Content Refresh vs Rewrite: How to Choose

If a page is underperforming, you have to decide whether to edit it or write it again. This article is for founders, marketing leads and small content teams who already know a specific page needs work and have to decide how much. It compares the two options, gives a five-step test for choosing, and says what the available evidence does and doesn’t show.

If your own archive is the lever, the next step is deciding which existing pages to improve first. how to choose between a refresh and a rewrite That decision depends on whether an underperforming page needs a light edit or a full rewrite, and the test it offers makes that call easier to justify.

Refresh or rewrite: the short answer

Refresh the page when it still answers the query people are searching and the problems are factual, structural or cosmetic. Rewrite it when the query now wants a different answer, or when the page’s shape can’t be fixed without rebuilding it. Before either, rule out technical causes. Also consider merging the page into a stronger one, or leaving it alone.

Be wary of anyone who quotes a confident return for either option. The two datasets reviewed for this article point in different directions. Neither is an experiment, and no controlled comparison of a refresh against a rewrite on matched pages turned up. So base the choice on what you can observe about your own page.

Refresh vs rewrite at a glance

In this comparison the two sides are editorial interventions, so there are no vendor prices. The table compares time, attention and risk instead. Read the rows as trade-offs, not a ranking.

Factor	Content refresh	Content rewrite
Definition used here	Keep the page and make targeted edits	Discard the existing draft and write a new article for the same query, usually at the same URL
Strengths	Smaller effort; no migration; keeps the structure and links that already work	Can fix a wrong structure or a changed intent that edits can’t reach
URL	Unchanged	Usually unchanged; a new URL is possible but adds redirect work
Evidence behind it	Raptive’s observational analysis of 103,000 refreshes found a 4.50% average increase in Google pageviews. Elmrank’s analysis of its own edits found no difference from untouched posts	Indirect only. Raptive’s largest word-count-change band averaged the biggest gain. Elmrank found rewrites of 500+ words did worse than smaller edits, which its author says could be chance
Fits when	Intent is right; facts, examples or sections are stale; the structure is sound	Intent has shifted, or most headings would have to be replaced
Time and attention	Usually smaller: a focused review of specific sections	Larger: research, drafting, review and any redirect work
Limitations	Small edits may change little; a date change without real edits is the practice Google asks about	A rebuilt page can lose what the old one did well, and the bigger effort carries no proven proportionate return

The two definitions follow the convention in SEO Rocket’s refresh-vs-rewrite guide. It’s an editorial convention, not a standard.

Rule out technical causes first

A page that is blocked, redirected or canonicalised to another URL doesn’t need better writing. Page Refresh AI makes the same point: check the basics before choosing an intervention. Check three things:

Status and indexing. Confirm the page returns a normal response and isn’t blocked from indexing.
The canonical Google selected. A declared canonical is a signal, not a command. Google describes redirects and rel="canonical" as strong signals and sitemap inclusion as a weak one. It also decides which pages are duplicates (Google, canonical documentation). Check which URL Google actually chose.
Robots and sitemap signals. Google notes it may still index URLs disallowed in robots.txt, without their content. Leaving a page out of a sitemap isn’t proof of an indexing problem by itself (same source).

Technical and editorial problems can coexist. Fix what is broken, then decide whether the content also needs work.

Step 1: is the search intent still the one you wrote for?

This test outranks the others. Search the target query and read the top results. If searchers now seem to want a comparison, a tool, a definition or a checklist, and your page answers something else, no amount of sentence-level editing will fix it.

If the intent has shifted, stop here. The choice is a rewrite, a new URL, a merge into a better-matching page, or removal. If the intent still matches, go to step 2.

Step 2: is only the factual layer stale?

This is the clearest case for a refresh. Figures have changed, a product has been renamed, an example no longer applies, or a cited source has moved. The structure is sound and the reader’s questions are still in the right order.

Be realistic about what a factual edit buys. Elmrank analysed edits on three sites it runs (manhwa and anime, video game news, trading card games). An edit counted if 100+ words changed or the title or meta description was new. Of 190 measurable edited posts, 70 (37%) beat comparison posts of similar age on the same site. In the untouched group, 484 of 1,302 post-weeks (37%) did the same. Edited posts drew 11% fewer Search Console clicks than predicted and untouched posts 13% fewer, over four weeks after each edit.

That doesn’t show small edits do nothing everywhere. The sites are young entertainment sites, and the author says another field may see something different. It is a warning that a factual refresh isn’t a guaranteed lift.

Step 3: is the structure salvageable without rebuilding it?

List the questions the page should answer in one column and its headings in another. If the right ground is covered in a poor order, or some sections belong to other questions, reordering and tightening may be enough. If most headings have to be replaced, if the argument no longer holds, or if every section is thin, a rewrite is the honest option.

Once the cause is clear, decide whether the page needs a light pass or a full rebuild, because that choice determines how much work the fix takes. refresh or rewrite It walks through the same heading and section test you can apply to each row before picking the fix.

Step 4: what does the page’s existing performance justify?

Industry guidance often uses search position to choose. InkSTR advises refreshing an article that “still ranks on page one or two, but lower than you’d like”. It advises rewriting one that “has been below position 20 for an extended period”, because then “something is fundamentally wrong with how it addresses the topic, not just a quality gap that a targeted edit can close”. ércule frames it as ROI: “A refresh works when the bones of your content are solid but the execution needs updating.”

If the weaker page holds little unique value, folding its content into the stronger URL is often cleaner than editing both. merge a weaker page into a stronger one It compares refresh and rewrite criteria based on ranking position and content quality, which helps decide when that move makes sense.

Treat these as rules of thumb. The sources reviewed for this article gave no test, sample or calibration behind the cut-offs, so position 20 isn’t a measured tipping point.

Your own Search Console data can help you decide which pages to look at first. For the page, review its queries over a defined period, with clicks, impressions and average position for each. Three limits apply:

Average position is an aggregate across impressions, not a fixed rank.
The query rows you can see are incomplete.
Tables show at most 1,000 rows. Rare queries can be omitted from the table while still counting in chart totals (Search Console Help).

Use this to rank pages for attention. It can’t pick the intervention or predict the result.

Step 5: what does the URL owe you?

The usual advice is to keep the URL, and usually that’s right. Page Refresh AI’s FAQ says: “Usually no. Keep the same URL when the topic and page role are still valid.” ércule gives the same answer for a page whose topic is fundamentally unchanged. Both are vendor advice.

Keeping the URL avoids a migration, redirects and the upkeep they bring. It doesn’t guarantee performance survives a major rewrite. Google supports a narrower point: “301 and other permanent redirects don’t cause a loss in PageRank”. Google says to keep redirects “for as long as possible, generally at least 1 year” in its guidance on site moves and migrations, so don’t stretch either statement into a claim that every signal survives any URL change.

There is a counter-case. Content Marketing Institute’s 2020 account of its republishing practice keeps the original URL and publishes an updated post at a new URL when the original gets substantial organic traffic and the information is still reasonably current. Its URLs include the year and month of publication, so the search result would keep showing the old date. It is a historical account from 2017 to 2020 with no outcome measurement, and it depends on a date-stamped slug that most sites don’t have. It isn’t advice to run two near-identical live pages, which is what Google’s canonical guidance addresses.

If the URL does change:

Map each old URL to its new one.
Use permanent redirects where the old page is being replaced.
Update internal links that point to the old address. How to Build Internal Links for a Small but Growing Blog covers how to find and track them.
Afterwards, check which URL Google selected as canonical.

What the date can and cannot do

Changing the “last updated” date on unchanged content is a common shortcut. Google’s helpful-content guidance asks directly: “Are you changing the date of pages to make them seem fresh when the content has not substantially changed?” (Google’s helpful content guidance).

A visible date that matches a real edit fits Google’s byline date guidance, which recommends a user-visible date with an appropriate label. It doesn’t promise the date will appear in results. No source reviewed shows that bumping the date on unchanged content works.

How big a change is worth making?

Two independent datasets bear on this, and they disagree.

Raptive is an ad network that analysed refreshes on its own publisher pages. Its analysis of 103,000 refreshes covers 103,000 refreshed pages across 38 verticals over a 16-month period (calendar dates not disclosed). It compared Google pageviews, not sessions, leads or revenue, over the 60 days before and after each refresh. Across all 103,000 refreshes, pages averaged a 4.50% increase. In a matched arm of more than 5,500 pairs, matched on site, vertical, content age, baseline traffic and baseline rank, refreshed URLs averaged 28% growth against 9% for similar pages left alone. The pairs were matched, not randomised.

Raptive also grouped refreshes by word-count change: 0–10% averaged 2% growth, 11–30% averaged 11%, 31–100% averaged 11%, and more than 100% averaged 44%. Word count is a rough proxy for effort, since a page can be deeply rewritten without changing length much, and the study reports no costs. Its authors say: “The larger gains from extreme rewrites are worth noting, but they’re not a reason to rewrite every post from scratch.”

Elmrank, described in step 2, found no difference between edited and untouched posts. Its depth result runs the other way: edits of 100–499 words beat their comparisons in 47 of 114 cases (41%), and rewrites of 500+ words in 19 of 61 (31%). The author reports the gap “could be chance (p = 0.25 against the smaller edits)”. It describes its method as “an inference from comparisons, not an experiment”.

These aren’t competing estimates of one quantity. Raptive measured refreshed pages across a large network over a long window. Elmrank measured individual edits on three young sites, with a median post age of 72 days at the time of edit, over four weeks. Populations, windows and edit types differ, and both are observational.

No first-party data from The Blog Writer covers this question either, so nothing here comes from our own refreshes or rewrites. What is missing from the public record is a controlled test. The study that would settle it assigns URLs with the same search intent and similar baseline traffic at random to three arms: a targeted refresh, a full rewrite and an untouched control. It would log editing hours for each page and compare organic clicks and impressions over equal windows afterwards, so you could see return per hour of effort. We found no such study.

What survives is narrower than either headline. Larger edits were associated with bigger gains in one large dataset and weaker results in a smaller one. Neither shows that a bigger intervention buys a proportionate return. Decide from your diagnosis of the page, not from the budget you happen to have.

When neither is right: consolidate, remove or leave alone

Two options aren’t always enough. Ritner Digital frames the choice as delete, refresh, rewrite or consolidate. Page Refresh AI’s matrix also reaches consolidation for overlapping URLs.

A page with the wrong intent doesn’t need better writing. It is competing for a query nobody who wants to buy from you will type. If another page on your site already answers the right query, merging into it can beat both interventions.

Leaving a page alone is also a decision. Elmrank reads its own data as showing edits like theirs “didn’t move a post’s clicks either way in the next four weeks, and whatever the post was already doing carried on”. For a page that doesn’t matter to your business, that is a reason not to spend the week.

The create, refresh or consolidate decision also fits inside a broader process. How to Build an SEO Content Workflow From Research to Publication shows where it sits between intake and briefing.

Worked example: a hypothetical page

This case is illustrative. It isn’t a real customer, a measured result or a page from any site.

Imagine a 1,400-word guide written two years ago on choosing project management software. Several product names are out of date, the pricing paragraph describes plans that no longer exist, and one example uses a discontinued feature. The main query’s average position has slipped, but the page still gets impressions on related queries.

Technical check. The page returns a normal response, is indexable, and Google’s selected canonical is the page’s own URL. No technical cause.
Step 1, intent. The top results are now comparison pages and buyer checklists. The guide was written as a general explainer. The intent has moved toward evaluation, so a refresh would polish a page that answers the wrong question.
Step 3, structure. Most headings would need to change to serve an evaluation reader.

Decision. Rewrite for the evaluation intent and keep the URL, unless a stronger evaluation page already exists on the site, in which case merge the two. The stale facts get fixed along the way, but they aren’t what decided it.

Measurement plan. Record the page’s query set, clicks, impressions and average position before the change, then review the same measures over a fixed period afterwards. Treat any change as an association. Seasonality, competing pages and changes in query mix can all move the numbers.

When review marks a page for rework, decide between a light refresh and a full rewrite before changing anything. Content Refresh vs Rewrite: How to Choose the Right Intervention It also covers recording baseline metrics first so you can judge whether the change actually helped.

The verdict: five variables

Deciding between a light update and a full rewrite is often the hardest call in a backlog. how to rank refresh and rewrite work against your other SEO tasks It gives a clear framework for weighing those options alongside everything else on your list.

“It depends” is only useful when you name what it depends on. Five variables decide the common case:

Intent match. Does the query still want the answer the page gives? If not, don’t refresh. Rewrite, merge or remove.
Factual currency. Is the stale material limited to facts, examples or small sections? If so, a refresh is the natural first step.
Structural salvageability. Can the reader’s questions be answered in a sound order under the current headings? If not, rewrite.
Business importance. Does the page support a product, service or audience you care about? If not, consider consolidating it or leaving it alone. Ahrefs says its own full rewrites are usually done “when you have a secondary business goal—beyond just improving organic traffic”. That is its practice, not a rule, but it is a fair question to ask before you commission one.
Your capacity to measure. Can you record the page’s queries, clicks and impressions before and after, over a defined period? If not, you won’t learn what the work did, so set up the measurement first.

For the common case, where intent still matches and the problems are mostly factual, refresh. Keep the URL, and change the visible date only when the content has really changed. Rewrite only when the intent has moved or the structure can’t be repaired. Justify that larger effort by the diagnosis, not by an expected multiplier.

Other options to consider

If your team lacks the time to research the query, pull the page’s search data and make this call, a managed SEO and content service such as The Blog Writer can take on that work. Whoever does it should show their research and explain each proposed decision, so you can check the reasoning yourself.