versus a rewrite, with ranking positions four through twenty queued

Budget the path the salvage test already chose. About 70% of a consultant refresh queue is surgical work at 60 to 90 minutes; about 30% is a rewrite at 3 to 5 hours. If the foundation still holds, do not schedule the longer block. If the structure is broken, a 90-minute pass will not fix it.

70%
of a typical refresh queue is surgical work
30%
needs a full rewrite
4–20
positions where refresh pays off fastest

Positions 4 to 20 are where that refresh work pays off fastest. Pages in that band are the update queue; thin pages that cannot match the current SERP format are the rewrite queue. A same-topic net-new page typically takes months to match an existing URL’s signals, so spend the hours on the URL you already have.

Key Takeaway

Queue the hours — Match the clock to the salvage result: 60 to 90 minutes when the argument still holds, 3 to 5 hours only when the structure is broken, and put the effort on positions 4 to 20 rather than a new URL.

Minutes: surgical update vs rewrite

Sources

Does a date change count as a refresh, and how do you know it worked?

Changing the publish date is not a content refresh next to a truly expanded article

Once a page is in the update queue, the next mistake is treating a new publish date as the work. It is not a refresh, and neither is a 0 to 10% word-count tweak. Studies associate gains with meaningful expansion, not a cosmetic stamp.

In a controlled study of 14,987 URLs, a 0 to 10% change averaged a 0.51 position loss and an 11 to 30% change averaged a 2.18 position loss. Expanding 31 to 100% gained 5.45 positions on average, while never-updated pages lost 2.51.

Across 103,000 refreshes, refreshed URLs averaged 28% Google pageview growth versus 9% for matched pages that were not refreshed, a 19-point lift. By size of change, 0 to 10% edits averaged 2% growth, 11 to 30% and 31 to 100% averaged 11%, and rewrites that more than doubled word count averaged 44%.

Image changes were associated with a 14% pageview increase across 28,000 pages. Title changes and paragraph-count changes each averaged 11%. Treat those as associations, not proof that one element caused the gain.

Judge the edit after the page is reindexed and the new version has had time to settle. A rewrite can take longer to stabilize than a surgical update. Do not call a new date a win.

Key Takeaway

Not a refresh — A date change or a tiny word-count tweak is not a refresh. Call it a win only if positions or pageviews move after reindex, not because the timestamp did.

Pageview lift tied to on-page edits

Sources

Check content decay before you rewrite

Content decay curve with exits to update, consolidate, or leave alone before a rewrite

Run the decay-signal check before you choose a rewrite, so you do not rebuild a URL that is losing clicks for the wrong reason. That diagnosis lives on the Search Console checklist — use the result, and do not treat a traffic slide as a broken foundation.

The Fractl sample also splits by format: how-to content lost 61% of peak traffic, listicles 82%, and research and data pieces 89%. That spread means the rule is recurring work, not a one-time rebuild.

Consolidate or leave the URL alone when there is no demand, or when another URL already owns the intent. No impressions and no query match is a leave-alone or consolidate call, not a rewrite. Whether to brief a new page lives on the refresh calendar — do not spawn a competing page from this decision.

Key Takeaway

Diagnose first — A click drop is a reason to diagnose decay, not to rebuild. No demand means leave the URL alone or consolidate it.

Peak traffic lost within 3 months

Sources

How AI search freshness changes the update versus rewrite decision

AI search freshness allowing fact updates while rewrite, URL change, and publish stay locked

AI search freshness raises the cost of leaving a sound page stale. It does not justify a full rewrite when the angle still matches the query.

AirOps reported that 95% of pages cited by generative-AI systems were updated within the last 10 months. Pages with a visible last-updated timestamp earned 1.8 times more citations than those without, and pages not updated in more than a year were more than twice as unlikely to be cited.

Update facts, examples, and extractable answers. Rewrite only when the angle no longer matches the query. An autoblogging system may draft that surgical update on a URL that already passed the salvage test. It must not publish a rewrite, change the slug, stamp a new date on an unchanged page, or create a competing URL without a human gate.

Key Takeaway

Freshness — AI citation freshness makes a sound page more expensive to leave stale. The response is still a surgical update of facts and extractable answers, not a rewrite or a date stamp.

Sources

How to brief a writer or AI after you choose update or rewrite

Writer or AI brief with update and rewrite modes and the existing URL locked

The brief starts where the decision ends. Name the mode first — surgical update or same-URL rewrite — and do not ask the writer or the model to choose again.

For an update, list the sections to keep, the facts and examples to replace, and the extractable answer to refresh. For a rewrite, state the broken foundation, the new structure, and the URL and intent that must not change.

Automation may draft the update; a human must approve any rewrite before publish. Stop at that handoff, and point the writing contract to your existing article-writing and AI-writer guidance instead of rebuilding those briefs here.

Key Takeaways

[01]
Surgical update vs same-URL rewriteKeep the URL and refresh in place when the core argument and search intent still hold; rebuild on that same URL only when structure, positioning, or competitive thinness breaks the foundation.
[02]
The salvage cutIf more than 60% of the piece is still accurate and well written, a surgical update is faster; if less than 40% is worth keeping, a rewrite is the better use of time. Salvage is whether the claim still answers the query, not a word-count jump.
[03]
What not to changeDo not change the slug, publish a second page, or treat a date stamp or a 0–10% tweak as a refresh. Equity and internal links live on the URL Google already trusts; risk comes from moving that trust, not from correcting a stale fact.
[04]
Where the queue pays offAbout 70% of a typical refresh queue is 60–90 minutes of surgical work. Prioritize positions 4–20; give positions 1–3 light fact updates only; send thin pages that cannot match the SERP format to the rewrite queue.
[05]
How you know it workedMeasure 30 days before against 30 days starting about 14 days after reindex, and allow 45–60 days for a rewrite to settle. Meaningful expansion and real content changes associate with gains; cosmetic edits do not.
[06]
Decay and AI freshnessTraffic decay is recurring work, not an automatic rewrite. No impressions, no query match, or intent already owned by another URL is a leave-alone or consolidate call. AI citation favors recently updated pages, which raises the cost of leaving a sound page stale—it does not justify rewriting an angle that still matches intent.

Run your next ranking URL through the foundation test, then brief a surgical update or a same-URL rewrite before anyone changes the slug or publishes a competing page.

Frequently Asked Questions

When should you update content instead of rewriting it?
Update when the core argument and search intent still hold and most of the page is still accurate. Artikle’s operator rule treats more than 60% still sound as a surgical update, not a rebuild.
When is a rewrite the better choice?
Rewrite when the foundation is broken: less than 40% is worth keeping, the structure or positioning no longer fits, or a thin post cannot compete with the current SERP format. Do that rewrite on the same URL.
Does changing the publish date count as a content refresh?
No. A date stamp without a real revision is cosmetic. RepublishAI found 0–10% content changes averaged a 0.51 position loss, while adding 31–100% more content averaged a 5.45 position gain versus never-updated pages.
Should a rewrite get a new URL?
No. Changing the slug breaks internal links and drops the backlink equity the original URL already earned. Artikle calls a URL change during a rewrite the most common mistake.
Which pages belong in the update queue?
Pages ranking roughly positions 4–20, where Ten Speed says refresh work pays off fastest, especially when they already earn impressions but weak clicks. Pages with no demand should be left alone or consolidated, not rewritten.
Does AI citation freshness mean every stale page should be rewritten?
No. AirOps reported that 95% of pages cited by generative-AI systems were updated within the last 10 months, so freshness matters. Update facts, examples, and extractable answers on a sound page; rewrite only when the angle no longer matches intent.
Sources

You Might Also Like

View All Articles